Datasets:

bzantium

/

LongPerplexity

like 0

accurate distances to planetary nebulae ( pne ) are critical for calculating the size , mass , luminosity , age , and other properties of pne and their central stars . distances to pne are also crucial when studying properties of the galaxy such as the galactic rotation , and scale height of their progenitor stars . in addition , distances are required to ascertain the planetary nebula luminosity function in order to use pne as standard candles for cosmological distance studies ( ciardullo _ et al . _ 1988 ) . despite being bright and containing ( dominated by ) a rich line spectrum , pne have distances that are remarkably ill - constrained . in fact , distances to many extragalactic pne are better known than distances to most galactic pne . many different distance measurement techniques have been applied to pne in the past . most of have been statistical in nature , presuming that some property is common to the entire population ( hajian & terzian 1996 ) . the most commonly used pn distance estimation technique is known as the shklovsky method ( shklovsky 1956 ) , and assumes a constant ionized gas mass for all pne . unfortunately , these statistical techniques yield errors that are large , often a factor of two or more , as reviewed by terzian ( 1997 ) . in the case of the shklovsky method , @xmath1 , where @xmath2 is the nebular distance and @xmath3 the ionized mass . since pne progenitor masses can span a factor of @xmath410 , and assuming that @xmath3 scales with the progenitor mass , distance errors of several hundred percent are typical . the most reliable distance measurements for pne so far have been provided by measuring their expansion parallaxes ( terzian 1997 ) . pn shell expansion velocities are typically @xmath5 km s@xmath6 , or @xmath7 mas yr@xmath6 at distances of @xmath8 kpc ( mas = milliarcsecond ) . expansion parallax distance determinations require two epochs of image data separated by a time baseline @xmath9 yr , which can be used to measure the angular expansion `` parallax '' , @xmath10 , of a fiducial feature . the physical expansion velocity of the features must be spectroscopically obtained , and converted to a tangential velocity , @xmath11 ( _ i.e. _ , the component of the velocity normal to the line of sight ) using a spatiokinematic model of the nebula . the distance to a pn can then be computed directly by dividing @xmath11 by @xmath10 ( hajian , terzian , & bignell 1993 ) . distances to nine pne have been previously measured using radio images from the very large array ( vla ) at different epochs ( masson 1989a ; masson 1989b ; gomez _ et al . _ 1993 ; hajian , terzian , & bignell 1993 ; hajian , terzian , & bignell 1995 ; hajian & terzian 1996 ; haryadi & seaquist 1998 ) . one past attempt has been made to measure the expansion of ngc 6543 using the vla . however , this expansion parallax measurement was unsuccessful due to insufficient image quality in the radio - wavelength data ( hajian & terzian 1996 ) . the many well - defined , optically bright features of ngc 6543 serve as ideal fiducials in measuring its angular expansion . the ability to measure the expansions of individual features within this extraordinary nebula provides us with a way to sort through the nebula s recent history , to probe the mass ejection behavior of its nucleus , and to directly determine the distance to this pn . as discussed by hajian , terzian , & bignell ( 1995 ) , the optical expansion of a pn is particularly detectable and measurable with high accuracy ( @xmath1225% ) using the hubble space telescope ( hst ) . in this paper , we report the detection of the systematic expansion of ngc 6543 between 1994 and 1997 from hst archival observations . in order to quantify the magnitude of the angular expansion rate , we present three separate astrometric analyses of these recorded images . these observations , along with a spatiokinematic model of the nebula derived from independent measurements of the doppler velocity , permit three direct distance determinations , all of which are in excellent mutual agreement . in addition , the expansion pattern reveals that the kinematic ages of features approximately scale with their angular size . this provides insight into the highly organized and orchestrated mass loss history of the nucleus , presently a wr - of star ( heap & augensen 1987 ) . this two - in - one paper discusses both the distance to and the nebular evolution of ngc 6543 . the bright core of ngc 6543 just fills the higher - resolution planetary camera ( `` pc '' ) of the wide field and planetary camera 2 ( `` wfpc2 '' ) whose projected pixel size is 45.5 mas . archived unsaturated hst narrowband images from go program 5403 and corresponding calibration images from go program 6943 were used for this study ( table 1 ) . the data epochs are separated by 35 months . images through f502n ( hereafter `` [ o iii ] '' ) , f656n ( `` h@xmath13 '' ) , and f673n ( `` [ s ii ] '' ) filters were obtained at both epochs . the use of the same instrumentation for the two epochs , albeit at differing orientations , reduces systematic errors . measurements of the expansion were based solely on images in the [ o iii]@xmath14502 nm and [ s ii]@xmath14673 nm lines . the [ o iii ] image is optimal for this study since it is characterized by well - defined , bright structures with excellent signal - to - noise ratios in the wfpc2 images . the h@xmath13 images appear essentially redundant to their [ o iii ] counterparts . for this reason the h@xmath13 images were not analyzed . the [ s ii ] images were used to confirm the expansion measurements from the [ o iii ] images , and to investigate the proper motions of specific , bright , low - ionization features . the pointing and orientation of the hst were different for the 1994 and 1997 observations . therefore , after cosmic rays were removed and the relative image intensities were calibrated , the images had to be regridded to a common center and orientation . we used the iraf tasks `` drizzle '' , `` imshift '' , and `` rotate '' to remove geometric distortions from , align , and rotate the images . the shifts and rotations were dithered until patterns of residuals of the corresponding images from the two epochs showed no trace of alignment or orientation errors . we estimate the accuracy of the alignment to be @xmath15 0.05 pixel ( 2 mas ) and the orientation to be @xmath16 0.1@xmath17 . blinking of the 1994 and 1997 images shows clearly that the nebula has changed its scale size ( _ i.e. _ the pattern changes are radial ) . our strategy is therefore to average shifts in the locations of fiducials on equal and opposite sides of the nucleus ( along radial lines ) using various methods . the angular expansion rate , @xmath10 , is best measured using fiducials with narrow peaks and/or sharp edges . in our analysis , we concentrated on the edges of several nebular features shown in figure 1 . using the nomenclature of miranda & solf ( 1992 ; hereafter `` ms92 '' ) , we analyzed the `` p.a . 25@xmath17 ellipse '' ( e25 ) and the `` p.a . 105@xmath17 ellipse '' ( e105 ) . in the outer regions , we were able to measure the motion of the bright polar caps and the faint polar condensations , referred to as ( d - d@xmath18 ) and ( f - f@xmath18 ) , respectively , in ms92 . these structures are identified in figure 1 . we used three methods to extract @xmath10 from the images . the most direct approach , called the `` profile method '' , is to compare one - dimensional radial brightness distributions , or `` profiles '' , from images of two epochs taken with the same filter . the relative positions of sharp nebular features are found by analyzing the flux gradients in the radial profiles in the residual image of the two epochs , as pioneered by masson ( 1986 ) . this method allows @xmath10 to be estimated along many one - dimensional `` cuts '' through the nebula . the random errors appear to be dominated by photon statistics . a summary of the `` profile method '' can be found in hajian , terzian , & bignell ( 1993 , 1995 ) and in hajian & terzian ( 1996 ) , referred therein as the `` expansion parallax algorithm '' . the analysis for hst images is identical to the analysis for vla data except that hst data are not obtained in the fourier plane . the method requires high signal - to - noise ratios in the residuals after the images of two epochs are subtracted , and it assumes no change in surface brightness or nebular shape . using this method , @xmath10 can be computed for any bright feature intersecting a radial cut : @xmath19 where @xmath20 is the flux gradient of the nebular feature and @xmath21 is the peak value of the difference map near the nebular feature . we take the uncertainty in expansion measurement along an individual profile to be equal to ( 1 ) the difference between expansion rates inferred from the corresponding features on opposite sides of the nucleus , or ( 2 ) the unweighted mean and standard deviation of multiple cuts intersecting an extended structure , where appropriate . the `` radial fitting '' method is a similar and somewhat redundant technique . the positions of bright nebular features along 1-d radial flux profiles are measured by performing gaussian fits , by evaluating the flux - weighted centroid of a fiducial s peak , or by graphically measuring the shift of a sharp edge in each of the two epochs . the radial fitting method has a somewhat increased potential for yielding expansion estimates for nebular features that are too faint for similar measurements with the profile method . this is due to the increased noise level in a difference map relative to the epoch maps ( @xmath22 for epoch maps with equal noise levels ) . the random errors of the measurement are dominated by the uncertainties in the fit , which include the judgement involved in defining a fiducial . for fits to peaks , the measurements of a gaussian feature in a complicated environment are sensitive to choices of the edges and base of the fitted gaussian curve . great care is necessary to use the same procedures for the images of both epochs . finding the relative positions of sharp edges is straightforward . it is done by overplotting the 1-d flux profiles of the two epochs , and graphically measuring the radial displacement at multiple positions along the steep flux dropoffs of the edges . we assume a minimum 20% uncertainty in measured expansion rates since we only employ the radial fitting method once for each fiducial in each filter . the `` magnification method '' , is the simplest but least accurate of the methods . a magnification factor is applied to the first - epoch image which minimizes the rms of the difference , or residual , image in the regions of the bright fiducial features . for each emission line , [ o iii ] and [ s ii ] , we used the iraf task `` geotran '' to magnify the 1994 image by various magnification factors , @xmath23 , in the range of 1.001 to 1.005 . for radial motions with no brightness changes , a given feature will disappear in the residual image after the magnified 1994 image is subtracted from the aligned 1997 image . we can then compute @xmath10 using the following equation : @xmath24 where @xmath25 is the angular distance from the nucleus to the feature . the advantages of the magnification method are that the entire image is utilized , that the intuitive pattern recognition capability and judgement of the brain can be exploited , and that the residuals may uncover deviations from simple radial expansion in the complex nebula that would not easily be recognized in one - dimensional analyses of the brightness distribution . the major disadvantage to the magnification method is that errors are difficult to quantify . we assume an uncertainty of 25% in our magnification method expansion rates . once @xmath10 has been measured using the above methods , and @xmath11 has been determined based on spectroscopic observations and a spatiokinematic model used to convert radial into tangential velocities , the distance to the nebula in parsecs can be calculated : @xmath26 the tangential component of the expansion velocities of various features were taken from the detailed observations and the nebular models of ms92 . in addition , the kinematic age , @xmath27 , of a fiducial is given by : @xmath28 we also note that a more accurate distance can be computed by taking all of the measured values of @xmath10 into account ( _ i.e. _ , @xmath10 at multiple p.a.s for multiple features , rather than just the minor axis of e25 , which our distance is based upon ) . however , since ( 1 ) the only available kinematic information for ngc 6543 was obtained in an emission line , [ n ii]@xmath29 , that is different from the line(s ) used to determine the astrometric expansion of the nebula , [ o iii]@xmath30 and [ s ii]@xmath31 ; and ( 2 ) uncertainty in the adopted expansion velocity is approximately @xmath1610% , we did not feel it was necessary to fit a complex spatiokinematic model to the data . this will be more appropriate when we obtain positionally resolved velocities in the summer of 1999 . we next discuss the measurements of @xmath10 using all three methods . the results are compiled in table 2 . also shown in the table is the expected value of @xmath10 assuming uniform nebular expansion whose rate is calibrated to the average measurement of @xmath10 along the minor axis of e25 ( see below ) . for the [ o iii ] images , we applied the profile analysis to various radial cuts through the nucleus in @xmath32 increments , as shown in figure 2 . the resulting angular expansion rates are displayed in units of mas yr@xmath6 wherever the profile intersects a bright , thin feature . for the [ s ii ] images we measured @xmath10 along radial cuts through the central star at intervals of 2 degrees as shown in figure 3 . the residuals of the magnification method are shown in figures 4 and 5 , and the corresponding values of @xmath10 are shown in table 2 . the changes in the nebular structure are seen most clearly in the residual images of the unmagnified 1994 - 1997 image pairs . see the upper left panels of figures 4 and 5 . although the residuals are only about 1% of the image brightnesses , the patterns of change are quite clear . none of the residuals from any of these techniques yield evidence of nonradial motions . the best estimates of the distance rely on sharply defined symmetric structures such as spheres or prolate ellipsoids with sharp edges . we capitalize on a geometric trick : the measurement of @xmath10 along the minor axis of a prolate ellipsoid is ideal for deriving an expansion distance since no geometric correction for inclination is required . e25 is the optimum structure for expansion studies because of the prominent and sharp edges along its minor axis and its well - measured expansion velocity . e105 is somewhat more diffuse than e25 , and ms92 s doppler velocity mapping has shown it to be a tilted circular ring rather than an prolate ellipsoid ; however , it serves as a useful consistency check . _ e25 : _ for the profile method , the best measured expansion rate of e25 s minor axis in the [ o iii ] images is @xmath33 mas yr@xmath6 . adopting ms92 s expansion velocity of 16.4 km s@xmath6 from the kinematic model described earlier in this section , and a 10% velocity uncertainty results in a distance d@xmath34 pc . based on a size of @xmath25 = 36 for the minor half - axis of e25 , the kinematic age is t@xmath35 yr . the radial fitting method was used on both the [ o iii ] and the [ s ii ] images . using the iraf command `` splot '' , gaussians were fitted to measure the shift of the fiducial features between the two epochs . for [ s ii ] , the radial cuts were offset slightly to avoid a bright cosmic ray @xmath16 15 to the south of the central star . we measured an angular expansion rate @xmath36 mas yr@xmath6 for the minor axis of e25 using this method . the result agrees to within 10% with the result of the profile method . the kinematic model described above implies a distance d@xmath37 pc . the corresponding kinematic age t@xmath38 yr . the magnification method gives very consistent results . the expansion of most of the inner core ( within the intersection of e25 and e105 ) is well characterized by a single magnification factor midway between 1.0025 and 1.003 . we adopt a value of 1.00275 and an error of 25% ( _ i.e. _ , @xmath39 mas yr@xmath6 ) . judging from the darkened residuals along e25 , there may have been a slight decrease in surface brightness as well . the distance derived from the magnification method results is d@xmath40 pc . we derive a kinematic age t@xmath41 yr . combining the results for the minor axis of e25 with equal weights and averaging the errors , we derive a best - estimate distance @xmath42 pc and a kinematic age @xmath43 yr . this conservative approach of averaging our uncertainties together is warranted because our measurements all rely on the same images and radial velocities , and are therefore not statistically independent . next we consider the measurement of @xmath10 along the major axis of e25 . the structure at the ends of the major axis is much more poorly defined than along the minor axis . accordingly , the various techniques produce somewhat scattered results . we measure 4.51 , 1.95 and @xmath165 mas yr@xmath6 using the profile , radial , and magnification methods respectively . the disagreement in the results is substantial . the important point is that on average all of the methods fall about a factor of two short of the expected result for uniform ( `` hubble law''-like ) expansion , 7.7 mas yr@xmath6 . in other words , the kinematic age of the major axis of e25 is @xmath44 yr . similarly large expansion ages are obtained not just along the major axis of e25 but throughout the periphery of the core of ngc 6543 by almost all methods . whether the peripheral structures are truly older than the material along the minor axis of e25 , or whether the outer material is coeval and decelerated is an issue that will be addressed in 4 . in principle , the major axis of e25 can be exploited to give a check on the expansion distance since we are able to measure its expansion rate @xmath10 . after some consideration we decided this would be a mistake for several reasons : ( 1 ) @xmath10 must be corrected for inclination , and the angle is uncertain ; ( 2 ) the structure of e25 along its major axis is irregular and faint ; ( 3 ) our data show that the ellipse e25 does not expand uniformly , as ms92 assumed , and that the major axis deviates the most from the uniform expansion pattern ; and ( 4 ) there is some doubt that the tips of e25 are actually part of the same physical structure as the minor axis ( balick , wilson , and harrington , in progress ) . e105 has almost the same projected size and shape as e25 , so it is tempting to use it to derive an independent distance to ngc 6543 . however , e105 is lumpier and more diffuse , and measuring its size and expansion rate is commensurably more uncertain . in addition , its expansion velocity is not as well known . all three methods again yield consistent results on the expansion rate of e105 along its major and minor axes . moreover , those results concur with the expansion rates derived for e25 . that is , the values of @xmath10 for e105 follow a uniform hubble - law expansion relation based on the angular size and @xmath10 that we measured for the minor axis of e25 . this is comforting . this means that e105 and the minor axis of e25 share the same kinematic ages . let us now employ the ms92 model wherein e105 is a planar circular ring which surrounds e25 and shares the same symmetry axis . since e105 is a tilted circular ring , the ring s expansion rate measured along its major axis , @xmath45 mas yr@xmath6 , needs no geometric correction for tilt . ( this value is an average of the results of the various methods in table 2 . ) the measured expansion velocity , is 28 km s@xmath6 ( ms92 ) , and the derived distance @xmath46 is @xmath47 pc , where the errors in the expansion velocity ( which may be substantial for this feature ) have been ignored . satisfyingly , d@xmath48 is consistent with @xmath49 . _ caps @xmath50 : _ the caps are lumpy complexes which lie within thin regions in the outermost extremities of the nebular core . the locations of the caps at the extreme edges of the core suggest that the pair of them lie in or near the plane of the sky . this means that their doppler shifts should be , in principle , unmeasurable , and no distance can be estimated from them . on the other hand , this same geometry guarantees a straightforward computation of the kinematic age once the angular expansion rate of the caps has been measured . if ngc 6543 expands uniformly at the rate determined from the minor axis of e25 then we would expect the caps to exhibit angular expansion rates @xmath51 10 mas yr@xmath6 . using the profile method , we detected the signature of angular expansion from the polar caps in both [ o iii ] and [ s ii ] . in the case of [ o iii ] , the expansion rate of the polar caps averages to 6.70@xmath01.37 mas yr@xmath6 . in the case of [ s ii ] , the average expansion of the polar caps is @xmath52 mas yr@xmath6 . the discrepancy from uniform expansion lies well outside of measurement errors . these results are verified by the radial method , which gives expansion rates of @xmath53 mas yr@xmath6 and @xmath54 mas yr@xmath6 for [ o iii ] and [ s ii ] , respectively . the residuals of the caps are minimized for values of @xmath55 in both the [ s ii ] and [ o iii ] images . our value of @xmath56 corresponds to @xmath10 of @xmath57 mas yr@xmath6 , which agrees very well with the results from the other two methods . when all three methods are averaged together , the caps have an expansion rate of @xmath58 mas yr@xmath6 , which is more than a third smaller than the expected uniform expansion rate . they are characterized by a kinematic age of @xmath59 yr . the caps , like the major axis of e25 have significantly larger kinematic ages than the inner regions of e25 . _ condensations @xmath60 , & jets @xmath61 : _ if the condensations follow the expansion rate of the minor axis of e25 then we expect to measure a large angular expansion rate , @xmath62 mas yr@xmath6 . yet their proper motions are barely discernible ( table 2 and figure 5 ) . this is even more peculiar in light of their relatively large doppler shifts , @xmath6325 km s@xmath6 ( ms92 ) which , after ( presumably ) large correction for inclination suggest true space velocities @xmath64 km s@xmath6 relative to the nucleus . it is therefore quite curious that their proper motions are so small . unfortunately the roll angle of hst excluded the interesting polar jets j - j@xmath18 from the 1997 images . future hst observations of ngc 6543 using the 1994 roll angle should allow measurement of their proper motion . _ decreasing expansion rates . _ all of the methods seem to agree that the apparent expansion rates do not increase linearly with increasing radius , _ i.e. _ the core of ngc 6543 does not expand uniformly . either the outer nebular gas is decelerating , or it was ejected considerably earlier than was the gas within @xmath65 of the nucleus . we shall return to these points later . based primarily on e25 we have derived a distance to ngc 6543 of 1001@xmath0269 pc . this result is consistent with the previous distance determinations of 1100 , 1170 , and 980 pc by cudworth ( 1974 ) , castor _ et al . _ ( 1981 ) , and cahn _ et al . _ ( 1992 ) , respectively , but is discrepant from the 640 pc distance computed by daub ( 1982 ) by about 1.4@xmath66 . the various methods that we used to measure angular expansion rates all appear to be feasible . ( the glaring exception is @xmath67 measured for the e25 major axis in [ o iii ] line . ) indeed , based on the experience that we gained , it seems that proper motions of 0.25 pixels , or about 12 mas with the pc camera , can be measured with errors of order 20% when the signal is strong . this applies not only to images obtained under ideal conditions ( same guide stars , centering , spacecraft orientation ) but under situations in which substantial regridding of the data and corrections for geometric distortions and cosmic rays are required . according to the papers by hajian , terzian , & bignell ( 1993 , 1995 ) and hajian & terzian ( 1996 ) , the profile method might even achieve accuracies of 2 mas under better conditions ( sharper features , better control of the observations , ideal exposure durations , etc . ) than those in this study . however , the ultimate technique would exploit the two dimensions of the image data , perhaps based on the optimum reduction of the residuals along various filamentary features or the nebula as a whole . of course , at some point expansion parallax studies will be limited by systematic errors in the wfpc2 detectors , but apparently this limit has not been reached for motions as small a few mas . however , the most severe limitation of any pn expansion distance method remains a combination of assumptions about the three - dimensional geometry of the target and the measurement of expansion velocities , especially with spatial resolutions obtainable from the ground in complex targets such as ngc 6543 . ms92 built a heuristic ( empirical ) model for the evolution of ngc 6543 from the best ( ground - based ) images and their detailed long - slit spectroscopy available in 1991 . they assumed that each feature in the core of the nebula was axisymmetric and uniformly expanding . from the data and this assumption they could derive the dimensions and inclination of each feature . e25 was modeled as a closed , prolate ellipsoid , and e105 is a circular ring surrounding e25 in its equatorial plane . in their model a second , or `` outer '' thick ellipsoid surrounds the nebular core . its inner edges are approximately delineated by the caps @xmath2 and @xmath68 on its major axis and the projected tips of e105 along its minor axis . ms92 found that both e25 and the outer ellipsoid expand uniformly along nearly the same symmetry axis and inclination . the newer and more detailed hst images require an entirely new geometric interpretation . our heuristic model for the geometry of ngc 6543 is shown in figure 6 . we retain the idea of e25 as a prolate ellipsoid , albeit with numerous small bulges and irregularities near its tips . we are confident that e25 is a prolate ellipsoid for two reasons , ( 1 ) ms92 s spectroscopic data showed with high certainty that it is a closed ellipsoid , and ( 2 ) the hst images are highly suggestive of a prolate ellipsoidal structure . the outer thick ellipsoid is replaced with a figure - eight outline of two roughly - spherical bubbles with truncated outer regions where the caps are seen . e105 is situated at the locus where the bubbles are conjoined at the waist , much like a figure eight with a fat waist , or `` fat eight '' ( 8) . e25 forms one coherent structure , and e105 and the fat eight together form an ensemble which we denote the `` conjoined bubbles '' . both e25 and the conjoined bubbles have sharp edges which are nearly unresolved . if they are @xmath69 yr old or older , and if these sharp edges ( which are strong nebular pressure gradients ) were free to relax ( expand ) at their internal sound speed of 10 km s@xmath6 then they would be @xmath513 x 10@xmath70 cm , or @xmath512 , in diameter . however , almost all of the features in the core of ngc 6543 have edges which are @xmath71 thick , suggesting that they are constrained by some pressure , perhaps thermal or ram pressure . what s more , if the edges of the expanding features have been accreting material with low specific momentum then they will have been decelerated . the further they have traveled the more they will have slowed , nicely in accord with the expansion patterns discussed in 3.2 . in any event , the growth patterns of the core of ngc 6543 provide a very interesting and useful glimpse into the physics of its formation and evolution , as we discuss next . _ the standard concept of e25 as a wind heated , thermally expanding bubble moving forward supersonically ( @xmath16 mach 1.5 ) is compatible , on the whole , with its morphology , measured doppler shifts , and pattern of angular expansion . the gas which the bubble has displaced and swept up is now compressed into the bright rim which outlines it . the leading shell seems to have developed thin - shell instabilities , or lumps , of small amplitude during its evolution . this happens because the outward flow of the thermally - driven expanding bubble surface encounters dense regions upstream and bends around them , forming surface lumps . ( however , the scale of the lumps is too small to allow studies of their individual expansion patterns . ) the 1994 and 1997 observations show that although e25 is likely to be a prolate ellipsoid , it is not expanding uniformly . indeed , its expansion rate is smallest at its outermost tips . the expansion of the conjoined bubbles is more difficult to ascertain ; however , it s angular expansion rate is roughly that of e25 . therefore , the conjoined bubbles are either older or more decelerated than e25 , or both . lumpy ellipsoids like e25 are common in elliptical planetary nebulae such as ngc 6543 . close counterparts are seen in hst images of ngc 5882 , 6826 , 6884 , and especially ngc 7009 ( balick _ et al._1998 ) . in the last case the bubble has essentially the same morphology as does e25 . emanating from the tips of the bubble in ngc 7009 are highly collimated gas jets with bright ansae at their termini . the symmetry axis of ngc 7009 appears to be oriented close to the plane of the sky . however , if viewed at an inclination of @xmath72 , the bubble , jets , and ansae would resemble e25 and the jets j - j@xmath18 in ngc 6543 . the elongated shape of e25 and the bubble of ngc 7009 are generally believed to be the consequence of a high - pressure waist of confining material that forces the thermally - expanding bubble to grow most rapidly along the direction orthogonal to the disk ( _ e.g. _ , icke _ et al . _ thick disks or tori produce prolate ellipsoids because the expansion in a large equatorial region is hindered . thin disks on the other hand , produce pairs of figure - eight lobes as in the homunculus of @xmath73 carinae ( morse _ et al . _ 1998 ) and the conjoined bubbles of ngc 6543 , because the restricting material is confined to a narrow , dense region , while allowing free expansion in the rest of the nebula . the concept of e25 as a thermally expanding bubble fails two tests . firstly , the torus that putatively constrains the equatorial growth of e25 should be situated astride the minor axis of e25 and probably interior to e105 . such a torus is readily discerned in ngc 7009 . however , no such torus is apparent in ngc 6543 , perhaps owing to the confusion of brighter features nearby . secondly , the prolateness develops because the wind - driven bubbles expand fastest in the polar directions where the upstream confining pressures are smallest . therefore the angular expansion rate of e25 is predicted to be largest along its major axis . rather , just the reverse is seen . the kinematic age of the tips of e25 is roughly twice as large than the kinematic age of its waist . if the observed trends continue , the e25 ellipsoid will evolve to become increasingly less prolate in time . to rescue the concept of e25 as a thermally expanding bubble one might argue that although the tips of the bubble originally expanded much faster than its waist , the tips have now displaced and accreted considerable amounts of gas with lower specific momentum and have been decelerated . alternately the gas behind the tips may have cooled adiabatically , so the thermal pressure driving the expansion has abated . _ conjoined bubbles . _ almost certainly the conjoined bubbles represent the projected edges of a pair of bubble - like shells conjoined along their waist . along its waist a ruffled and slightly irregular series of ragged low - ionization , elongated knots ( e105 ) which seem to point away from the nucleus . these knots may be the result of an instability formed as the edges of the expanding bubbles pinch and compress the trapped gas in their waist , like the blades of a pair of scissors . alternately they are ablation flows from dense neutral knots subjected to ionizing radiation ( mellema _ et al._1998 ; redman & dyson 1999 ) . although e25 has counterparts in many other planetaries , the conjoined bubbles do not . the closest analogues may be ngc 7027 and some bipolars such as ngc 650 - 1 , 2371 - 2 , and 6309 . interestingly , the conjoined bubbles also mimics @xmath73 carinae , except that the waist of the latter object looks like a large thin disk with radial streaks , and its figure eight is more pinched . if @xmath73 carinae follows the pattern of decreasing expansion with radius that presently characterizes the conjoined bubbles , then it is not difficult to imagine that its overall morphology will evolve over time to resemble the conjoined bubbles ! dwarkadas & balick ( 1998 ) showed how this might happen if the hydrodynamic evolution of the system is dominated by thermal expansion and adiabatic cooling . _ historical evolution . _ one key fact is that the waist of the conjoined bubbles , e105 , and e25 have the same kinematic age within our error bars . hence e25 and the conjoined bubbles formed at roughly the same time . based on the increasing kinematic age with radial distance from the nucleus , it appears that e25 is overtaking some of the more distant nebular regions such as the tips of e25 and the caps d - d@xmath18 . we propose a fairly simple framework for understanding the evolution of the core of ngc 6543 . about 1000 yr ago , the star ejected a pulse of material which now forms the conjoined bubbles . however , the absence of limb brightening along the edges of the conjoined bubbles ( except at @xmath74 ) suggests that the ejection event filled the interior with low - density gas . for example , this might have happened if the gas formed a shock which left some of the outflowing gas in a hot state . starting shortly after the pulse of mass ejection the nucleus has been blowing a high - speed wind , perhaps much like the wind observed today : @xmath75 @xmath76 m@xmath77 yr@xmath6 and @xmath78 km s@xmath6 ( perinotto , cerruti - sola & lamers 1989 ) . this wind has swept into the interior of the expanding conjoined bubbles . it has displaced and plowed the upstream gas and has formed the feature e25 : therefore , e25 defines the `` realm '' of the wind - affected portion of the core of the nebula . e25 is now seen as a prolate balloon pushing into the interior of the conjoined bubbles , fastest near the equator , and decelerating towards the poles . as we noted before , the plowing of slower upstream gas and adiabatic cooling inside the tips may have slowed the expansion of e25 along its major axis . the upstream portions of the conjoined bubbles have yet to know of the existence of e25 . thus the morphology of the conjoined bubbles is a result of events that shaped it initially . this feature `` remembers '' its origins . its pinched waist suggests that a thin , dense disk shaped it into a bipolar when it was formed . e105 is the remnant of this disk . this fanciful scenario fails to account for the symmetrically placed caps , condensations , and jets of ngc 6543 . we can only conjecture that the caps are dense , neutral material left behind by an even earlier ejection of mass . their low specific momentum and neutral , dense gas form impediments to the polar growth of the conjoined bubbles and , at the same time , cause dense ionization fronts to form . the origins of the condensations and jets are even less obvious . we shall return to these issues in a later paper in which the detailed ionization structure of the nebula is discussed . the excellent agreement of each of the three methods presented confirms that we have , for the first time , measured the optical expansion parallax of a pn , and that our error estimates are adequate . we have shown that accurate measurements of the expansion of nebular features are readily obtainable with the wfpc2 camera with a reasonable time baseline . we have further shown that the e25 major axis , the polar caps , the polar condensations , and the conjoined bubbles are all expanding significantly slower than expected of a hubble - like expansion . this result is confirmed in both the [ o iii ] and [ s ii ] images , and indicates a region in which ejected gas is either decelerating , or left behind from an earlier epoch , or both . combining the proper motion of bright fiducials of the nebula with corresponding spectroscopically obtained radial expansion velocities ( ms92 ) has allowed us to determine the distance to ngc 6543 of 1001@xmath0269 pc . our deduced kinematic age of the brightest inner parts of the core of the nebula , based on three measurements along the minor axis of the tilted prolate ellipsoid e25 , is 1039@xmath0259 yr . the outer regions , which resemble a pair of conjoined bubbles with a large waist , have longer kinematic ages . a heuristic and highly conceptual explanation of the evolution of these features is as follows . a major outburst occurred a thousand years before today s nebula formed . even earlier outbursts produced the huge halo that surrounds ngc 6543 , the polar caps , and other features beyond about a radius of 20@xmath79 from the nucleus . these debris form the environment for the latest ejection . the major outburst formed a dense equatorial disk , the remnants of which are e105 . in addition , the ejected gas was shaped by this disk into a pair of expanding bubbles which join at e105 . the `` conjoined bubbles '' were created in such a way that their interiors are uniform and probably in a hot state . these bubbles are expanding and adiabatically cooling . stellar winds began to blow at high speeds shortly after the latest massive outburst . these have created an overpressured hot bubble which expends into the interior of the conjoined bubbles . the bubble attained a prolate geometry by rapid expansion along its polar direction . the subsequent evolution of the bubble has formed e25 . like many other interior bubbles in planetary nebulae , e25 is forming lumps on its surface , perhaps the result of the buckling of its thin expanding shell . the expansion of the tips of the prolate bubble has displaced and accreted material of lower specific momentum , so the tips have decelerated much more than its waist . e25 is therefore becoming less prolate as it expands . in the future e25 may overtake the conjoined bubbles . when this occurs rayleigh - taylor instabilities will set in , shredding the remnants of the nebula , and forming a clumpy halo perhaps much like the larger one that presently surrounds ngc 6543 ( _ e.g. , _ balick _ et al . _ 1993 ) . the ongoing go 7501 program has selected 30 pne for wfpc2 observations for its ongoing multi - epoch expansion parallax program . first epoch observations are nearly complete . subsequent observations will allow accurate expansion parallax distance measurements of these pne . third epoch observations of ngc 6543 are planned for late in the year 2000 during which the observing configuration of the 1994 observations by harrington & borkowski ( go 5403 ) will be replicated in the bright , complementary nebular lines of [ o iii ] and [ n ii ] as closely as possible . this observation should provide excellent data spanning six years under highly controlled conditions , allowing even more accurate measurements of the expansion patterns and the distance to ngc 6543 . we are extremely grateful to virginia player for her diligent efforts in making astrometric measurements for our analyses . we are most grateful to the space telescope science institute for providing archived data used in this study . support for this work was provided by nasa through grant number go 7501 from the space telescope science institute , which is operated by aura , inc . , under nasa contract nas 5 - 26555 . [ o iii ] ( f502n ) & 2x200s , 2x600s & 17@xmath80 58@xmath81 28@xmath82 & 66@xmath8337@xmath84 & 274.105@xmath83 & 9 - 18 - 94 & 5403 & harrington ( f502n ) & 200s & @xmath85 58@xmath81 28@xmath86 & 66@xmath833814@xmath87 & 305.434@xmath83 & 8 - 17 - 97 & 6943 & casertano e25 minor & 72 ( in [ o iii ] ) & profile & 3.63@xmath00.82 & & & rad . fit & 3.36@xmath00.86 & & & mag . & 3.4@xmath00.9 & & & & & e25 major & 161 ( in [ o iii ] ) & profile & 4.51@xmath00.85 & & & rad . fit & 1.95@xmath01.01 & & & mag . & 4.8@xmath01.2 & & & linear ( hubble ) expectation & 7.7 & & & & & e105 minor & 90 ( in [ s ii ] ) & profile & 5.8@xmath01.5 & & & rad . fit & 3.98@xmath01.17 & 3.28@xmath00.7 & & mag . & 4.6@xmath01.2 & 4.6@xmath01.2 & & linear ( hubble ) expectation & 4.3 & & & & & e105 major & 146 ( in [ s ii ] ) & profile & 7.2@xmath01.5 & & & rad . fit & 4.68@xmath02.34 & 8.59@xmath01.7 & & mag . & 7.5@xmath01.9 & 7.5@xmath01.9 & & linear ( hubble ) expectation & 7.0 & & & & & polar caps ( dd@xmath89 ) & 212 ( in [ s ii ] ) & profile & 6.70@xmath01.37 & 6.94@xmath01.42 & & rad . fit & 6.56@xmath01.3 & 5.86 @xmath01.2 & & mag . & 6.4@xmath01.6 & 6.4@xmath01.6 & & linear ( hubble ) expectation & 10.2 & & & & & condensations ( ff@xmath89 ) & 247 ( in [ s ii ] ) & rad . fit & & 2.73@xmath01.64 & & mag . & & 4.2@xmath01.1 & & linear ( hubble ) expectation & 11.9 & * figure 1 : * images of ngc 6543 . left : a linear representation of the [ o iii ] surface brightness . various features are identified ( ms92 terminology shown in italics ) . white dots show the locations of the radial fits of angular expansion rates of table 2 . center : a negative linear representation of the [ o iii ] surface brightness superimposed on a logarithmic display of the faint nebular background in the same hst image . right : like the center panel , but for the [ n ii ] line . the [ s ii ] image is noisier but otherwise similar in appearance to this panel . * figure 2 : * a linear representation of the [ o iii ] surface brightness of ngc 6543 showing the location of the profiles analyzed in 3.2 . values of the angular expansion rate in mas yr@xmath6 are shown corresponding to the locations ( black dots ) where the profile cuts intersect bright regions of the nebula . * figure 3 : * same as for figure 2 , except for the [ s ii ] surface brightness of ngc 6543 . the values listed correspond to the measured angular expansion rates in mas yr@xmath6 where the profiles intersect the bright polar caps . * figure 4 : * a montage of f502n ( [ o iii ] ) residual images after the 1997 images were subtracted from magnified and aligned 1994 images . the magnification factors that were applied to the 1994 images are shown in the upper right corners of each panel . * figure 6 : * schematic of a conceptual geometric model of ngc 6543 . a plan view of the model is shown on the right side . how it might appear if tilted at about 30@xmath17 is shown in projection on a combined [ n ii ] and [ o iii ] image .

chromospheric evaporation was first suggested by @xcite to explain the origin of the hot , dense , soft x - ray ( sxr ) emitting plasma confined in the coronal loops during solar flares . the basic scenario is as follows . magnetic reconnection , believed to be the primary energy release mechanism , heats the plasma and accelerates particles high in the corona . the released energy is transported downward along the newly reconnected closed flaring loop by non - thermal particles and/or thermal conduction , heating the chromospheric material rapidly ( at a rate faster than the radiative and conductive cooling rates ) up to a temperature of @xmath0@xmath1 k. the resulting overpressure drives a mass flow upward along the loop at a speed of a few hundreds of km s@xmath2 , which fills the flaring loop with a hot plasma giving rise to the gradual evolution of sxr emission . this process should also result in a derivative of the sxr light curve in its rising portion that closely matches the hard x - ray ( hxr ) light curve , which is called neupert effect and is observed in some ( but not all ) flares @xcite . hydrodynamic ( hd ) simulations of chromospheric evaporation have been carried out with an assumed energy transport mechanism ( e.g. , electron `` beam '' or conductive heating ) @xcite leading to various predictions on the uv - sxr spectral lines produced by the evaporated plasma , as well as the density and temperature profiles along the flaring loop . most of the observational tests of these predictions rely on the blue - shifted components of sxr emission lines produced by the up - flowing plasma , first reported by @xcite and @xcite using spectra obtained from the _ p78 - 1 _ spacecraft . similar observations were subsequently obtained from x - ray spectrometers on the solar maximum mission ( _ smm _ ; antonucci et al . 1982 , 1984 ) , the _ hinotori _ spacecraft @xcite , the _ yohkoh _ spacecraft @xcite , and the solar and heliospheric observatory ( _ soho _ ; brosius 2003 ; brosius & philips 2004 ) . @xcite , on the other hand , observed co - spatial sxr blue - shifts ( up - flows ) and h@xmath3 red - shifts ( down - flows ) as expected from hd simulations @xcite . a summary of relevant observations from _ smm _ can be found in @xcite . all the aforementioned observations , however , were indirect evidence in the sense that the evaporation process was not imaged directly . based on hd simulations , @xcite derived the expected x - ray brightness profile across the evaporation front and suggested that _ yohkoh_/sxt or x - ray imagers with equivalent or better spatial and temporal resolution should be able to detect the front . indeed , @xcite found that the hxr and sxr sources of the 1994 june 30 flare moved toward the loop top ( lt ) during the impulsive phase . since the flare was located near the center of the solar disk , they identified such motions as the horizontal counterpart of the line - of - sight motion revealed by the blue - shifted emission lines observed simultaneously by _ yohkoh_/bcs . the reuven ramaty high energy solar spectroscopic imager ( _ rhessi _ ) with its superior spatial , temporal , and spectral resolution @xcite provides us with opportunities to study the chromospheric evaporation process in unprecedented detail . we report in this paper our analyses of the spatial and spectral evolution of a simple flare on 2003 november 13 with excellent _ rhessi _ coverage . because the flare occurred near the solar limb it presented minimum projection effects and a well - defined loop geometry that allows direct imaging of the hxr brightness profile along the loop . the observations and data analyses are presented in [ obs ] , followed by a derivation of the evolution of the density profile along the flaring loop in [ dens ] . we summarize the major findings of this paper and draw conclusions in [ conclusion ] . the flare under study is a _ goes _ m1.7-class flare that occurred on 2003 november 13 in the active region 0501 after it appeared on the east limb . this event followed a period of extremely high solar activities in late october and early november when a series of x - class flares including the record setting x28 flare of 2003 november 4 took place @xcite . _ rhessi _ had an excellent coverage of this flare . figure [ lightcv ] shows the _ rhessi _ and _ goes_-10 light curves . the _ goes _ 8 - 1 ( 1.6 - 12.4 kev ) and 4.0 - 0.5 ( 3.1 - 24.8 kev ) fluxes rise gradually and peak at 05:00:51 and 05:00:15 ut , respectively . the _ rhessi _ high energy ( @xmath4 kev ) count rates , on the other hand , exhibit two pulses peaking at 04:58:46 and 05:00:34 ut with the first one stronger . the steps in the _ rhessi _ light curves are due to the attenuator ( shutter ) movements @xcite . before 04:57:57 ut and after 05:08:59 ut , there were no attenuators in and between the two times the thin attenuator was in except for a short period near 05:05 ut when the attenuator briefly moved out . figure [ mosaic ] shows the evolution of the flare at different energies , which may be divided into three phases : ( 1 ) before 04:57:57 ut , corresponding to the rising phase when the emission mainly comes from a flaring loop to the south . ( 2 ) between 04:57:57 and 05:08:59 ut , the impulsive phase , during which another loop to the north dominates the emission . this loop appears to share its southern footpoint ( fp ) with the loop to the south , which is barely visible because of its faintness as compared with the northern loop and _ rhessi _ s limited dynamic range of @xmath010 . ( 3 ) after 05:08:59 ut , the decay phase , when the shutters are out and two off - limb sources ( identified as the lts of the two loops ) dominate . the relatively higher altitudes compared with earlier lt positions are consequences of the preceding magnetic reconnection , as seen in several other _ rhessi _ flares @xcite . clearly the southern loop , which extends to a relatively higher altitude , evolves slower and is less energetic than the northern one . we shall focus on the evolution of the northern loop during the first hxr pulse ( 04:58 - 05:00 ut ) in this paper . it is necessary to check if pulse pileup is important in this flare , before we can make more quantitative interpretation of the data . the reason is that although we have applied the first order pileup correction @xcite in our spectral analysis , such a correction is challenging for images and not available at present . there are several ways to do the check , the detector livetime being the first and simplest indicator . we first accumulated spatially integrated spectra for every 1 s time bins during the interval of 04:58:01 - 04:59:49 ut , ( see text about figure [ prof_3e ] ) , which covers the bulk of the first hxr pulse . ] using the front segments of all the nine detectors except detectors 2 and 7 which have degraded energy resolution @xcite . we then obtained the livetime ( between data gaps ) from the spectrum object data and averaged it over the seven detectors being used . the resulting livetime generally decreases with time , ranging from 96% to 89% , with a small modulation produced by the spacecraft spin . in this m1.7 flare , such a livetime is comparably high ( cf . livetime of @xmath5 during the 2002 july 23 x4.8 flare and @xmath6 during the 2002 february 20 c7.5 flare ) and indicates minor pileup severity . another approach involves inspecting the change of the spectrum due to pileup . we accumulated spectra over every 1 spacecraft spin period ( @xmath7 s , with the same set of detectors mentioned above ) and used the pileup correction to obtain the relative fraction of the pileup counts among the total counts as a function of energy @xcite . we find that the pileup counts amount to less than @xmath8 of the total counts at all the energies until 04:59:01 ut when the livetime drops to 91% . after that , the relative importance of the pileup counts continues increasing , but remains below @xmath9 of the total counts before 04:59:17 ut . toward the end of the first hxr pulse ( 04:59:45 - 04:59:49 ut , livetime @xmath10 ) , the pileup counts to total counts ratio exceeds @xmath11 in the entire 20 - 40 kev range and humps up to @xmath12 near 28 kev . we respectively integrate the pileup counts and total counts over the 20 - 40 kev band , and plot their ratio versus time , as a general indicator of pileup severity ( see figure [ pileup ] ) . clearly this ratio is @xmath13 during the first 2/3 of the interval shown and does not reach the moderate @xmath14 level until the very end . we therefore conclude that pileup effects are generally not very significant for this flare , especially during the first minute of the impulsive phase , because the count rate is not too high and the thin shutter is in at times of interest which further attenuates the count rate . it should be noted that the two piled up photons ( that result in a single photon seen in the image ) most probably originate from the same location on the sun and pileup of photons across difference sources is relatively unimportant ( g. hurford , private communication ) . therefore the source geometry would not be significantly affected by pileup , except that there could be a ghost " of a low - energy source appearing in a high - energy image for very large ( e.g. x - class ) flares . however , the spectra of individual sources derived from images are distorted , which is relatively more significant at the lt than at the fps . this is because ample low - energy photons dominate in population over high - energy photons , having the highest probability to produce pileup , and generally most of the low - energy photons are emitted by the lt source . we now examine the images in greater detail . top left panel _ of figure [ mdi ] shows _ rhessi _ clean @xcite images of the northern loop at 9 - 12 , 12 - 18 , and 28 - 43 kev for 04:58:22 - 04:58:26 ut . ( although the 4-second integration time is rather short , the image quality is reliable with a well - defined source structure . ) at 9 - 12 kev the lt dominates and the emission extends towards the two fps , which dominate the emission at 28 - 43 kev and above with the northern fp ( n - fp ) much brighter than the southern one ( s - fp ) . one of the most interesting features of the source structure is that emission from the legs of the loop dominates at the intermediate energy ( 12 - 18 kev ) . similar structures are also observed for several other time intervals during the first hxr pulse ( see discussions below ) . we find emission from the legs is a transient phenomenon at intermediate energies , because when integrating over a long period and/or a broad energy band , the lt and/or fp sources become dominant . to our knowledge no images like this have been reported before . we attribute this in part to the relatively short integration time and _ rhessi _ s high energy resolution . to be compared with observations at other wavelengths , the same images at 9 - 12 and 28 - 43 kev ( solid contours ) are shown with the _ soho_/eit , mdi magnetogram , and mdi white light maps in the other three panels of figure [ mdi ] , where the dashed contours depict the southern loop at 6 - 9 kev for 04:57:40 - 04:57:52 ut . the eit image at 04:59:01 ut ( _ upper right panel _ ) shows emission at 195 co - spatial with the sxr emission from the northern loop . the brightest 195 emission , an indicator of the highest differential emission measure ( thus the highest density ) at @xmath15 k , appears to be close to the n - fp which is also the strongest fp in hxrs . k ( see * ? ? ? 12 ) , and emission intensity would be lower for both higher and lower temperatures . ] the _ lower left panel _ displays the x - ray emission along with the post - flare ( 05:57 ut ) mdi magnetogram . this clearly shows that the northern loop straddles across a polarity reversal with the brighter n - fp associated with a stronger magnetic field . the southern loop ( dashed contours ) is associated with a even weaker magnetic field . here we show the mdi magnetogram recorded one hour after the flare s impulsive phase because during a flare there are many uncertainties in the magnetic field measurement . the _ lower right panel _ shows the mdi continuum map at 12:47 ut ( about 8 hours after the flare ) , suggesting that the flare occurred above the lower sunspot region ( dark area ) . note that during this interval the sunspot has moved westward for about @xmath16 in heliographic longitude . we do not plot the mdi white light map at the time of the flare because then the sunspot was nearly on the limb and barely visible . next we consider the evolution of the northern loop . we notice that , as shown in the four columns for 04:58:00 - 04:59:20 ut ( boxed by the dotted line ) in figure [ mosaic ] , the fps initially appear at all energies but later on dominate only in the high energy bands , while the lt is first evident at low energies and becomes more and more prominent at relatively higher energies as indicated by the dashed diagonal line . the emission from the lt also extends towards the legs at intermediate energies and in a given energy band the emission concentrates more and more at the lt with time . these are expected to be common features of flares with a single loop because of chromospheric evaporation that can increase the plasma density in the loop making the lt dominate at progressively higher energies . however , because the 20 second integration time is relatively long , these images do not uncover the details of the evaporation process . to remedy this we have carried out three different but complementary analyses of the images with higher temporal or energy resolution . \1 . to study the source morphology change over short time intervals , we model the loop geometry and study the evolution of the hxr brightness profile along the loop . we first made clean images in two energy bands of 6 - 9 kev ) photons by the attenuator and escape of the germanium k - shell fluorescence photons that are produced by photoelectrical absorption of higher energy ( 10 - 20 kev ) photons in the germanium detector ( see * ? ? ? . however , for the flare under study , the 6 - 9 kev image most likely reveals the real lt morphology , because there are ample thermal photons at lower energies originating from the lt source and photons at slightly higher energies seem to come from the same location . ] and 50 - 100 kev over the time interval of 04:58:12 - 04:58:53 ut that covers the plateau portion of the first hxr pulse . from these two images we obtained the centroids ( indicated by the white crosses in figure [ loop - model]_a _ ) of the sources identified as the lt ( 6 - 9 kev ) and two fps ( 50 - 100 kev ) , respectively . assuming a semi - circular loop that connects the three centroids , we located the center of the circle marked by the plus sign in figure [ loop - model]_a_. the grey scale in figure [ loop - model]_a _ was obtained by superposition of 30 images ( six 8-second intervals from 04:58:08 to 04:58:56 ut in five energy bands : 9 - 12 , 12 - 15 , 15 - 20 , 20 - 30 , and 30 - 50 kev ) reconstructed with the pixon algorithm @xcite . figures [ loop - model]_b _ and [ loop - model]_c _ respectively show the intensity profiles perpendicular to and along the loop ( averaged over the respective orthogonal directions ) . the inner and outer circles ( at @xmath17 and @xmath18 ) in figure [ loop - model]_a _ show the positions of the 50% of the maximum intensity in figure [ loop - model]_b_. however , for inferring the intensity profile along the loop we use radially integrated flux down to the 5% level . this enables us to include as much source flux as possible ( with little contamination from the southern loop ) . we define the mean of the radii at the 5% level as the radius of the central arc of the loop ( the white dot - dashed line in figure [ loop - model]_a _ ) . with the above procedure , one can study the evolution of the brightness profile along the loop at different energies . figure [ prof_3e ] shows the results obtained from pixon images with an integration time of 1 spacecraft spin period ( @xmath04 seconds ) from 04:58:01 to 04:59:49 ut for three energy bands ( 20 - 30 , 15 - 20 , and 12 - 15 kev ) . using a simple algorithm we determine the local maxima whose slopes on both sides exceed some threshold value and mark them by filled circles . we compare each profile with its counterpart obtained from the clean image ( with the same imaging parameters ) , and use the rms of their difference to estimate the uncertainty as indicated by the error bar near the right - hand end of the corresponding profile . for each panel , the rms difference of all the profiles , as a measure of the overall uncertainty , is shown as the error bar in the upper - right corner . this uncertainty is about 10% for the three energy bands ; as expected , it increases slightly at higher energies which have lower counts . figure [ prof_3e]_a _ displays the profile at 20 - 30 kev , which as expected ( see figure [ mosaic ] ) shows emission from the two fps with fairly constant positions until the very last stage when the lt emission becomes dominant . at this stage , the s - fp becomes undetectable and the n - fp has moved very close to the lt . at the lower energy ( 15 - 20 kev , figure [ prof_3e]_b _ ) the maxima tend to drift toward the lt gradually and eventually merge into a single lt source . at the even lower energy ( figure [ prof_3e]_c _ ) this trend becomes even more pronounced and the drift starts earlier except here the shift is not monotonic and there seems to be a lot of fluctuations . we also repeated the same analysis at a higher cadence ( every 1 second , @xmath7 second integration interval ) with both pixon and clean algorithms . the evolution of the resulting profiles ( although oversampled and thus not independent for neighboring profiles ) appears to be in line with that shown here at a 4-second cadence obtained with pixon . the general trends of these results indicate that high energy hxr producing electrons lose their energy and emit bremsstrahlung photons higher and higher up in the loop as the flare progresses . this can come about simply by a gradual increase of the density in the loop , presumably due to evaporation of chromospheric plasma . from the general drift of the maxima we obtain a time scale ( @xmath010 s of seconds ) and a velocity of a few hundred km s@xmath2 consistent with the sound speed or speed of slow magnetosonic waves . as stated above at low energies we see some deviations from the general trend , some of which do not appear to be random fluctuations . if so and if we take one of the evident shorter time scale trends shown by the dashed line in figure [ prof_3e]_c _ we obtain a large velocity km s@xmath2 @xcite and 800 km s@xmath2 @xcite , obtained from blue - shifted fe xxv spectra . ] ( @xmath0@xmath19 km s@xmath2 ) which is comparable to the alfvn or fast magnetosonic wave speed . this may indicate that another outcome of energy deposition by non - thermal particles is the excitation of such modes which then propagate from the fps to the lt and might be responsible for the circularly polarized zebra pattern observed in the radio band @xcite . this , however , is highly speculative because the spatial resolution ( @xmath0@xmath20 ) is not sufficiently high for us to trust the shorter time scale variation . the longer time scale general trend , however , is a fairly robust result . \2 . instead of examining the source structure with high time resolution , we can investigate it with higher energy resolution at longer integration intervals as tradeoff for good count statistics and image quality . to this end , we have made pixon images during three consecutive 24 s intervals starting from 04:58:00 in 20 energy bins within the 6 - 100 kev range . figure [ box ] shows a sample of these images at 04:58:24 - 04:58:48 ut . figures [ prof_24s]_a_-[prof_24s]_c _ show the x - ray emission profile along the loop at different energies for the three intervals . ) . ] as in figures [ prof_3e]_a_-[prof_3e]_c _ the high energy emission is dominated by the fps but there is a decrease of the separation of the fps with decreasing energies and with time . again at later stages the lt dominates and the profile becomes a single hump . the general trend again suggests an increase of the gas density in the loop . at lower energies ( @xmath21 kev ) , the profile is more complicated presumably due to many physical processes ( in addition to chromospheric evaporation ) , such as thermal conduction and transport of high energy particles , thermal and non - thermal bremsstrahlung , wave excitation and propagation , wave - particle coupling , and even particle acceleration , which may be involved . we believe that a unified treatment of acceleration and hd processes with physical conditions close to the flare is required for interpreting these results to uncover the details . to quantify this aspect of the source structure evolution , we divided the loop into two halves as shown by the boxes in figure [ box ] and calculated their emission centroids . the resulting centroids at the three times together with the central arc of the model loop are plotted in figure [ centroids ] . as can be seen , for each time interval the centroids are distributed along the loop with those at higher energies being further away from the lt , and the entire pattern shifts toward the lt with time . figure [ centr - dist ] shows the centroid positions of the northern half of the loop ( where the source motions are more evident ) along and perpendicular to the loop during the three intervals . this again shows that higher energy emission is farther away from the lt and the centroids shifted towards the lt with time , but similarly there are some complicated patterns at low and intermediated energies . all these are consistent with the general picture proposed above for the chromospheric evaporation process . \3 . to further quantify the source motions , we obtained the brightness - weighted standard deviation or the 2nd moment of the profiles . in general the moment measures the compactness of the overall emission , but does not yield the sizes of individual sources whose measurement is still challenging for _ rhessi _ @xcite . hence our attention should be paid to the general trend of the moment rather than its absolute values , which may be subject to large uncertainties and thus less meaningful . the moments of the profiles resulting from clean images ( in three energy bands over 8 s intervals ) are plotted in figure [ time]_b_. there is a general decrease of the moment with the decline starting earlier at lower energies . such a decrease is expected if the two fps move closer to each other . however , caution is required here because a decrease of this quantity could also come about by other causes , say by an increasing dominance of the brightest source . we therefore checked the original images and the corresponding profiles when interpreting our results . to estimate the uncertainty of the moment , for each energy band we repeated the calculation with different integration time ( _ panel c _ ) . the resulting moments remain essentially unchanged and as excepted the fluctuations of the moment decrease with increasing integration time . we also plot in _ panel c _ the moment ( solid line ) obtained from pixon images with an integration time interval of two spin - period ( @xmath22 s ) , which basically agrees with its clean counterpart in the general trend . the gradual decrease of the moment is consistent with the motion of the centroids of sources up along the legs of the loop , which can take place by a continuous increase of the gas density in the loop due to evaporation . spectral analysis can be used to study the evaporation process as well . with an isothermal plus a power law model , we fitted the spatially integrated _ rhessi _ spectra down to 6 kev @xcite for every 8 second interval during the impulsive phase . the emission measure ( @xmath23 ) and temperature of the isothermal component ( asterisk symbols ) are plotted in _ panels d _ and _ e _ of figure [ time ] , respectively . the @xmath23 rises almost monotonically with time from 0.6 to 14.2 @xmath24 @xmath25 . this translates into an increase of the plasma density ( @xmath26 ) by a factor of @xmath05 assuming a constant volume @xmath27 . the temperature remains almost constant with a trend of slight decrease with time . the @xmath23 and temperature derived from the _ goes _ data ( plus symbols ) are also shown for comparison . in general the _ goes _ results are smoother and the temperature increases monotonically but remains below that of the _ rhessi _ , consistent with previous results @xcite . this is expected because _ rhessi _ is more sensitive to higher temperatures than _ goes_. however , surprisingly , the _ goes _ emission measure is also lower than that of _ rhessi _ as opposed to what is the case more generally ( see holman et al . it is not clear whether or not this is due to a problem related to the _ rhessi _ calibration at low energies . nevertheless , the continuous increase of the @xmath23 s at comparable rates does suggest a gradual increase of the plasma density . the best fit parameters of the power law component with a low energy cutoff are plotted in figure [ time]_f_. the power - law index @xmath28 ( plus symbols ) is anti - correlated with the high energy light curves ( see figure [ time]_a _ ) and shows a soft - hard - soft behavior . it starts with 4.43 at 04:58:02 ut , drops to 3.82 at the impulsive peak ( 04:58:26 ut ) , and rises up to 7.12 at 04:59:46 ut . the high indexes ( @xmath29 ) may be an indicator of high - temperature thermal rather than non - thermal emission . thus in what follows we limit our analysis to time up to 04:59:20 ut . the low - energy cutoff ( asterisks ) of the power law is about @xmath30 kev and is near the intersection of the isothermal ( exponential ) and power - law components . the neupert effect is commonly quoted as a manifestation of chromospheric evaporation @xcite and a simple energy argument ( e.g. li et al . 1993 ) is often used to account for the relationship between sxr and hxr fluxes ( @xmath31 and @xmath32 ) . in the thick - target flare model , the non - thermal @xmath32 represents the _ instantaneous _ energy deposition rate ( @xmath33 ) by the electron beam precipitating to the chromosphere , but the thermal @xmath31 is proportional to the _ cumulative _ energy deposited , i.e. , the time - integral of @xmath33 . it naturally follows that the temporal derivative of the sxr flux , @xmath34 , should be related to @xmath32 . the simplest test of the neupert effect is usually carried out by plotting @xmath34 and @xmath32 in some energy band . there are many reasons why a simple linear relationship would not be the case here . the first and most important is that @xmath33 is related to @xmath32 through the bremsstrahlung yield function @xmath35 ( @xmath36 ) which is not a constant and depends on the spectrum of the electrons or hxrs ( see e.g. petrosian 1973 ) . here the most crucial factor is the low energy cutoff ( @xmath37 ) of the non - thermal electrons , but the spectral index also plays some role . the total yield of all the bremsstrahlung photons produced by a power law spectrum of electrons with energies above @xmath37 ( in units of 511 kev ) is @xmath38 and the yield of the photons whose energies are greater than @xmath37 is @xmath39 where @xmath40 , @xmath41 is the coulomb logarithm , and @xmath42 is the spectral index of the power - law electron flux . as shown in figure [ time]_f _ both the low energy cutoff and spectral index of the non - thermal emission vary during the pulse , indicating variations in the electron spectrum and thus breaking the linearity of the sxr - hxr relationship . other factors which can also produce further deviations are energy deposition by protons ( and other ions ) , by conduction and other possible ways of dissipation of energy than simply heating and evaporating the chromospheric plasma by non - thermal electrons . a detailed treatment of the problem requires solutions of the combined transport and hd equations , which is beyond the scope of this paper . @xcite by inclusion of some of these effects in an approximate way found that the expected relationship was mostly not present in several _ rhessi _ flares . finally one must include the fact that the chromospheric response of sxr emission will be delayed by tens of seconds depending on the sound travel time ( and its variation ) and other factors . the flare under study has shown no indication of gamma - ray line emission which means that the contribution of protons most probably is small . in the currently most favorable model where the electrons are accelerated stochastically by turbulence ( see e.g. petrosian & liu 2004 ) the turbulence can suppress heat conduction during the impulsive phase and possibly also during the decay phase @xcite . because there does not appear to be large changes in the shape of the loop during the impulsive phase other energy dissipation processes such as cooling by expansion may also be negligible . assuming these to be the case we have performed the neupert effect test in two ways , the first being the common practice of examining the relation between @xmath34 and @xmath32 . we then examine the relation between @xmath33 and @xmath34 by taking into account the variation of the bremsstrahlung yield . the temporal derivatives of the fluxes of the two _ goes _ channels are shown in the _ bottom panel _ of figure [ lightcv ] . as evident , during the rising portion of the _ goes _ fluxes both channels derivatives indeed match the first pulse of the _ rhessi _ hxr light curves ( @xmath43 kev ) , but not during the second weaker pulse ( where the 1 - 8 derivative shows some instrumental artifacts ) . this may be due to the fact that the neupert effect of the second pulse is overwhelmed by the cooling of the hot plasma produced during the first stronger pulse . nevertheless , the sxr light curves ( of both _ goes _ and _ rhessi _ ) exhibit slightly slower decay rate than expected from the first pulse alone . this most likely is the signature of the energy input by the second pulse , which slows down the decay of the first pulse alone . we note in passing that the sxr light curves start rising several minutes prior to the onset of the hxr impulsive phase . this is an indication of preheating of the plasma before production of a significant number of suprathermal electrons . the 6 - 12 kev curve rises faster than the _ goes _ curves at lower photon energies , which is consistent with the picture that the primary energy release by reconnection occurs high in the corona where the relatively hotter plasma is heated before significant acceleration of electrons ( as suggested in petrosian & liu 2004 ) , and before transport of energy ( by accelerated electrons or conduction ) down the flare loop to lower atmosphere where cooler plasmas are heated subsequently and produce the _ goes _ flux . on the other hand , the increase of the sxr flux at the beginning is dominated by the southern loop , which shows little evidence of chromospheric evaporation . the phenomenon therefore may be a unique feature of this flare . to quantify the sxr - hxr relationship , we cross - correlated the _ rhessi _ 30 - 50 kev photon energy flux ( @xmath44 , figure [ neupert]_a _ ) and the derivative of the _ goes _ low energy channel flux ( @xmath34 , figure [ neupert]_c _ ) in the sxr rising phase ( 04:58:00 - 04:59:51 ut ) . the resulting spearman rank - order correlation coefficient ( see figure [ neupert]_f _ ) , an indicator of an either linear or nonlinear correlation , shows a single hump with a maximum value of 0.91 ( corresponding to a significance of @xmath0@xmath45 ) at a time lag of 12 s. this suggests a delay of @xmath34 relative to @xmath44 , which is expected given the finite hydrodynamic response time ( on the order of the sound travel time of @xmath020 s for a loop size of @xmath0@xmath46 cm and @xmath47 k ) required for redistribution of the deposited energy . such a delay is evident in the numerical simulations of @xcite who in addition found that the density enhancement contributes more to the total sxr emissivity than the temperature increase for longer duration ( @xmath48 s ) hxr bursts during the decay phase . in figure [ neupert]_d _ , we plot the two quantities with the _ goes _ derivative shifted backward by 12 s to compensate the lag of their correlation . a linear regression ( dotted line ) gives @xmath49 with an adjusted coefficient of determination ( or so - called r squared ) @xmath50 close to 1 suggesting a good linear correlation . we also carried out the same analysis for the electron energy power @xmath33 , assuming a thick - target model of power - law electrons with a low - energy cutoff of @xmath51 kev . we first obtained the energy flux of all the photons with energies greater than @xmath37 , @xmath52 , from the 30 - 50 kev photon energy flux @xmath44 : @xmath53 where @xmath54 is the photon flux distribution at the sun ( photons kev@xmath2 s@xmath2 ) which is obtained from spectrum fitting ( see [ spec ] ) and assumed to extend to infinity in energy space . we then calculated the power of the electrons by @xmath55 where the bremsstrahlung yield @xmath56 is given by equation ( [ y ] ) . ) here . however , one can still use equation ( [ y ] ) with a simple correction factor of @xmath57 in the range @xmath58 to achieve an accuracy of @xmath59 . ] the resulting @xmath33 is plotted versus time and _ goes _ derivative in _ panels b _ and _ e _ of figure [ neupert ] respectively . the dotted line in _ panel e _ shows a linear fit ( @xmath60 ) to the data : @xmath61 the corresponding spearman correlation coefficient has a peak value of 0.78 ( significance @xmath0@xmath62 ) at a time lag of 3 s ( figure [ neupert]_f _ ) . as evident , @xmath33 yields no better correlation with @xmath34 than @xmath44 , which is similar to the conclusion reached by @xcite . during the hxr decay phase ( after 04:59:20 ut ) the spectrum becomes softer ( @xmath63 ) and @xmath64 decreases much slower than @xmath44 since the bremsstrahlung yield ( equation [ [ y ] ] ) decreases with the spectral index . as noted above for these high spectral indexes , the emission might be thermal rather than non - thermal . the inferred electron power is thus highly uncertain for these times . as stated earlier , the total energy of the non - thermal electrons is very sensitive to the low - energy cutoff @xmath37 which is generally not well determined ( cf . sui et al . we thus set @xmath37 as a free parameter and repeated the above calculation for different values of @xmath37 ( ranging from 15 to 28 kev ) . we find that , as expected , the temporal @xmath33-@xmath34 relationship highly depends on @xmath37 . for a small @xmath37 ( @xmath65 kev ) , @xmath33 keeps rising till @xmath004:59:50 ut ( near the bottom of the @xmath44 light curve ) , which makes the @xmath33-@xmath34 correlation completely disappear . on the other hand , for a large @xmath37 ( @xmath66 kev ) the correlation is generally good during the impulsive pulse ( through 04:59:10 ut ) and the larger @xmath37 the better the correlation . this is because the conversion factor @xmath67 in equation ( [ factor ] ) is an increasing ( decreasing ) function of the photon spectral index @xmath28 if @xmath37 is sufficiently small ( large ) . for a small @xmath37 , for example , the photon energy flux @xmath52 may have a somewhat large value in the valley of the @xmath44 light curve when @xmath28 is high . in addition , during this time interval the bremsstrahlung yield @xmath56 becomes small since @xmath42 is large ( see equation [ [ y ] ] ) and consequently this may result in a very large @xmath33 by equation ( [ edot ] ) . as to the magnitude of the energy flux of non - thermal electrons , @xcite in their hd simulations found that the dynamics of the flare loop plasma is very sensitive to its value . for a low energyflux ( @xmath68 ergs @xmath69 s@xmath2 ) , the up - flow velocity of the evaporating plasma is @xmath010 s of km s@xmath2 ; for a high energy flux ( @xmath70 ergs @xmath69 s@xmath2 ) , a maximum up - flow velocity of @xmath0100 s of km s@xmath2 can be produced . for the flare under study , we estimate the area of the cross - section of the loop to be @xmath71 @xmath72 , where the upper limit corresponds to the loop width determined by the 5% level in figure [ loop - model]_b_. we read the maximum electron power of @xmath73 ergs s@xmath2 from figure [ neupert]_b _ , which is then divided by @xmath74 ( assuming a filling factor of unity ) to yield the corresponding electron energy flux : @xmath75 ergs @xmath69 s@xmath2 . the source velocity on the order of a few hundreds of km s@xmath2 estimated in [ morp ] is consistent with that predicted by @xcite . for comparison , we note that @xcite from _ rhessi _ data also obtained an energy flux of @xmath76 ergs @xmath69 s@xmath2 for an m2.2 flare during which an up - flow velocity of @xmath0230 km s@xmath2 was inferred from simultaneous co - spatial _ soho_/cds doppler observations . in summary , the _ goes _ sxr flux derivative @xmath34 exhibits a neupert - type linear correlation with the _ rhessi _ hxr flux @xmath44 during the first hxr pulse . however , unexpectedly , the correlation between the electron power @xmath33 and @xmath34 is not well established based on the simple analysis presented here , which suggests that a full hd treatment is needed to investigate the chromospheric evaporation phenomenon ( see discussions in [ conclusion ] ) . for the 1994 june 20 disk flare , @xcite interpreted the moving sxr sources as thermal emission from the hot ( @xmath77 mk ) plasma evaporated from the chromosphere based on the good agreement of the emission measure of the blue - shifted component and that of the sxr from the fps . for the limb flare under study here , doppler shift measurements are not available . meanwhile , a purely thermal scenario would have difficulties in explaining the systematic shift of the centroids towards the fps with increasing energies up to @xmath070 kev as shown in figure [ centr - dist ] . a non - thermal scenario appears more appropriate . that is , the apparent hxr fp structure and motions can result from a decrease in the stopping distance of the non - thermal electrons with decreasing energy and/or increasing ambient plasma density caused by the chromospheric evaporation ( as noted earlier in [ morp ] ) . one can therefore derive the density distribution along the loop from the corresponding x - ray emission distributions ( e.g. , figure [ prof_24s ] ) , without any pre - assumed density model ( c.f . * ? ? ? this approach is described as follows . for a power - law x - ray spectrum produced by an injected power - law electron spectrum , @xcite obtained a simple empirical relation ( also see * ? ? ? * ) for the x - ray intensity @xmath78 per unit photon energy @xmath79 ( in units of 511 kev ) and column depth @xmath80 ( in units of @xmath81 = 5 \times 10^{22}$ ] @xmath69 for @xmath82 cm and @xmath83 ) : @xmath84 where @xmath28 and @xmath42 @xmath85 are the photon and electron spectral indexes , respectively , @xmath86 is a constant normalization factor , and @xmath87 , where @xmath88 is the distance measured from the injection site . this equation quantifies the dependence of the emission profile ( or source morphology ) on the electron spectral index and column depth . in general , when @xmath42 decreases ( spectrum hardening ) , the intensity at a given photon energy rises ( drops ) at large ( small ) @xmath80 s and thus the emission centroid shifts to larger @xmath80 s . this is expected because for a harder spectrum , there are relatively more high energy electrons that can penetrate to larger column depths and produce relatively more bremsstrahlung photons there . the opposite will happen when the spectrum becomes softer . during the impulsive peak showing a soft - hard - soft behavior ( see [ spec ] ) , one would expect that the emission centroids shift first away from and then back toward the lt ( if the density in the loop stays constant ) . knowing the spectral index , the emission profile can therefore yield critical information about the density variation in both space and time . to compare the above empirical relation with observations , we first integrate @xmath78 over an energy range @xmath89 $ ] , @xmath90 and then integrate @xmath91 over @xmath80 to obtain the cumulative emission , @xmath92 k^{-\gamma } dk , \label{ftau}\ ] ] where we have chosen @xmath93 so that @xmath94 . comparison of @xmath95 with the observed emission profiles gives the column depth @xmath96 whose derivative with respect to @xmath88 then gives the density profile along the loop . specifically for this flare , we assume that the non - thermal electrons are injected at the lt indicated by the middle vertical dotted line in figure [ prof_24s ] and denote the profile to the right - hand side of this line ( i.e. , along the northern half of the loop ) as @xmath97 , where @xmath89 $ ] is the energy band of the profile . the observed cumulative emission is then given by , @xmath98 where @xmath99 ( corresponding to @xmath100 ) is the maximum distance considered and @xmath101 has been properly normalized . then @xmath102 can be obtained by inverting @xmath103 where the integration over @xmath79 in equation ( [ ftau ] ) can be calculated numerically . it should noted that , however , not all the profiles in figure [ prof_24s ] are suitable for this calculation , because low energy emission is dominated by a thermal component especially in the lt region and at later times . we thus restrict ourselves to the energy ranges of 12 - 72 , 13 - 72 , and 17 - 72 kev , respectively for the three 24 s intervals . the lower bound is the energy above which the power - law component dominates over the thermal component , determined from fits to the spatially integrated spectrum for each interval as shown in figure [ spec_24s ] . within these energy ranges , separate leg or fp sources rather than a single lt source can be identified in the corresponding image , which is morphologically consistent with the non - thermal nature of emission assumed here . to further minimize the contamination of the thermal emission in our analysis , we have excluded the lt portion of the emission profile in excess of the lowest local minimum ( if it exists ) between the lt and leg ( or fp ) sources . an example of this exclusion is illustrated with the hatched region in figure [ prof_24s]_c _ for the 19 - 21 kev profile . this was done by simply replacing the profile values between the lt and the local minimum positions with the value at the minimum . we calculated @xmath104 for every emission profile within the energy ranges mentioned above for the three intervals in figure [ prof_24s ] , with the photon indexes , @xmath105 4.46 , 3.97 , and 4.23 respectively . from the geometric mean of the column depths obtained at different energies , @xmath106 , we derived the density profile @xmath107 for each time interval . the results are shown in figure [ dens - fig ] , where we shall bear in mind that attention should be paid to the overall trend rather than the details of the density profile and its variation , because the profile here only spans about 3 times the resolution ( @xmath108 ) and thus is smoothed , making neighboring points not independent . as can be seen , between the 1st and 2nd intervals , the density increases dramatically in the lower part of the loop , while the density near the lt remains essentially unchanged . the density enhancement then shifts to the lt from the 2nd to the 3rd interval . this indicates a mass flow from the chromosphere to the lt . the density in the whole loop is about doubled over the three intervals , which is roughly consistent with the density change inferred from the emission measure ( see figure [ time]_d _ ) . these results are again compatible with the chromospheric evaporation picture discussed in [ morp ] . we have presented in this paper a study of _ rhessi _ images and spectra of the 2003 november 13 m1.7 flare . _ rhessi _ s superior capabilities reveal great details of the hxr source morphology at different energies and its evolution during the impulsive phase . the main findings of this paper are as follows . ( 1 ) the energy dependent source morphology in general shows a gradual shift of emission from the lt to the fps with increasing energies . over some short integration intervals emission from the loop legs may dominate at intermediate energies . ( 2 ) the emission centroids move toward the lt along the loop during the rising and plateau portions of the impulsive phase . this motion starts at low energies and proceeds to high energies . we estimate the mean velocity of the motion to be hundreds of km s@xmath2 , which agrees with the prediction of the hydrodynamic simulations by @xcite . there are also shorter time scale variations that imply much higher velocities ( @xmath0@xmath19 km s@xmath2 ) but we are not certain if they are real because of instrumental limitations . ( 3 ) fits to the spatially integrated _ rhessi _ spectra with a thermal plus a power law model reveal a continuous increase of the emission measure while the temperature does not change significantly . the _ goes _ data show a similar trend of the @xmath23 but a gradual increase of the temperature . ( 4 ) the time derivative of the _ goes _ sxr flux is correlated with the _ rhessi _ hxr flux with a peak correlation coefficient of 0.91 at a delay of 12 s in agreement with the general trend expected from the neupert effect . however , the correlation between the electron power and the _ goes _ derivative is no better than the sxr - hxr correlation . ( 5 ) from the observed brightness profiles we derive the spatial and temporal variation of the plasma density in the loop , assuming a non - thermal thick - target bremsstrahlung model . we find a continuous increase of the density , starting first at the fps and legs and then reaching to the lt . all these results fit into a picture of continuous chromospheric evaporation caused by the deposition of energy of electrons accelerated during the impulsive phase . several of the new features of this event ( such as the leg emission at intermediate energies ) may be common to many solar flares . expanding the sample of flares of this kind will be very helpful to understand the underlying physical processes . the new findings reside near the limit of _ rhessi _ s current temporal , spatial , and spectral resolution . as advanced imaging spectroscopy capabilities are being developed and spatial resolution is being improved in the _ rhessi _ software @xcite , it will be critical to obtain the spatially resolved photon spectrum along the loop . this will yield incisive clues to the nature of the moving x - ray sources and relevant energy transport mechanisms and will be useful to check the reality of the short time scale variations . there are several important questions that need to be further addressed in future observational and theoretical investigations : ( 1 ) what is the nature of the moving x - ray sources ? could they be characterized as thermal emission from the evaporated hot plasma or non - thermal emission from the precipitating electrons , or a mixture of both ? could they be related to mhd waves or evaporation fronts ? ( 2 ) what are the roles of different heating agents of the chromosphere , i.e. , electron beams , thermal conduction , and/or direct heating by turbulence or plasma waves during the impulsive phase ? we have pointed out to some of the many physical processes that come into play in answering such questions . here we describe possible directions for future theoretical studies . we have shown that a more physical based test of the neupert effect between the electron power and sxr flux derivative does not reveal a better correlation than the usual hxr vs sxr derivative correlation . although the observed source velocity agrees with those of hd simulations , there are some features that current simulations have not addressed . to answer these questions requires an updated numerical calculation where one combines the model of particle acceleration and transport with the hd simulation of the atmospheric response to energy deposition to form a unified picture of solar flares . for example , one can use the output electron spectrum from the stochastic particle acceleration model @xcite as the input to the transport and hd codes , rather than simply assume a power - law electron spectrum as previous hd simulations . such a study can shed light on the relative importance of particle beams and thermal conduction in evaporating chromospheric plasma and the roles that mhd waves may play in heating the flaring plasma , in particular , addressing our tentative observation of the fast source motion which suggests possible presence of mhd waves in the flare loop . a better understanding of their propagation , damping , and excitation mechanisms is necessary for uncovering the energy release process during flares . this is particularly true in the context of the stochastic particle acceleration model .

a long gamma - ray burst ( grb ) is one of the most energetic explosions in the universe , which is observed via soft gamma - ray emission followed by afterglow in lower energy range . it is now broadly agreed that at least some of long grbs originate in core - collapse of massive stars together with core - collapse supernovae ( cc sne ) . however , cc sne do not always accompany long grbs , and the criteria discriminating between long grbs and general cc sne is one of the most important questions regarding long grbs . some theoretical studies on the origin of long grbs using stellar evolution models suggest that low - metallicity may be a necessary condition for a long grb to occur ( @xmath3 , e.g. @xcite ) . metallicity distribution of long grb host galaxies provides us with an important clue to study the relation between metallicity and a long grb occurrence . although metallicity of a long grb host galaxy is not necessarily identical to that of the progenitor ( e.g. , @xcite ) , galaxies with low - metallicity would have higher grb production efficiency [ i.e. , long grb rate to star formation rate ( sfr ) ratio ] than high - metal galaxies if low - metallicity is a necessary condition for a long grb to occur . with a sample of 5 host galaxies ( 3 long grbs and 2 x - ray flashes , xrfs ) , ( , hereafter s06 ) showed that the metallicity distribution of the host galaxies at redshifts @xmath4 is significantly biased towards low - metallicities compared with star forming galaxies at similar redshifts , suggesting long grbs really occur in low - metallicity environment . however their sample is too small to determine the relation between metallicity and long grb occurrence rate . the sample number of spectroscopically studied long grb host galaxies has been dramatically increased during the 10 years after s06 ( e.g. , @xcite ) . however , the relation between metallicity and long grb occurrence rate is not understood quantitatively . one difficulty is that we can determine redshifts and/or identify host galaxies of only a small fraction of long grbs ( @xmath1 30% in recent observations ) , which is likely biased with respect to the distributions of redshifts and host galaxy properties ( the @xmath5-determination / host - identification bias ) . furthermore , even when the redshift and the host galaxy of a long grb are known , the host galaxy is not always studied in detail . the targets of the spectroscopic studies have been selected non - uniformly by numerous independent researchers making the effect of the sampling bias unevaluable ( the reporting bias , e.g. , @xcite ) . there have been some previous long grb host observation campaigns invented to overcome the sampling biases ( so called unbiased surveys , @xcite ) . in the unbiased surveys , the samples are selected without using informations about redshifts and host galaxies to avoid the @xmath5-determination / host - identification bias , and hence the samples span a very wide range of redshifts ( @xmath6 0-6 ) . in such a wide range of redshift , properties of general galaxies largely evolve making it difficult to reach enough statistics to study the environmental requirements for a long grb occurrence by comparing the properties of grb host galaxies to those of general galaxies . some recent studies pick out grb host galaxies in specific ranges of redshifts from the unbiased surveys and compared their properties to those of general galaxies at similar redshifts @xcite . however , the properties of general galaxies , to which we compare long grb host galaxies to investigate the environmental requirements for a long grb occurrence , are still not well understood at high - redshifts . metallicity measurement of high - redshift galaxies is an especially challenging issue ( at @xmath7 , e.g. , @xcite ) . on the other hand , the sample number becomes very small at low - redshifts in the unbiased surveys ( @xmath8 grbs at @xmath9 ) because the cosmic long grb rate density is low and the comoving volume element is small , although long grb host galaxies at low - redshifts are of special importance to understand the relation between metallicity and long grb occurrence rate as we describe below . hence , although the unbiased surveys are important to understand what bias we would suffer by selecting long grbs whose redshifts and host galaxies are known , it is difficult to unveil the environmental requirements for a long grb occurrence only with the unbiased surveys . metallicity measurements of the host galaxies of all long grbs known at low - redshifts ( @xmath10 ) can dramatically improve our understanding of the relation between metallicity and long grb occurrence rate . there are advantages of studying grb host galaxies at low - redshift in some aspects . 1 . the success rate of grb redshift - determinations / host - identifications would be higher at lower redshifts ( i.e. , the @xmath5-determination / host - identification bias is weaker ) , because faint galaxies can be detected without very deep observations and spectroscopies of host galaxies can provide redshift informations even when grb afterglows are not bright enough for spectroscopies . the reporting bias can be eliminated by finding host galaxies of all grbs known in the redshift range and measuring metallicities of all host galaxies , which is difficult to do at higher redshifts ( @xmath11 ) . 3 . a wealth of spectroscopic studies of general galaxies at low - redshifts , such as the sloan digital sky survey ( sdss ) , provides us with control sample of galaxy properties , to which we can compare long grb host properties to investigate the relation between long grb rate and metallicity . the advantages and the disadvantages of a redshift selected study of grb host galaxies at low - redshifts can be summarized as in table [ tab : procon ] , in comparison with the other strategies of grb host studies . in this study , we present the overall metallicity distribution of the host galaxies of grbs known at @xmath12 that occurred before the end of march 2014 . for a few of this low - redshift host galaxies , sufficient spectroscopic data to constrain their metallicity were not previously available in the literature , or signal - to - noise ratio ( s / n ) of important metallicity indicating emission lines was low . we perform spectroscopic metallicity measurements of the host galaxies of three low - redshift grbs , 060614 , 090417b , and 130427a , to obtain better constraints on the metallicity distribution . we compare the metallicity distribution of this low - redshift sample to the predictions of empirical models of galaxies , and constrain the relation between metallicity and grb occurrence rate . this paper is organized as follows . we describe our sample selection , observations , and analysis of the obtained spectra in section [ sec : sample ] , [ sec : obs ] , and [ sec : linefit ] , respectively . in section [ sec : prop ] , we discuss the properties of the low - redshift long grb host galaxies including metallicities . in section [ sec : moedel ] , we describe models of galaxy metallicity distributions , to which we compare the low - redshift sample of host galaxies to constrain the relation between metallicity and grb occurrence rate . the results of the comparison are presented in section [ sec : result ] . in section [ sec : discussion ] , we discuss uncertainties of our galaxy model , possible subpopulations in the low - redshift grbs , and sampling biases we may suffer by collecting long grbs with known redshifts . we summarize our conclusions in section [ sec : conclusion ] . throughout this paper , we assume the fiducial cosmology with @xmath13 , @xmath14 , and @xmath15 70 km s@xmath16 mpc@xmath16 . the magnitudes are given in the ab system . [ tab : procon ] @xmath17 this work . we collect low - redshift grbs from online databases : swift gamma - ray burst table , gamma - ray burst online index ( grbox ) , j. greiner s grb table , and grb host studies ( ghosts ) . although there are various classes of gamma - ray burst like events : e.g. , long grbs , short grbs , ultra - long grbs , and xrfs , we focus on host galaxies of long grbs in this study . the classification of burst events are still controversial ( e.g. , @xcite ) , and important burst properties for the classification such as burst duration @xmath18 , and spectral peak energy @xmath19 are uncertain in some cases . however , for simplicity , we select long grbs simply by reported @xmath18 and @xmath19 as follows . the prompt emission properties are taken from @xcite , @xcite , @xcite , and @xcite , in addition to the online databases mentioned above . first , we collect bursts with 2 @xmath20 sec as long grbs , considering those with @xmath21 sec as short grbs and @xmath22 sec as ultra - long grbs . we also exclude short grbs with extended emissions , grb 050709 , 050724 , 061210 , and 071227 from our sample . from the bursts which meet the duration criteria , we remove bursts with very soft spectra ( xrfs ) which may be a different population from long grbs . xrfs are often defined according to their hardness ratio which depends on energy bands of x - ray detectors . to take into account burst events discovered by various instruments , we define bursts with @xmath23 kev as xrfs in this study . for some _ swift _ detected bursts , @xmath19 estimation is not available . in that case , we distinguish grbs and xrfs using photon spectral index in the energy range of _ swift_-bat ( @xmath24 ) . we consider bursts with @xmath25 which is typical of long grbs with @xmath26 kev @xcite as long grbs , and those with @xmath27 as xrfs . within the criteria described above , 15 long grbs occurred before the end of march 2014 at spectroscopically measured redshifts @xmath12 ( table [ tab : grbs ] ) however , we exclude grb 050219 from further discussion in this study , leaving 14 low - redshift long grbs in our sample , because it is unclear whether this burst really has occurred at a low - redshift . grb 050219 occurred close to ( but @xmath1 offset from ) an early - galaxy for which the redshift @xmath28 is measured @xcite . given that long grbs result from death of massive stars , the association of grb 050219 with an early - type galaxy is surprising , although it is not strictly prohibited because early - type galaxies can harbour small amount of star formation ( e.g. , @xcite ) . the significant offset of grb 050219 from its claimed host galaxy also suggests that the galaxy might be aligned with the grb by chance without any physical association with the burst . it is possible that the remaining sample is also contaminated by galaxies that coincide with grbs by chance . we discuss this issue in section [ sec : contamination ] . among the 14 grbs , grb 060614 , 090417b , 111225a , and 130427a were without sufficient spectroscopic observations to significantly constrain their host metallicity . we perform spectroscopic observations of the host galaxies of grb 060614 , 090417b , and 130427a in this study . there are also some recently published spectroscopic observations of the host galaxies of grb 060614 and 130427a . we also discuss those observations in section [ sec : obs ] . the host galaxy of grb 111225a missed our spectroscopic observation because its redshift had not been measured as of the end of march 2014 , and no metallicity information is available for this host galaxy . the redshift of this burst was reported @xmath1 2.3 years after the burst @xcite . we take into account this host galaxy as a part of the uncertainty in the metallicity distribution of the grb host galaxies . we discuss the implications for the sampling bias in the redshift selected sample from the late redshift report of this burst in section [ sec : fainthost ] . hereafter , grb means long grb unless otherwise stated . [ tab : grbs ] long grbs whose redshifts are spectroscopically determined to be @xmath12 . short grbs , xrfs , and ultra - long grbs are not included . + @xmath17 if ` host galaxy ' , the redshift is determined by emission lines of the host galaxy . if ` afterglow ' , the redshift is determined by absorption lines in the afterglow . the numbers in the square brackets are the references as listed below . + references : 1 . @xcite , 2 . @xcite , 3 . @xcite , 4 . @xcite , 5 . @xcite , 6 . @xcite , 7 . @xcite , 8 . @xcite , 9 . @xcite , 10 . @xcite , 11 . @xcite , 12 . @xcite , 13 . @xcite , 14 . @xcite , 15 . @xcite , 16 . @xcite , 17 . @xcite , 18 . @xcite , 19 . @xcite , 20 . @xcite + @xmath29 bursts with confirmed sn associations @xcite + @xmath30 we do not include grb 050219 in the analyses in this paper , because the association of this bust with its host galaxy for which the redshift is measured is highly uncertain . we perform optical spectroscopy of the host galaxies of grb 060614 , 090417b , and 130427a with the gemini multi - object spectrographs ( gmos , @xcite ) , and use flux ratios of the emission lines : [ o]@xmath313727 , h@xmath32 , [ o]@xmath314959 , [ o]@xmath315007 , h@xmath33 , and [ n]@xmath316584 , as metallicity indicators ( hereafter , [ o ] , [ o ] , and [ n ] mean [ o]@xmath313727 , [ o]@xmath315007 , and [ n]@xmath316584 , respectively , unless otherwise stated ) . the apparent spatial extent of the three host galaxies are @xmath1 , and major part of their flux is collected in the slit . the spectra obtained with gmos were calibrated using spectroscopic standard stars observed on different nights from the observations of the host galaxies , and hence the absolute flux scale is not accurate . emission line fluxes of the grb 130427a host galaxy obtained during a target - of - opportunity program ( too , pi : n. kawai ) using subaru / focas @xcite are presented together . we also performed imaging observations of the host galaxy of grb 111225a . we describe the observations of each target in the following subsections . see table [ tab : obs ] for the summary of our observations . the data were reduced in a standard manner using pyraf/ iraf , together with the gemini iraf package and the focasred package . grb 060614 is known as a grb unassociated with a bright sn at @xmath34 @xcite . although some spectroscopic observations of the grb 060614 host galaxy were performed soon after the burst @xcite , the metallicity measurements were uncertain because the fluxes of [ n ] and [ o ] lines were not well constrained . recently , @xcite also reported metallicity measurement of the host galaxy of grb 060614 . the metallicity they derived is broadly consistent with ours which we discuss in section [ sec : metal ] . we performed spectroscopy of the grb 060614 host galaxy with gmos - south using the slit and two different settings of grisms and order - cut filters . one is the b600 grating ( no - filter ) which covers 35006000 , and the other is the r400 grating + the og515 order - cut filter covering 53009000 . the signals were @xmath35 pixels binned . the b600 spectroscopy was performed on 2015 april 25 ( ut ) . the integration time was 1000 sec @xmath36 , and the slit position angle was set to the parallactic angle to minimize the effect of differential atmospheric refraction . the r400 spectroscopy was performed on 2015 april 28 , and may 14 with the integration time of 1200 sec @xmath37 and the position angle set to the parallactic angle . grb 090417b is an optically dark grb at @xmath38 with a very long duration of @xmath39 sec @xcite . the spectroscopy was performed with gmos - north on 2015 march 25 . we use the slit , the r831 grating , and the rg610 order - cut filter covering 73009200 . at this redshift , the h@xmath33 line coincides with a strong night - sky emission line at 8827 . to accurately subtract sky emission lines , we utilize the nod and shuffle sky subtraction . the signals were 2 pixels binned along the spectral direction . the total integration time was 2640 sec ( 60 sec @xmath40 2 positions @xmath40 11 cycles @xmath40 2 sequences ) . the slit position angle was set to the parallactic angle . grb 130427a is a bright burst at @xmath41 associated with a broad - lined type ic sn 2013cq @xcite . although detection of the host galaxy emission lines over the afterglow and the sn are reported by @xcite and ( , hereafter k15 ) , the detection of the [ n ] line is marginal in either case . the spectroscopic follow up observation of grb 130427a with subaru / focas was performed on 2013 may 17 ( i.e. , @xmath1 20 days after the burst ) , under a weather condition with cirrus clouds ( pi : n. kawai ) . the slit , the 300b grating , and the y47 filter covering 47009000 were used . the slit position angle was set to to put both of the center of the host galaxy and the grb position into the slit . the signals were 2 pixels binned along the spatial direction . the integration time was 9600 sec . unfortunately the atmospheric diffraction corrector ( adc ) could not be used at the time of the observation , however the spectroscopy was performed with elevation angles @xmath42 , and atmospheric refraction would be @xmath43 throughout the wavelength range . although the s / n of the focas spectrum is higher than those previously reported , the significance of the detection of the [ n ] line was still low @xmath44 ( @xmath45 in k15 ) . furthermore , the strong night - sky emission line at 8827 overlaps with the [ n ] line of the host galaxy at this redshift . hence accurate subtraction of the sky emission is essential for the spectroscopy of this host galaxy . we performed further spectroscopy of the grb 130427a host galaxy with gmos - north on 2015 march 27 to measure the [ n ] line flux more securely with higher significance of the line detection . the total integration time was 2640 sec . the slit position angle was set to the parallactic angle . we used the nod and shuffle sky subtraction to subtract the sky emission line accurately . the signals were 2 pixels binned along the spectral direction . @xcite reported the redshift of grb 111225a , @xmath46 , in april 2014 ( @xmath1 2.3 years after the burst ) by analyzing the archival data with an updated software . thus grb 111225a was not included in the target list of our spectroscopic observations which was compiled in march 2014 . we perform imaging observations of the host galaxy of grb 111225a in @xmath47 , and @xmath48band filters with subaru / focas on 2015 september 22 , under a photometric condition . the integration time is 1800 sec for @xmath49band , and 600 sec for each of @xmath50 , and @xmath48band . signals were @xmath35 pixels binned . photometric calibration is performed with respect to @xcite standard stars in the pg0231 + 051 field observed in the same night . [ tab : obs ] the observations of the grb host galaxies we present in this study . + @xmath17 use of nod & shuffle sky subtraction . + @xmath29 parallactic angle + @xmath30 the integration time of 1 nod & shuffle sequence consists of 60 sec @xmath40 2 positions @xmath40 11 cycles . + @xmath51 too , poor weather condition to minimize the effect of stellar absorption features on the line flux measurements , we first subtract stellar spectral energy distribution ( sed ) models of the host galaxies from the observed spectra . we perform sed fittings to optical / near - infrared ( nir ) broad - band photometries in the literature ( figure [ fig : sedfit ] ) using the _ sedfit _ software package @xcite which utilizes population synthesis models by @xcite . the extinction law by @xcite and the initial mass function ( imf ) by @xcite are assumed . we examine six cases of stellar metallicity : @xmath52 , and five cases of star formation history : simple stellar population ( ssp ) , constant star formation , and exponentially decaying star formation with @xmath53 and 5 gyr . as mentioned in section [ sec : obs ] , absolute scale of our spectra is not accurate . furthermore , it is possible that a fraction light is lost at the slit . in the following discussion , we rescale the observed spectra so that the continuum flux agrees with the best fit sed models . the broad - band photometries , the best fit sed models , and the rescaled spectra are shown in figure [ fig : sedfit ] . the parameters of the best fit stellar sed models are listed in table [ tab : pop ] . we derive metallicities of the grb host galaxies using relative flux ratios between different emission lines as indicators ( section [ sec : metal ] ) , and hence the derived metallicities are not significantly affected by the error of absolute flux scale . [ tab : pop ] the parameters of the best fitting stellar sed models @xcite that reproduce the broad - band photometries in the literature ( the upper left panel of figure [ fig : sedfit ] ) . the extinction law by @xcite and the imf by @xcite are assumed . there are some archival spectra of the grb 060614 host galaxy as mentioned in section [ sec : obs:060614 ] , and those by @xcite and @xcite are obtained with gmos - south using the slit and the r400 grating as well as our r400 spectrum . although the gmos archival data alone are not deep enough to constrain the metallicity of the host galaxy , we stack them with our r400 spectra to maximize the s / n . the gmos - south detectors were replaced between the archival observations and ours , the pixel scale remained @xmath54 / pix with the r400 grating . the integration time of the archival observations is 1800 sec @xmath40 4 and 1200 sec @xmath40 4 , respectively . we note that the detector replacement improved the efficiency at the wavelength of h@xmath33 and [ n ] at @xmath34 by @xmath1 35% . we subtract the best fit sed models from the rescaled spectra , and measure emission line fluxes by fitting the lines with gaussian plus flat continuum models . the residual spectra after the model subtraction and the gaussian fits are shown in figure [ fig : fitha ] and figure [ fig : fitoxhb ] . although [ o ] line is a doublet , we fit it with a single gaussian as well as the other lines , because the doublet is not resolved in our spectrum . the obtained emission line fluxes are corrected for the foreground extinction in the milky way ( mw ) using the extinction map by @xcite and the extinction law by @xcite , and shown in table [ tab : lines ] . the continuum component in the spectrum obtained by the focas too observation of grb 130427a is contaminated with the afterglow and the sn ( figure [ fig : too130427a ] ) . hence we can not calibrate the flux scale comparing the continuum spectrum with the best fit stellar sed model . instead , we firstly measure h@xmath33 flux without the absorption correction , and rescale the focas spectrum so that the absorption uncorrected h@xmath33 flux agrees with that obtained from the gmos spectrum . the emission line fluxes of the other low - redshift grb host galaxies in the literature are also shown in table [ tab : lines ] . the host galaxies of grb 980425 , 060505 , and 120422a , are spatially resolved in the spectroscopic observations @xcite , and emission line fluxes are measured at multiple positions within the galaxies . in that case , we use the integrated flux over the whole galaxy or the flux measured at the peak of the light profile of the host galaxy , for consistency with the spectroscopic data of the other grb host galaxies which are not spatially resolved . , but for the [ o ] , h@xmath32 , [ o ] lines of the grb 060614 host galaxy . the spectrum is stacked with the the archival data @xcite in the wavelength ranges of h@xmath32 and [ o ] . , width=302 ] the emission line fluxes of the host galaxies in the low - redshift sample ( see section [ sec : sample ] ) are listed in table [ tab : lines ] . with these emission line measurements , we derive metallicity of ionized gas , 12+log@xmath55(o / h ) , which would be close to the metallicity of young stars in the galaxies . we utilize both n2 index (= @xmath56 [ n]@xmath316584/h@xmath33 ) and @xmath57 index [ = ( [ o]@xmath313727+[o]@xmath314959+[o]@xmath315007)/h@xmath32 ] as indicators of metallicity . various calibrations are proposed for the relation between the indiators and metallicity , and different calibrations are often inconsistent with each other @xcite . in this study , we utilize the method of ( , kk04 ) , in which both n2 and @xmath57 indices are calibrated using the same photoionization model , to obtain consistent results from the two indicators . the solar - metallicity @xmath59 corresponds to 12+log@xmath55(o / h ) = 8.69 @xcite . in the kk04 calibration , the n2 and @xmath57 indices are related to metallicity taking the effect of ionization parameter ( @xmath60 ) into account . the @xmath60 parameter can be estimated from @xmath61 index [ = ( [ o]@xmath314959+[o]@xmath315007)/[o]@xmath313727 ] . the metallicity calibration by kk04 suffers from the model uncertainties of @xmath1 0.1 dex , and hence we consider error of a metallicity measurement to be 0.1 dex when the error propagation from the emission line fluxes is smaller than this value or flux error informations are not available . the fluxes in table [ tab : lines ] are corrected for the foreground extinction in the mw , but not for the extinction in the host galaxies themselves . to derive metallicity , we correct the line fluxes for the host extinctions , assuming the intrinsic value of the balmer decrement h@xmath33/h@xmath32 = 2.86 and the extinction law by @xcite . the positions of the host galaxies on the baldwin - phillips - terlevich ( bpt ) diagram @xcite with the extinction correction are shown in figure [ fig : bpt ] . in the bpt diagram , the grb host galaxies show emission line ratios that are consistent to result purely from star forming activities . however , @xcite showed that the spectrum of the host galaxy of grb 031203 is possibly affected by an active galactic nucleus ( agn ) investigating more numerous emission line fluxes than examined in the bpt diagram . hence the derived properties of the grb 031203 host galaxy might be systematically affected by an agn . we describe the metallicity determination of each host galaxy in the appendix . the derived metallicities and the extinctions are listed in table [ tab : metal ] & [ tab : msfr ] . when both of the n2 and @xmath57 indices are valid for a galaxy , we use the @xmath57 index , although the two indices agree with each other except for the host galaxies of grb 031203 and 011121 . the emission line ratio of the grb 031203 host galaxy might be affected by an agn as mentioned above , and the grb 011121 host galaxy is a remarkable outlier from the mass - metallicity ( @xmath62-@xmath63 ) relation of galaxies ( see section [ sec : result ] ) . the metallicity measurement of these objects should be considered with care . [ tab : lines ] fluxes are in units of @xmath64 erg s@xmath16@xmath65 , and corrected for the foreground extinction in the mw . when the estimated flux is consistent with zero within the 2@xmath66 error , we show the 95% upper - limits instead . + @xmath17 the flux errors are @xmath1 10% . + @xmath29 flux error informations are not available . + @xmath30 the h@xmath33 and [ n ] fluxes are measured from the gmos spectrum , and the other line fluxes are measured from the focas spectrum . indices do not agree well with each other for these host galaxies . the vertical dotted line represents the [ n]/h@xmath33 ratio of the host galaxy of grb 090417b for which the [ o]/h@xmath32 ratio is not known . , width=302 ] [ tab : metal ] @xmath67 kk04 calibration method is utilized . + @xmath68 the estimation errors are @xmath69 0.1 unless otherwise stated . + @xmath70 the best estimates as described in the appendix . + @xmath71 possibly contaminated by an agn . + @xmath72 derived from the emission line fluxes by k15 . + @xmath73 assumed value ( see the appendix ) + stellar mass ( @xmath62 ) and sfr of the low - redshift grb host galaxies are listed in table [ tab : msfr ] . the stellar masses of the host galaxies of grb 060614 , 090417b , and 130427a are derived by the sed fitting assuming the chabrier imf as described in section [ sec : linefit ] , and are broadly consistent with previously reported results @xcite . the grb 111225a host galaxy is detected only in 4 bands ( @xmath74 , and @xmath75 , see table [ tab:111225a ] ) , and it is difficult to derive the properties of this host galaxy from its sed . however , we can naively estimate its stellar mass as @xmath76 assuming typical mass - to - luminosity ratio of galaxies at similar redshifts @xcite . we collect @xmath62 measurements of the other host galaxies from the literature . when @xmath62 in the literature is derived assuming an imf other than the chabrier imf , we rescale the mass assuming : @xmath77 @xcite , for consistency . we derive sfr of the host galaxies from their extinction corrected h@xmath33 fluxes assuming the global kennicutt - schmidt law @xcite and the same scaling factor between the different imfs as mentioned above . we note that major part of the errors of sfr is propageted from the error of @xmath78 in most of the cases . the host galaxy of grb 060505 is spatially extended over a few @xmath40 10 arcsec , and only a part of the emission line fluxes is collected in the slit , and hence we adopt a sfr measured from ultraviolet photometries without dust correction instead @xcite . although the dust - uncorrected sfr is much larger than that suggested from the h@xmath33 flux with the large slit - loss , the actual sfr with dust correction might be further high . the emission lines of the grb 120714b host galaxy were detected over the optical afterglow ( @xcite ; k15 ) , and the fraction of the host galaxy light in the slit is not known . hence we consider the obtained sfr for the host galaxies of grb 060505 and 120714b as lower - limits . [ tab:111225a ] magnitudes are in the ab system , and not corrected for the foreground extinction in the mw [ @xmath79 , @xcite ] . + @xmath17 2@xmath66-limit . [ tab : msfr ] @xmath67 estimated from the balmer decrement + @xmath68 the chabrier imf is assumed . the numbers in the square brackets are references as listed below . + @xmath70 sfr is derived from the h@xmath33 flux in table [ tab : lines ] and @xmath80 in the second column unless otherwise stated . + @xmath71 possibly contaminated by an agn . + @xmath72 the extinction correction is based on the result of the sed fitting : @xmath81 . + references : 1 . @xcite , 2 . @xcite , 3 . @xcite , 4 . @xcite , 5 . this work , 6 . @xcite , 7 . @xcite , 8 . to investigate the relation between metallicity and grb occurrence , we compare the metallicity distribution of the low - redshift sample presented in section [ sec : prop ] with the metallicity distribution of star forming galaxies at similar redshifts . following @xcite and @xcite , we compute the metallicity distribution of star forming galaxies using the empirical formulation of stellar mass function [ @xmath82 , @xmath62-sfr relation ( so called `` galaxy main - sequence '' ) , and @xmath62-@xmath63 relation of galaxies . in this study , we assume @xmath83 by ( , blue population ) , the @xmath62-sfr relation by ( , see their equation 11 , scatter @xmath84 dex ) , and the @xmath62-@xmath63 relation computed with the kk04 calibration by @xcite . we only consider the mass range log@xmath85 8.0 , below which the properties of galaxies are not well constrained . this mass limit is also comparable to the lowest - mass grb host galaxies in the low - redshift sample . @xmath62 and sfr in the formulations are rescaled to be consistent with the chabrier imf . the scatter of the @xmath62-@xmath63 relation is largely dependent on @xmath62 in the sense that the scatter is larger at smaller @xmath62 . here we assume a simple formulation : @xmath86 which is broadly consistent with the scatter of the @xmath62-@xmath63 relation discussed in @xcite . it should be noted that @xcite derived the @xmath62-@xmath63 relation using galaxy spectra obtained by the sdss whose fiber spectrograph covers only the central of each galaxy . hence it is possible that the systematic loss of light from the outskirts of galaxies affects the derived metallicities , because galaxies often have lower metallicity in their outskirts than at their centers ( so called metallicity gradient , e.g. , @xcite ) . some of the low - redshift grb host galaxies might also suffer from similar problems . however , @xcite examined correlation between measured metallicities and fraction of light covered by the spectrograph using a sample of galaxies with similar @xmath62 and sfr in the sdss , and found that the dependence of derived metallicities on the fiber coviring fraction is small ( @xmath87 dex , see figure 5 of @xcite ) . thus we consider that the systematic loss of light from the outskirts of galaxies does not significantly affect our metallicity measurements although it is suggested that there is a correlation between sfr and metallicity of galaxies with similar stellar masses @xcite , the observed correlation between sfr and @xmath63 correlation is not quantitatively understood and possibly affected by sample selections and metallicity calibration methods ( e.g. @xcite ) . hence we do not consider the correlation in our baseline model , and discuss the effects of the sfr-@xmath63 correlation on our results in section [ sec : msfrz ] . in this study , we consider normalized probability distribution function ( pdf ) of grb host galaxy metallicities when we compare the model predictions to the observations . we note that the absolute scale of the stellar mass function ( @xmath88 ) , and also the absolute scale and the scatter of the @xmath62-sfr relation do not affect the predicted metallicity pdf . @xcite have shown that the choice of the formulations of the stellar mass function and the @xmath62-sfr relation does not significantly affect the predicted metallicity pdf , as far as the stellar mass function and the @xmath62-sfr relation that represent field star forming galaxies are selected . observations of some nearby galaxies ( mw , the magellanic clouds , and m31 ) show that inter - stellar medium ( ism ) in a galaxy is not chemically homogeneous . hence the metallicity of a grb progenitor star might be different from the metallicity of the host galaxy which we can measure by follow up spectroscopies . to examine metallicity variation of star forming regions within a galaxy , we consider observed metallicity distributions of h regions in nearby galaxies . although recent integral field unit ( ifu ) spectroscopies provides us with spatially resolved map of metallicity in some star forming galaxies , the spatial resolution of such observations are still limited to @xmath89 100 pc in most of the cases ( e.g. , @xcite ) . hence the spectroscopy of individual h regions ( except for giant ones ) can be performed only in a few nearby galaxies . in the upper panel of figure [ fig : zint ] , we show the metallicity distributions of h regions in the mw ( solar - neighborhood , @xcite ) , the large / small magellanic clouds ( lmc and smc , @xcite ) , and the andromeda galaxy ( m31 , @xcite ) . it is possible that the metallicity estimates of the faint h regions in the m31 sample by @xcite have systematic errors @xcite , and hence we consider only h regions with h@xmath33 luminosity @xmath90 [ erg s@xmath16 ] in the m31 sample . the h@xmath33 luminosities of the h regions are taken from @xcite . see @xcite for the detail of the catalog matching between the samples of @xcite and @xcite . the metallicities of the h regions can be represented by log - normal distributions but with different median value and dispersion in different galaxies . the median and the scatter of the h region metallicity distributions in the nearby galaxies [ 12+log@xmath55(o / h)@xmath92 and @xmath93 are shown in the lower panel of figure [ fig : zint ] . although the sample number is small , it is seen that the galaxies with the higher median metallicity have larger scatter of the distribution . as a baseline model , we assume a broken linear relation between 12+log@xmath55(o / h)@xmath92 and @xmath94 as shown in the lower panel of figure [ fig : zint ] : @xmath95 & ( -3.7 \leq { \rm log_{10}(\frac{o}{h})_{gal } } < -3.4 ) \\ 0.3 & -3.4 ( \leq { \rm log_{10}(\frac{o}{h})_{gal } } ) \\ \end{array}\right . , \end{aligned}\ ] ] considering 12+log@xmath55(o / h)@xmath92 equals to the representative metallicity of a galaxy which we obtain by spectroscopically observing the galaxy without spatially resolving it . we also discuss the results with different @xmath94 formulations in section [ sec : sigma_zint ] . in figure [ fig : zint ] , we plot the metallicity distribution of h regions in the grb 980425 host galaxy together with those in the local galaxies . the grb 980425 host galaxy is the nearest grb host galaxy known and the only grb host galaxy for which metallicity is measured with a spatial resolution @xmath96 kpc , although the resolution @xmath97 pc is much larger than that achieved for the local galaxies ( e.g. , @xmath98 pc for m31 , @xcite ) . the internal metallicity variation observed within the grb 980425 host galaxy is naively consistent with those seen in the magellanic clouds . metallicity distribution of young stars which are formed in a galaxy might be significantly different from the metallicity distribution of h regions in the galaxy , if metallicity and sfr of h regions are correlated with each other . @xcite examined the correlation between h@xmath33 luminosity and metallicity of h regions in m31 using the narrow band photometric data by @xcite and the spectroscopic data by @xcite . although it is suggested that metallicity measurements of faint h regions with h@xmath33 luminosity @xmath99 [ erg s@xmath16 ] have luminosity dependent systematic error depending on the metallicity calibrator , no correlation was seen for h@xmath33 luminosity @xmath90 [ erg s@xmath16 ] . hence , we consider that the correlation between metallicity and sfr of h regions would not be strong even if it exists , although the existence of the correlation is not robustly ruled out . the metallicities of the nearby h regions discussed in this section are measured using different metallicity calibration methods than the kk04 calibration which we utilize in this study , and it is known that different metallicity calibrations are often inconsistent with each other @xcite . we consider this possible inconsistency as a part of the uncertainty of @xmath94 which we discuss in section [ sec : sigma_zint ] . it is broadly agreed that the metallicity decreases as the galactocentric radius increases in many galaxies ( e.g. , @xcite ) . however , we note that a scatter around the gradient at each radius is also significant . the scatter of metallicity at each galactocentric radius can be comparable to ( or evan larger than ) the radial variation ( e.g. , @xcite ) , and lowest - metallicity h regions in a galaxy may reside close to the center of the galaxy . it should be noted that the scatter of the metallicities in each galaxy may in part result from uncertainties of the metallicity measurements . however , similar variation of metallicity is also found in the mw using stars with ages of @xmath100 gyr as a tracer of metallicity @xcite . furthermore , in the m31 h region sample of @xcite , the weak auroral line [ o]@xmath314363 is detected for four h regions . with the electron temperature ( @xmath101 ) method of metallicity measurement using the auroral line ( e.g. , @xcite ) , which is not affected by uncertainties of the photoionization models , they showed that the four h regions have 12+log@xmath55(o / h ) @xmath1 8.38.4 . it is generally difficult to detect an auroral line of a high - metallicity h region , and it is not surprising that the four h regions with the auroral line detections are biased towards low - metallicities . however , the @xmath101 method measurements provides a confirmation that some h regions in m31 have low - metallicity in reality . [ erg s@xmath16 ] ) , and the host galaxy of grb 980425 which resides at @xmath102 . the best fit log - normal distributions are plotted together . _ lower panel _ : the median [ 12+log@xmath55(o / h)@xmath92 ] and the scatter ( @xmath94 ) of the h region metallicity distributions in the nearby galaxies . the pentagonal symbol represents the grb 980425 host galaxy . the correlations between 12+log@xmath55(o / h)@xmath92 and @xmath94 which we examine in this study : a broken linear model ( solid line ) , a single linear model ( dashed line ) , and constant models ( dotted lines ) are plotted together . , width=302 ] we parameterize the relation between grb production efficiency ( @xmath104/sfr ) and metallicity of a population of young stars . as discussed in section [ sec : zint ] , the metallicity of the stellar population may be different from that of the host galaxy . here we consider that @xmath105 is suppressed above a threshold metallicity ( o / h)@xmath106 by a factor of @xmath107 which is not necessarily zero ( step function model ) : @xmath108 we note that absolute scale of the efficiency ( i.e. , @xmath109 ) is marginalized when we consider normalized metallicity pdf of galaxies . we also consider the cases with the power - law @xmath105 model : @xmath110^{\alpha } & { \rm o / h > ( o / h)_{cut } } \end{array}\right . , \label{eq : eff_pow}\end{aligned}\ ] ] and with the exponential @xmath105 model : @xmath111^{\beta } ) . \label{eq : eff_exp}\ ] ] the schematic picture of the @xmath105 models is shown in figure [ fig : eff ] grb rate of a galaxy with a given set of sfr and 12+log@xmath55(o / h)@xmath92 ( median metallicity of star forming regions in the galaxy ) can be obtained by performing the following integration : @xmath112 where @xmath113 is the metallicity distribution of star forming regions in the galaxy which is represented by a log - normal distribution with a median value that corresponds to 12+log@xmath55(o / h)@xmath92 and a log scale scatter @xmath94 ( see section [ sec : zint ] ) . ) , power - law ( dashed , eq . [ eq : eff_pow ] ) , and exponential ( dotted , eq . [ eq : eff_exp ] ) models of @xmath105 as functions of metallicity . , width=302 ] assuming @xmath83 , the @xmath62-sfr relation , and the @xmath62-@xmath63 relation , we compute number density of galaxies in the parameter space of @xmath62 , sfr , @xmath63 : @xmath114 [ mpc@xmath115dex@xmath115 ] . the expected metallicity distribution of grb host galaxies can be derived integrating @xmath114 over @xmath62 and sfr weighted with @xmath117 ( equation [ eq : rgrb ] ) . each @xmath105 model discussed in section [ sec : eff ] has two free parameters : [ ( o / h)@xmath106 , @xmath107 ] , [ ( o / h)@xmath106 , @xmath33 ] , or [ ( o / h)@xmath106 , @xmath32 ] . the parameter ( o / h)@xmath106 determines the cutoff metallicity above which the grb production efficiency is suppressed , and @xmath107 , @xmath33 , or @xmath32 determines the sharpness of the cutoff . we search for parameter sets that reproduces the observed metallicity distribution of the low - redshift sample , through the parameter ranges @xmath120 , @xmath121 , @xmath122 , and @xmath123 with intervals of @xmath124 , and @xmath125 . the goodness of fit is determined using the kolmogorov - smirnov ( ks ) test against the `` median '' distribution shown in the upper left panel of figure [ fig : result ] . although the metallicity of the grb 111225a host galaxy is not measured and those of the host galaxies of grb 061021 and 090417b are not precisely constrained ( see section [ sec : metal ] ) , the uncertainties of these metallicities do not significantly affect the overall metallicity distribution of grb host galaxies . the resulting parameters from the fitting are not significantly changed within the range of this uncertainty . the bestfit model distributions are plotted in the upper left panel of figure [ fig : result ] together with the metallicity distribution of the low - redshift sample . the best fitting metallicity distributions with the three different @xmath105 models are very similar to each other . the sfr weighted metallicity distribution predicted from the same @xmath83 , the @xmath62-sfr relation , and the @xmath62-@xmath63 relation , which the grb host galaxies would follow if @xmath105 is not dependent on metallicity , is also shown together . the acceptable range of the parameters are shown as contour maps of the k - s test probability @xmath126 in figure [ fig : result ] . the parameter ranges that the @xmath127 are : step function : : : 12+log@xmath55(o / h)@xmath128 , @xmath129 power - law : : : 12+log@xmath55(o / h)@xmath130 , @xmath131 exponential : : : 12+log@xmath55(o / h)@xmath132 , @xmath133 the degeneracy between the two parameters is seen in the cases of the power - law and exponential models . the best fit 12+log@xmath55(o / h)@xmath106 is @xmath1 8.3 , and the sharper cutoff reproduces the observations better in any of the models . this means that majority of low - redshift grb host galaxies have higher metallicity than 12+log@xmath55(o / h)@xmath106 above which grbs can not be produced . in other words , most low - redshift grbs take place in star forming regions whose local metallicity is much lower than the representative value of their host galaxies . the model distribution with the step function model of @xmath105 does not significantly depend on @xmath107 when log@xmath134 , and the distribution with @xmath135 is quite similar to the best fit distribution shown in the upper left panel of figure [ fig : result ] . we also note that , in the limit of @xmath136 ( @xmath137 ) , the power - law ( exponential ) model of @xmath105 is identical to the step function model with @xmath135 . in the following , we only consider the step function model of @xmath105 with @xmath135 to examine the predicted properties of low - redshift grb host galaxies with our models . in figure [ fig : effgal ] , we compare the step function model of @xmath105 [ 12+log@xmath55(o / h)@xmath106 = 8.3 , @xmath135 ] with that convolved over galaxy scale ( @xmath139 ) as defined in equation [ eq : rgrb ] . @xmath139 steeply declines with increasing metallicity at around 12+log@xmath55(o / h)@xmath106 . however , galaxies with 12+log@xmath55(o / h)@xmath140 12+log@xmath55(o / h)@xmath106 have @xmath141 due to the internal variation of metallicity , and their contribution to the cosmic grb rate density is significant because they play a dominant role in the cosmic star formation at low - redshifts . the steep decline of @xmath139 at 12+log@xmath55(o / h)@xmath142 is also reported by @xcite based on a grb host galaxy sample collected from a wider range of redshifts where metallicity measurements of grb host galaxies are incomplete . we note that @xmath143 likely evolve with redshift because it depends on the properties of galaxies ( e.g. , @xmath94 ) , unlike @xmath144 which would be determined by stellar physics . in some cases , grb positions are spatially resolved from the center ( or brightest part ) of the host galaxies ( e.g. , @xcite ) . @xcite has shown that the metallicities at the grb positions are systematically lower than that of their host galaxies by @xmath145 dex , but not necessarily as low as suggested by our results . @xcite pointed out that the actual metallicity of a grb explosion site can not be obtained with a spatial resolution @xmath146 kpc which is typical of the observations of grb positions at @xmath147 , because the length scale of the ism metallicity variation is @xmath96 kpc in nearby galaxies ( e.g. , @xcite ) . our results suggest that the actual metallicity of the grb explosion sites ( or the progenitor star itself ) which is buried within the spatial resolution is significantly lower than the kpc scale metallicity currently observed , by up to @xmath1 0.5 dex . ) of a stellar population with a single metallicity [ solid line , the step function model with 12+log@xmath55(o / h)@xmath106 = 8.3 and @xmath135 ] , and the efficiency convolved over galaxy scale taking the internal variation of metallicity within each galaxy into account ( dashed line , [ eq : rgrb ] ) . , width=302 ] the @xmath117 weighted distribution of galaxies on the @xmath62-@xmath63 and @xmath62-sfr parameter planes , predicted using the step function model of @xmath105 with 12+log@xmath55(o / h)@xmath106 = 8.3 and @xmath135 , is plotted as contours in figure [ fig:2ddist ] . the sfr weighted model distribution , the assumed @xmath62-@xmath63 relation @xcite , the @xmath62-sfr relation @xcite , and the observed properties of the low - redshift sample ( table [ tab : metal ] & [ tab : msfr ] ) are plotted together . the observed grb host galaxies construct a sequence which is offset from the @xmath62-@xmath63 relation by 0.10.2 dex towards lower - metallicities , as previously noted by @xcite with a smaller incomplete sample , except for one remarkable outlier : the grb 011121 host galaxy , whose n2 and @xmath57 metallicities do not agree well with each other . the @xmath117 weighted model distribution also shows similar offset on the @xmath62-@xmath63 plane due to the higher @xmath139 in galaxies with lower 12+log@xmath55(o / h)@xmath92 . although the offset of the observed grb host galaxies may be slightly larger than that of the model distribution , this possible contradiction can be resolved by taking the sfr-@xmath63 correlation into account as we discuss in section [ sec : msfrz ] . in the @xmath62-sfr plane , the predicted galaxy distributions with the sfr and @xmath117 weightings are offset from the @xmath62-sfr relation by @xmath1 0.5 dex ( @xmath148 , as generally expected when the scatter of a relation is represented by a log - normal distribution ) . the observed grb host galaxies are broadly consistent with the predicted offset , although the sfr scatter of the host galaxies with log@xmath149 is large . the host galaxy of grb 031203 , which might have an agn , agrees well with the sequence of other grb host galaxies in the @xmath62-@xmath63 plane , while it is largely offset towards higher - sfr in the @xmath62-sfr plane . the outliner in the @xmath62-@xmath63 plane , the grb 011121 host galaxy , does not show any peculiarity in the @xmath62-sfr plane . the host galaxy with the lowest sfr is that of grb 060614 which is not associated with a sn . the grb 060614 host galaxy does not show any peculiarity in the @xmath62-@xmath63 plane . vs. @xmath63 ( _ upper panel _ ) , and @xmath62 vs. sfr ( _ lower panel _ ) . the solid and dotted contours represent the predicted distributions weighted with the @xmath117 and sfr , respectively . the contours are drawn at a logarithmic interval of 0.25 dex from the peaks of the distributions . the vertical dashed lines indicate naively estimated @xmath62 of the host galaxy of grb 111225a for which 12+log@xmath55(o / h)@xmath92 and sfr is not measured . the horizontal dashed lines represent 12+log@xmath55(o / h)@xmath92 and the lower - limit sfr of the host galaxy of grb 120714b whose @xmath62 is not known . the dashed lines with gray shaded regions are the @xmath62-@xmath63 relation @xcite and the @xmath62-sfr relation @xcite . the shaded regions represent 1@xmath66-scatter of the relations . the following host galaxies are marked with specific symbols : grb 011121 ( square , @xmath62-@xmath63 outlier ) , 031203 ( solid circle , possible agn , @xmath62-sfr outlier ) , 060614 ( dashed circle , @xmath62-sfr outlier ) . the vertical dotted line indicates the lower - limit of the @xmath62 range which we consider when we predict the metallicity distribution of grb host galaxies . , width=302 ] the model predictions of the metallicity distribution depend not only on the @xmath105 model , but also on the underlying assumptions of the properties of galaxies . in this section , we examine how our results are affected by some galaxy properties which is not well understood , namely the sfr-@xmath63 correlation and the variation of metallicity within a galaxy . to investigate the effect of the sfr-@xmath63 correlation , we consider the @xmath62-sfr-@xmath63 relation by @xcite , which is an extension of the relation by @xcite towards lower-@xmath62 . however , the metallicity calibration method used in @xcite is different from that used in this study ( kk04 ) . to examine the effect of the sfr-@xmath63 correlation in a consistent metallicity scale , we compute the @xmath150 gradient of the @xmath62-sfr-@xmath63 relation around the galaxy main - sequence at each @xmath62 . and assume the gradient without changing the @xmath62-@xmath63 relation @xcite , which has been used in the previous sections . it should be noted that the scatter of the @xmath62-@xmath63 relation at each @xmath62 may partly result from the sfr-@xmath63 correlation . @xcite showed that the metallicity scatter around the @xmath62-sfr-@xmath63 relation is smaller by 50% than that of the @xmath62-@xmath63 relation . thus we assume the scatter of the @xmath62-sfr-@xmath63 relation is 50% of that of the @xmath62-@xmath63 relation defined in equation [ eq : sigmamz ] . the resulting @xmath151 contour map and the bestfit metallicity distribution using the step function @xmath105 model are shown in figure [ fig : gmodel_pks ] ( top panel ) and figure [ fig : gmodel_dist ] , respectively . both the best fit parameters and the goodness of the fit are not significantly different from the case without the sfr-@xmath63 correlation . this situation is similar with the power - law and exponential models of @xmath105 . the small effect of the sfr-@xmath63 correlation on the metallicity distribution is not surprising . the sfr-@xmath63 gradient around the galaxy main - sequence is @xmath152 in the low-@xmath62 range ( log@xmath153 ) and shallower in higher-@xmath62 ranges . given that the grb host galaxies have systematically higher - sfr than the main - sequence by @xmath1 0.5 dex ( section [ sec : result ] ) , the expected effect of the sfr-@xmath63 correlation is @xmath154 , which is smaller than the uncertainties of the metallicity measurements . the sfr and @xmath117 weighted model distributions of galaxies on the @xmath62-@xmath63 and @xmath62-sfr planes , predicted assuming the sfr-@xmath63 gradient are shown in figure [ fig:2ddist_fmr ] . the @xmath105 model and the parameters are the same as in figure [ fig:2ddist ] . the sfr weighted galaxy distribution ( i.e. , no metallicity effect ) on the @xmath62-@xmath63 plane is also shifted towards lower - metallicities due to the sfr-@xmath63 correlation , but the shift is smaller than that of the @xmath117 weighted distribution . furthermore , the grb host galaxies have systematically lower-@xmath62 than the typical @xmath62 of the sfr weighted population ( log@xmath155 ) , which can not be explained solely by the effect of sfr-@xmath63 correlation ( e.g. , @xcite ) . although the sfr-@xmath63 correlation does not significantly affect the overall metallicity distribution of the grb host galaxies as mentioned above , the predicted number of grb host galaxies above the @xmath62-@xmath63 relation is reduced ( figure [ fig:2ddist_fmr ] , top panel ) . this improves the consistency between the predicted and observed distributions of grb host galaxies on the @xmath62-@xmath63 plane compared with the case without the sfr-@xmath63 correlation ( figure [ fig:2ddist ] ) . , but with different assumptions of the properties of galaxies . the step function model of @xmath105 is utilized . _ ( a ) _ : the correlation between sfr and log@xmath55(o / h ) is incorporated . _ ( b ) _ : @xmath94 is linearly correlated with 12+log@xmath55(o / h)@xmath92 . _ ( c ) and ( d ) _ : @xmath94 is a constant value ( 0.1 and 0.3 dex , respectively ) . , width=302 ] , but with different assumptions of the properties of galaxies . the step function model of @xmath105 is utilized . the plotted models are the same as those investigated in figure [ fig : gmodel_pks ] . the black histogram is the same as that in the upper left panel of figure [ fig : result ] , while the gray histogram represents the metallicity distribution without the host galaxies of grb 011121 , 031203 , 060505 , and 060614 whose burst classifications or metallicity measurements are uncertain . , width=302 ] , but with the sfr-@xmath63 correlation . the anomalies in the contour lines at log@xmath156 result from a discontinuity of the @xmath62-sfr-@xmath63 relation by @xcite . , width=302 ] as mentioned in section [ sec : zint ] , the detailed investigation of the internal ism metallicity variation is performed only for a small number of nearby galaxies . therefore , @xmath94 of general star forming galaxies and its correlation with other galaxy properties are highly uncertain . given that the nearby high - metallicity galaxies , namely the mw and m31 , have higher @xmath62 than the grb host galaxies in our sample , it is possible that ism in the grb host galaxies have different properties to those in the nearby galaxies . to investigate how our results depend on the assumed @xmath94 , we perform the parameter fittings assuming other @xmath94 formulations than the baseline model . here we consider a single linear relation between 12+log@xmath55(o / h)@xmath92 and @xmath94 , and also constant @xmath94 of 0.1 , 0.2 , and 0.3 dex ( figure [ fig : zint ] , bottom panel ) . however , it should be noted that the constant @xmath94 is disfavored by observations . the @xmath151 contour maps and the predicted metallicity distributions are shown in figure [ fig : gmodel_pks ] and figure [ fig : gmodel_dist ] for the cases with the single linear relation , and the constant @xmath94 of 0.1 and 0.3 dex . the single linear relation between 12+log@xmath55(o / h)@xmath92 and @xmath94 overpredicts the contribution of high - metallicity galaxies to the grb production even with the best fit parameters , due to the large @xmath94 of galaxies with 12+log@xmath55(o / h)@xmath157 . however , the statistical significance of the overprediction is low ( @xmath158 ) . the best fit parameters are similar to the case with the baseline model . when @xmath159 dex independently of 12+log@xmath55(o / h)@xmath92 , the preferred cutoff metallicity is 12+log@xmath55(o / h)@xmath160 . this result is consistent with the previous studies without the internal metallicity variation being explicitly treated in which the cutoff of the grb efficiency at 0.51@xmath59 is suggested ( e.g. , @xcite ) . even in this case , high @xmath105 in high - metallicity stellar population is disfavored ( @xmath161 ) . when @xmath162 dex , the cutoff metallicity is 12+log@xmath55(o / h)@xmath163 , which is close to the results with the baseline model . the results with the the constant @xmath164 0.2 dex are similar with those with 0.1 , and 0.3 dex , but with 12+log@xmath55(o / h)@xmath165 . some of the low - redshift grbs are less energetic than typical grbs that are observed up to higher - redshifts ( hereafter classical grbs ) , and possibly are different kind of phenomena . these bursts , so called low - luminosity grbs , have isotropic equivalent gamma - ray energy @xmath166 [ erg ] ( e.g. , @xcite ) . @xcite have shown that @xmath167 of a grb might be dependent on the metallicity of its host galaxy based on observations of the host galaxies of 5 grbs / xrfs at redshifts @xmath4 , while later studies @xcite have shown that there is no significant correlation between them with a larger sample from wider range of redshifts ( up to @xmath168 ) . because low - luminosity grbs can be detected only at low - redshifts @xmath169 , the comparison of the host galaxies and low - luminosity grbs and with those of classical grbs at @xmath170 might be affected by evolution of galaxy populations over the redshits . thus it is interesting to revisit the correlation between @xmath167 and host metallicity with our sample limited to low - redshifts but much larger than that of @xcite . @xmath167 of the low - redshift grbs and 12+log@xmath55(o / h)@xmath92 are shown in the left panel of figure [ fig : eiso_z ] . although the grbs in the pre-_swift _ era ( until 2004 ) shows a trend that grbs with larger @xmath167 occurr in lower - metallicity galaxies as reported in @xcite , we do not find significant correlation between @xmath167 and 12+log@xmath55(o / h)@xmath92 when grbs in the _ swift _ era are included ( correlation coefficient is -0.36 ) . grb 060505 and 060614 are widely known as grbs which are not associated with sne @xcite , and it is possible that they are actually short grbs that do not result from core - collapse of massive stars despite their long duration @xmath171 sec ( e.g. , @xcite ) . these bursts are marked with circles in figure [ fig : eiso_z ] . the host metallicities and @xmath167 of grb 060505 and 060614 are typical of low - redshift grbs in our sample . the metallicity distributions of the grb host galaxies excluding grb hosts with possible systematic error in their metallicity measurements or with uncertain burst classifications ( grb 011121 , 031203 , 060505 , and 060614 ) is shown in figure [ fig : gmodel_dist ] . the metallicity distribution without these peculiar objects does not significantly differ from the distribution of the whole sample . in this study , we have presented the overall metallicity distribution of the host galaxies of all grbs known at @xmath12 . however , the fraction of grbs with spectroscopic redshifts are @xmath1 30% even in recent observations ( e.g. , @xcite ) , and hence we may suffer from the redshift determination bias by collecting grbs with known redshifts . one possible source of the redshift determination bias is `` dark '' grbs ( e.g. , @xcite ) . a `` dark '' grb is a grb with a fainter optical afterglow than expected from its x - ray spectrum . roughly 1/3 of grbs are `` dark '' , and most of them ( @xmath172% ) result from dust extinction , while some may result from absorption by neutral hydrogen in inter - galactic medium at high - redshifts @xcite . redshift determination of grbs largely relies on optical spectroscopy of afterglows , and it is difficult to determine redshifts of `` dark '' grbs . thus a sample of grb host galaxy with known redshifts may be biased against dusty galaxies which may tend to have higher - metallicity . however , at low - redshifts , detection of a grb host galaxy is not quite challenging in most of the cases . assuming the stellar - mass to luminosity ratio derived by @xcite , a galaxy with @xmath173 , which is typical of the low - redshift grb host galaxies , would be observed with @xmath174band magnitude @xmath175 at @xmath176 . once the host galaxy of a grb is detected , x - ray afterglow localization by _ swift_/xrt with typical precision of a few arcseconds , which is available for @xmath172% of the _ swift_/bat - detected grbs , can associate the grb to the host galaxy with a certain level of confidence @xcite . and the redshift can be determined using emission lines of the host galaxy . in reality , many redshifts of the low - redshift grbs are determined by the emission lines of the host galaxies ( table [ tab : grbs ] ) , including that of a `` dark '' grb 090417b ( marked with a square in figure [ fig : eiso_z ] ) , while redshift determination by host emission lines is rarely available at higher redshifts . the fraction of the grbs whose redshifts are determined only by the host emission lines in the low - redshift sample ( 8/14 ) is naively comparable to the fraction of grbs without known redshifts in the _ swift_/bat - detected grbs . thus the bias against grbs in dusty environment is not necessarily effective at low - redshifts . it is possible that grbs whose redshifts are determined with the emission lines of their host galaxies are biased against grbs that occurred in galaxies with weak emission line fluxes . emission line luminosities of a galaxy is largely dependent on sfr in the galaxy , which correlates with @xmath62 and 12+log@xmath55(o / h)@xmath92 , in the sense that galaxies with higher @xmath62 and 12+log@xmath55(o / h)@xmath92 have more luminous emission lines . hence grbs that occurred in very low - metallicity host galaxies might be systematically missed in the low - redshift sample , especially at higher - end of the redshift range where detection of weak emission lines are difficult . however , such redshift dependent metallicity bias is not seen in our sample ( the right panel of figure [ fig : eiso_z ] , correlation coefficient is 0.007 ) . it is also interesting to compare the expected rate of low - redshift grbs to the actual finding rate to evaluate the success rate of identifying low - redshift grbs . based on the largest unbiased grb survey to date ( shoals ) , @xcite estimated the event rate density of bright grbs with @xmath177 [ erg ] at @xmath9 to be @xmath178 [ yr@xmath16gpc@xmath115 ] . considering the field of view of _ swift_/bat ( 1.4 steradian ) and the operation time until the end of march 2014 ( 9.3 years since december 2004 ) , the expected number of bright grbs detected by _ swift_/bat at @xmath12 in this time period is @xmath179 . on the other hand , the number of _ swift_/bat - detected grbs with @xmath177 [ erg ] in our sample is 4 ( figure [ fig : eiso_z ] ) , which suggests that the success rate of redshift determination is higher at low - redshifts compared with that of the overall population of grbs , although the estimation of the grb rate density is still uncertain . the faint end of the grb luminosity function is still controversial affected by sample selections and model assumptions ( e.g. , @xcite ) . hence it is difficult to estimate the actual rate and evaluate the redshift success rate of low - luminosity grbs . on the other hand , it is also true that redshift determination of a grb has failed ( partly ) due to the weak emission line of the host galaxy . as mentioned in section [ sec : obs:111225a ] , the redshift of grb 111225a has not been determined until @xcite reports detections of weak emission lines over the afterglow together with a marginal absorption line using an updated analysis tool . the failure of the redshift determination in the immediate follow up observations after the burst suggests that grb host galaxies with @xmath180 might be missed in redshift selected samples , given that the stellar mass of the grb 111225a host galaxy is naively @xmath181 ( section [ sec : host_msfr ] ) . another interesting case is the identification of the host galaxy of grb 020819b . grb 020819b has been considered to be a low - redshift burst at @xmath182 since radio afterglow observations by @xcite localized the burst on a `` blob '' which resides @xmath1 from a spiral galaxy at @xmath182 , and further observations of the host galaxy by @xcite suggested that the `` blob '' also resides at @xmath182 . however , recent extensive observations of the host galaxy by @xcite showed that the `` blob '' is actually a dusty starforming galaxy at @xmath183 , and the low - redshift spiral galaxy is possibly a foreground object unrelated to the grb . hence it is possible that a few of the low - redshift grbs in our sample whose redshifts are obtained only via their host galaxies actually reside in high - redshift galaxies which are too faint to be observed , and currently claimed host galaxies at low - redshifts are foreground objects . although most of the low - redshift grbs are closely located to the center of their host galaxies within @xmath184 unlike grb 020819b , it is difficult to robustly confirm that a `` low - redshift '' grb is actually at a low - redshift when the grb does not have an afterglow measured redshift or a confirmed sn association ( grb 060505 , 080517 , 060614 , 050826 , and 090417b , see table [ tab : grbs ] ) . furthermore , among the low - redshift grbs without robust redshift confirmations , grb 060505 and 080517 have offsets @xmath185 from the center of their host , although the large offsets are not surprising given their very low - redshifts @xmath186 . it should be noted that the originally claimed host galaxy of grb 020819b at @xmath182 was a massive - spiral galaxy with the highest metallicity in the known grb host galaxies . when a grb host galaxy sample is contaminated by unrelated forground galaxies , it is possible that the contaminants tend to be brighter galaxies because they can be found more easily than the actual host galaxies which are likely faint . missing redshift informations for grbs in faint galaxies and contamination by bright foreground galaxies may bias the metallicity distribution of the overall population towards higher - metallicities , in contrast to what is expected in the case of the bias against `` dark '' grbs . we have presented the overall metallicity distribution of the host galaxies of all grbs known at @xmath12 that occurred before the end of march 2014 , including the newly obtained emission line fluxes of the host galaxies of grb 060614 , 090417b , and 130427a . the low - redshift sample of grb host galaxies with complete metallicity measurements is essential to study the nature of grb progenitor in some aspects : 1 ) the high success rate of redshift determinations and host identifications , 2 ) the reporting bias can be eliminated by completing the metallicity measurements , and 3 ) a wealth of comparison sample is available . we have compared the metallicity distribution of the low - redshift sample with the predictions from the empirical formulations of the properties of low - redshift galaxies . to predict the metallicity distribution , we adopted models of grb efficiency ( @xmath187 ) as functions of metallicity . the metallicity of the progenitor stars considered here is not necessarily identical to that of its host galaxy . instead , we considered the metallicity variation within each galaxy motivated by the observations of h regions in nearby galaxies . the three different formulations of @xmath105 , namely the step function , power - law , and exponential models are examined . each of the three @xmath105 models has two free parameters that can be used to fit the metallicity distribution of the host galaxies . in either case , @xmath105 function with a sharp cutoff around 12+log@xmath55(o / h)@xmath188 8.3 ( @xmath189 ) reproduces the metallicity distribution of the low - redshift sample best . this cutoff metallicity is naively consistent with those predicted by the stellar evolution models ( @xmath1 0.1@xmath190 , @xcite ) . @xmath105 models with moderate or weak metallicity dependence [ @xmath191 are disfavored , although the statistical significance of the current constraints is still low . this result is in contrast to the results of some previous studies which have suggested the cutoff metallicity @xmath1 0.5@xmath192 ( e.g. , @xcite ) . this is because we take the internal variation of metallicity within each galaxy into account . our galaxy models also indicate the cutoff metallicity around 0.51@xmath59 , when the internal metallicity variation is assumed to be smaller ( 0.20.1 dex ) in any galaxy . however , the current model of the internal metallicity variation relies on the observed metallicity distributions of h regions in a small number of local galaxies , and the actual internal metallicity variation in grb host galaxies is hardly known . this uncertainty possibly affect our results significantly . the effect of the correlation beween sfr and metallicity of galaxies on the predicted metallicity distribution is smaller than the uncertainty of metallicity measurement . however , it may improve the agreement of the predicted and observed distribution of the grb host galaxies on the parameter plane of @xmath62 versus @xmath63 . the relation between the nature of grb progenitor stars and the observable properties of their host galaxy depends on the internal and global properties of galaxies . better understanding on the general population of galaxies at each redshift is essential to obtain more robust constraints on the properties of grb progenitors , in addition to gathering larger less biased sample of grb host galaxies . because we collected grbs with known redshifts ( @xmath0 ) in this study , it is possible that the sample of the host galaxies is biased for a kind of galaxy population in which grb redshift determination is easy ( e.g. , galaxies with smaller dust content ) , although the grb redshift determination is generally easier at lower - redshifts . however , the number of low - redshift grbs in our sample is close to the number expected from the latest unbiased surveys , suggesting a high success rate of redshift determinations at low - redshifts , although the statistical error is still large . furthermore , the obtained metallicities do not show significant systematic trend with respect to burst energy ( @xmath167 ) and redshift . thus there is no sign that the low - redshift sample of the grb host galaxies is biased compared to the actual population of grb host galaxies , although the tests are not robust yet . on the other hand , the failure of the redshift determination of grb 111225a in the immediate follow up observations after the burst suggests that faint ( and thus low - metallicity ) host galaxies might be systematically missed in the redshift selected sample . it is also possible that a few low - redshift galaxies which are aligned in the foreground of higher - redshift grbs contaminate the low - redshift sample as in the case of grb 020819b . the more precise estimation of the grb rate density and their properties , including those of high - redshift bursts which might contaminate low - redshift samples , will enable us to evaluate sampling biases in the redshift selected sample robustly . we thank gemini and subaru observatory staffs for their kind support for the observations . yn is supported by the research fellowship for young scientists from the japan society for the promotion of science ( jsps ) . this research has made use of the ghosts database ( www.grbhosts.org ) , which is partly funded by spitzer / nasa grant rsa agreement no . thanks are also due to nathaniel r. butler , jochen greiner , and daniel a. perley , for producing the useful online databases . we would also like to thank the anonymous referee for helpful comments . * * grb 980425 host galaxy * : the n2 method indicates @xmath193 and log@xmath194 = 7.3 , while the @xmath57 method does not have any valid solution . the [ o]@xmath314959 flux is not measured in @xcite , and we assume theoretically expected line ratio of [ o]@xmath314959/[o]@xmath315007 = 1/3 ( e.g. , @xcite ) . the lack of @xmath57 solution generally suggests @xmath195 8.18.7 , which is consistent with the n2 solution in this case . we adopt the n2 solution as the best estimate of the metallicity of this host galaxy . * * grb 060505 host galaxy * : the n2 method indicates @xmath196 and log@xmath194 = 7.2 ( the best estimate ) . the @xmath57 method does not have valid solution . * * grb 080517 host galaxy * : the emission line strength is presented as equivalent widths in @xcite . we derive the emission line fluxes assuming the continuum flux density to be 0.8 , 1.7 , and 1.5 @xmath197 erg s@xmath16@xmath65@xmath16 at the wavelength of [ o ] , h@xmath32[o ] , and h@xmath33[n ] , respectively . the n2 , @xmath57 upper - branch , and @xmath57 lower - branch solutions are @xmath198 , 8.68@xmath199 , and 8.38@xmath199 , with log@xmath194 = 7.2 , 7.2 , and 7.1 , respectively . we adopt the @xmath57 upper - branch solution as the best estimate . * * grb 031203 host galaxy * : the n2 method indicates @xmath200 ( log@xmath194 = 8.6 ) , and the @xmath57 method provides two solutions of @xmath201 and 8.32@xmath199 ( so called upper-/lower - branches , with log@xmath194 = 8.3 and 8.2 , respectively ) . + kk04 calibrated the metallicity indicators in a range of ionization parameter @xmath202 , and we need to extrapolate the relation when log@xmath194 is outside of this range . because the n2 index is sensitive to the ionization parameter , we do not consider the n2 solution for this host galaxy reliable , although the indicated super - solar metallicity suggests the upper - branch solution of the @xmath57 method is more likely than the lower - branch solution . thus we adopt the @xmath57 upper - branch solution as the best estimate . the @xmath57 upper - branch is less sensitive to the ionization parameter than the lower - branch and the n2 . + although the position of this galaxy on the bpt diagram is consistent with being a star forming galaxy ( figure [ fig : bpt ] ) , @xcite showed that the host galaxy might have an agn . the high @xmath60 parameter value also supports the existence of an agn in the galaxy . * * grb 060614 host galaxy * : due to the non - detection of [ n ] line , we can put only upper - limit on the n2 index ( n2 @xmath203 ) . the [ o]@xmath314959 line coincides with an strong night - sky emission line , and the flux can not be measured from the observed spectra . hence we assume the theoretically predicted line ratio of [ o]@xmath314959/[o]@xmath315007 = 1/3 . the upper- and lower - branch solutions of the @xmath57 method are @xmath204 and 8.35@xmath199 with log@xmath194 = 7.4 and 7.3 , respectively . assuming log@xmath194 = 7.37.4 , the n2 upper - limit indicates @xmath205 . thus we adopt the @xmath57 lower - branch solution as the best estimate . * * grb 030329 host galaxy * : the n2 , @xmath57 upper - branch , and @xmath57 lower - branch solutions are @xmath206 , 8.75@xmath199 , and 8.14@xmath199 , with log@xmath194 = 7.8 , 8.2 , and 7.8 , respectively . we adopt the @xmath57 lower - branch solution . * * grb 120422a host galaxy * : the n2 method indicates @xmath207 and log@xmath194 = 7.3 ( the best estimate ) . the @xmath57 method does not have valid solution . * * grb 050826 host galaxy * : the n2 , @xmath57 upper - branch , and @xmath57 lower - branch solutions are @xmath208 , 8.86@xmath199 , and 8.14@xmath199 , with log@xmath194 = 7.5 , 7.5 , and 7.3 , respectively . we adopt the @xmath57 upper - branch solution . * * grb 130427a host galaxy * : the n2 , @xmath57 upper - branch , and @xmath57 lower - branch solutions are @xmath209 , 8.67@xmath199 , and 8.33@xmath199 , with log@xmath194 = 7.4 , 7.4 , and 7.3 , respectively . we adopt the @xmath57 upper - branch solution as the best estimate . however , we note that the focas too spectrum from which we measure the emission line fluxes of [ o]@xmath313727 , h@xmath32 , and [ o]@xmath2104959,5007 , is obtained without adc . hence it is possible that the line fluxes suffer from differential atmospheric refraction . + to evaluate the effect of this systematic uncertainty , we also examine the independently measured emission line fluxes reported in k15 ( shown in table [ tab : lines ] ) . the k15 line fluxes are systematically weaker than those in the gmos+focas spectra possibly due to larger loss of light at the slit , however we can compare relative flux ratios between the lines . + with the k15 line fluxes , the n2 , @xmath57 upper - branch , and @xmath57 lower - branch solutions are @xmath211 , 8.70@xmath212 , and 8.34@xmath213 . we note that the detection of the [ n ] line is marginal in k15 ( s / n = 2.5 ) . although it is difficult to determine which of the @xmath57 solutions is true only with the k15 spectrum , if we adopt the upper - branch solution as suggested by our gmos spectrum , it agrees well with the estimate from the gmos+focas spectra . * * grb 090417b host galaxy * : the [ o ] , h@xmath32 , and [ o ] line fluxes are not known . assuming @xmath214 log@xmath215 ( the @xmath60 parameter range covered by the other grb host galaxies in the sample excluding that of grb 031203 ) , the n2 method indicates @xmath216 = 8.68.9 . * * grb 061021 host galaxy * : the [ n ] line is not detected providing n2 @xmath217 . the @xmath57 upper - branch and lower - branch solutions are @xmath218 and @xmath219 , respectively ( log@xmath194 = 7.5 in either case ) . the n2 upper - limit is not deep enough to reject either of the @xmath57 solutions . combining the @xmath57 upper - branch and lower - branch solutions , we estimate the metallicity of this host galaxy to be 8.288.66 . * * grb 011121 host galaxy * : the n2 , @xmath57 upper - branch , and @xmath57 lower - branch solutions are @xmath220 , 8.69@xmath199 , and 8.33@xmath199 , with log@xmath194 = 7.2 , 7.4 , and 7.2 , respectively . we adopt the @xmath57 lower - branch solution . * * grb 120714b host galaxy * : the n2 method indicates @xmath221 , although the detection of the [ n ] line is marginal . the @xmath57 method , which is not affected by the low - s / n of the [ n ] line , provides the upper - branch and the lower - branch solutions : @xmath222 and 8.43@xmath199 , respectively ( log@xmath194 = 7.5 in either case ) . we adopt the @xmath57 upper - branch solution , which agrees with the n2 solution .

nowadays cross - kerr nonlinearity is considered as a promising tool for quantum computation and non - classical state generation @xcite . an entanglement arising between modes participating in the cross - kerr interaction can be used for establishing an interface between matter qubits and ` flying ' light photonic @xcite , for generation of non - gaussian states and schrdinger - cat states , and for performing quantum gate operations @xcite . an interest to cross - kerr nonlinear interactions is heated up by both developing ways to implement effectively even very weak non - linearities ( which one commonly expects to have in practice ) , and by discovery of methods to produce sufficiently large self - kerr and cross - kerr nonlinearities ( such as implementation of the electromagnetically induced transparency ( eit ) and photonic crystals @xcite ) . decoherence is a main practical obstacle to implementations of schemes using kerr and cross - kerr nonlinearities . a genuine example of quantum state degradation due to losses is a decoherence of a quantum superposition state , the effect drastically enhanced if speaking of a macroscopic superposition state of the schrdinger cat type . already more than 20 years ago the self - kerr nonlinearity was proposed as a tool for generating such a schrdinger - cat state ( more precisely , a superposition of two coherent states with the same amplitudes but opposite phases ) @xcite . however , photon losses turn this superposition into statistical mixture of two coherent states with the rate proportional to the square modulus of the amplitude of these states . modifications of the scheme for the cross - kerr nonlinearities or four - wave mixing brought no advantage with respect to photon loss @xcite . this unfortunate circumstance made one look for the ways to circumvent the problem of decoherence that inevitably accompanies kerr nonlinearity . recent suggestions in this direction are based on the conditional preparation of desired states ( which brings into consideration an additional problem of the finite detection efficiency ) , and are aimed to exploit weak nonlinearities . recently , even a way to produce cat - states ` on demand ' was suggested using a source of single - photons @xcite . in our work we want to discuss an aspect of the decoherence which has been seldom discussed when considering an influence of losses on states generated via kerr nonlinearity . namely , we address losses arising _ in the process _ of generation and not due to propagation of the generated state via lossy channels . we concentrate our attention on a feature that might be quite significantly pronounced in modern schemes of generating large kerr nonlinearity : the modal loss can be strongly correlated . indeed , the modes occupy the same volume and interact with the same physical systems which form the reservoirs . also , if the kerr - nonlinearity scheme implies a sufficiently strong dispersive coupling of light modes to emitters , then coupling of these emitters to dissipative reservoirs might also appear to be quite strong . as a result , this would mean strongly correlated modal losses . for example , in photonic crystals high density of states in the vicinity of a modal frequencies and emitter s transition frequency can cause the strong emitter - field coupling ; but it would also imply higher population loss of emitters due to coupling to radiative reservoirs . dephasing losses of emitters would as well invoke a correlated modal dephasing . coupling to correlated reservoirs can drastically change state dynamics in comparison with loss to uncorrelated reservoirs . for example , it was demonstrated that coupling to the common reservoir preserves entanglement of a two - mode state @xcite . moreover , coupling to the common reservoir is capable of creating an entanglement between states of initially unentangled modes even in absence of any direct interaction between them . in our work we demonstrate both how the correlated loss arises via kerr nonlinear process , and how it affects the generated states . for this purpose we derive analytic solution generalizing a powerful and illustrative method of chaturvedi and srinivasan @xcite . on a number of examples we show how the correlated loss enhances and creates intermodal correlations and even entanglement , and might lead to generation of entangled states quite different from those generated in the same scheme without loss . correlated loss can result in the significantly enhanced robustness of the generation scheme . the outline of the paper is as follows . in sec . ii and in the related appendices we describe how cross - correlation terms emerge via correlations of markovian reservoirs ; we consider an example of the emitter - field interaction schemes producing correlated modal losses in the section iii . then in sec . iv we describe the method for obtaining exact solutions of the cross - kerr nonlinear interaction between modes in presence of losses to uncorrelated reservoirs and give generalization of the method for some cases of correlated losses . in the sec . v we analyze influence of losses in the nonclassical state generation process for the case of uncorrelated loss . some examples of correlated losses are considered in sec . to illustrate clearly an influence of correlated and uncorrelated losses , we restrict ourselves to the markovian loss accounting for photon losses and dephasing of interacting modes . we start from the general effective hamiltonian @xmath0 describing both self- and cross - interaction ( for the moment we refrain from detailing it ) and interaction of modes with reservoirs responsible for losses @xmath1 : @xmath2 here we use the interaction picture with respect to the free hamiltonians of reservoirs and the modes participating in the interaction process . these modes are described by usual bosonic creation and annihilation operators satisfying @xmath3=[a_2,a_2^{\dagger}]=1 , \quad [ a_1,a_2^{\dagger}]=[a_1,a_2]=0.\ ] ] the operators @xmath4 and @xmath5 describe reservoirs responsible , correspondingly , to the photon losses in modes @xmath6 and @xmath7 , and to dephasing of these modes . they may include also stochastic variables describing different realizations of reservoirs . it should be emphasized that reservoir operators @xmath4 and @xmath5 ( together with the initial state of the reservoir ) completely describe the reservoir properties with respect to the interaction with the modes . these operators are built on the basis of underlying microscopic model and account for all relevant physical parameters . for example , if the photon loss reservoir of the first mode is composed of electromagnetic field modes with frequencies @xmath8 , described by the creation and annihilation operators @xmath9 , then @xmath10 where each @xmath11 is the constant of interaction of the mode @xmath6 with the @xmath12th mode of the reservoir . sets of frequencies @xmath8 and interaction constants @xmath11 describe completely physical properties of the reservoir . in particular , if the reservoir is the set of electromagnetic modes of a non - absorbing dielectric structure , they are found from the eigensolutions of maxwell s equations for this structure @xcite . we make a number of standard assumptions in considering decoherence : coupling with the reservoirs is weak , the initial state of the modes @xmath13 are uncorrelated with initial state of reservoirs , and correlation times of reservoirs are small enough to enable an implementation of the born - markov approximation . then , using , for example , a time - convolutionless projection operator technique @xcite , one can obtain the master equation , the liouville equation of the following form : @xmath14 + { 1\over 2 } [ ( \gamma_1-\gamma_{12})\mathcal{l}(a_1)+ ( \gamma_2-\gamma_{12})\mathcal{l}(a_2)+ ( d_1-d_{12})\mathcal{l}(a_1^{\dagger}a_1)+(d_2-d_{12})\mathcal{l}(a_2^{\dagger}a_2)]\rho(t)+\\ { 1\over 2 } [ \gamma_{12}\mathcal{l}(a_1+a_2)+d_{12}\mathcal{l}(a_1^{\dagger}a_1+a_2^{\dagger}a_2)]\rho(t ) , \label{master_markov}\end{aligned}\ ] ] where the superoperator @xmath15 acts on the density matrix as @xmath16 here we set @xmath17 for simplicity . for more details of the derivation see appendix a. by construction , the master equation ( [ master_markov ] ) provides for non - negative definite @xmath18 for arbitrary @xmath19 . note that we call reservoirs `` correlated '' , if their integrated cross - correlation function is non - zero , for example , @xmath20 and , correspondingly , the coefficient @xmath21 defined in eq . ( [ gamma12 ] ) is non - zero . of course , one can always transform eq . ( [ master_markov ] ) to the diagonal form . however , in this case lindblad operators ( i.e. operators like @xmath22 in the diagonal form ( [ diagonal form ] ) ) will be the linear superpositions of the former lindblad operators , and the transformed equation will be still describing a coupling between physical objects represented by these original lindblad operators ( modes @xmath23 in our case ) . for example , the possibility that both modes are coupled to the same reservoir ( either the photon loss reservoir or the dephasing one ) corresponds to an equality in relations ( [ rates_unequalities ] ) in appendix a. in this case the operators describing the reservoir are proportional to each other , say , @xmath24 and @xmath25 . equation ( [ master_markov ] ) then reduces to @xmath14 + { 1\over 2 } [ \gamma_{1}\mathcal{l}(a_1+xa_2)+d_{1}\mathcal{l}(a_1^{\dagger}a_1+ya_2^{\dagger}a_2)]\rho(t ) . \label{master_markov_corr}\end{aligned}\ ] ] the most important point here is that in this equation both modes behave like a single object with respect to relaxation . such a ` decoherence ' is able to induce enduring entanglement between modes @xmath6 and @xmath7 ( an example is shown in the appendix b ) . furthermore , interacting light modes might experience correlated losses due to fact that they both interact with the same atoms . one can illustrate the mechanism of such a loss by the following qualitative consideration . let us consider a general hamiltonian describing the light - atom interaction @xmath26 plus terms describing interaction of atoms with the dissipative reservoirs @xmath27 where @xmath28 and @xmath29 are operators describing the atoms and dissipative reservoirs , respectively . the hamiltonian of this form describes a general interaction between field modes and emitters of some type as the correlated loss can occur in different physical systems . provided atom - reservoir interactions in ( [ ham - loss ] ) is sufficiently weak as not to perturb much interaction between light modes and atoms , the following approximation can be used for the atomic operators : @xmath30 where @xmath31 , and t denotes the time - ordering operator . after averaging out the atomic variable , the terms that describe reservoir - mode coupling in the effective interaction hamiltonian will take the form of @xmath32 . here @xmath33 and @xmath34 denotes the averaging over atomic states ( in general , over emitter s states ) . thus , one can see that coupling of light modes to the same atom ( emitter ) interacting with the dissipative reservoir under the condition of adiabatic elimination of emitter s variables leads directly to the mode - reservoir interaction terms in the resulting effective hamiltonian . note , that these terms remain linear in reservoir operators @xmath29 . hence one can derive a master equation averaging over the reservoir in a standard way . in the appendix b we give examples of the derivation of effective hamiltonians and the corresponding master equations discussing the simplest two - mode jaynes - cummings system ( fig . [ fig1](a ) ) . the described scenario of how correlated loss emerge is quite general , and can take place for a wide range of schemes involving light - shift - induced photonic nonlinearities . a three - level @xmath35-system interacting with classical driving and quantum fields represent a more realistic example of the scheme with the correlated loss . [ lambda - system ] consider the large cross - kerr nonlinearity generation suggested in ref . @xcite and depicted in fig . [ fig1](b ) . there two modes are coupled to the transition between 1 and 3 levels of the @xmath35-system in presence of two classical driving fields on transitions 1 - 2 and 1 - 3 . in the rotating - wave approximation and in the interaction picture with respect to free hamiltonian of the reservoir , the problem is described by the following interaction hamiltonian : @xmath36\sigma_{13}+\omega\exp\{{{\rm i}}\delta_2t\}\sigma_{23}+\mbox{h . c. } , \label{hamiltonian_lambda_system}\end{aligned}\ ] ] where @xmath37 is the rabi frequency of the driving fields ; @xmath38 , @xmath39 ; and @xmath40 are the interaction constants for the coupling of a light mode to an emitter ( atom ) . here for simplicity we have taken into account only losses on the transition 1 - 3 . the setup depicted in fig . [ fig1](b ) can be realized in toroidal microcavities , where @xmath6 and @xmath7 correspond to the clockwise and counter - clockwise propagating modes @xcite . as usually , we assume the markovian reservoir and the following conditions hold : @xmath41 from the hamiltonian ( [ hamiltonian_lambda_system ] ) , the dynamics is governed by @xmath42(\sigma_{33}-\sigma_{11})+{{\rm i}}\omega\exp\{{{\rm i}}\delta_2t\}\sigma_{21}. \label{31equation_lambda_system}\end{aligned}\ ] ] using the approach described in ref . @xcite , we assume that level 3 remains practically unpopulated , @xmath43 and @xmath44 , as well as @xmath45 . thus the @xmath35-system is prepared in the superposition of the metastable levels 1 and 2 ( namely , in the state @xmath46 ) . then , neglecting small and rapidly oscillating terms , one obtains from eqs.([hamiltonian_lambda_system ] , [ 31equation_lambda_system ] ) the following master equation @xmath47+{1\over2}\left[(\gamma_1-\gamma_{12 } ) \mathcal{l}(a_1)+(\gamma_2-\gamma_{12})\mathcal{l}(a_2)\right]\rho(t)+ \\ { \gamma_{12}\over2}\mathcal{l}(a_1+a_2)\rho(t ) , \label{master_lambda_model}\end{aligned}\ ] ] where @xmath48 the master equation ( [ master_lambda_model ] ) describes the cross - kerr and self - kerr interactions of two light modes plus their coupling to the correlated reservoirs . as follows from the model depicted in fig . [ fig1](b ) ) , it is actually the same reservoir : one can see that @xmath49 . naturally , loss rates @xmath50 are much less than the loss rate of the emitter . however , if one deals with input modes in a coherent state of rather large amplitude for generating large cross - kerr nonlinearities , then even comparatively small losses can strongly influence mode dynamics . below we consider examples of such an influence , e.g. , an example of a single - mode schdinger - cat state . the occurrence of a correlated modal loss due to presence of the emitter ( atom in our case ) has been noticed in ref . they have also pointed out that the loss rate should be proportional to the square of the ratio of the mode - emitter interaction constant and the detuning . however , we would like to emphasize that this loss rate does not depend on the population of level 3 , as it can be seen from eq . ( [ 31equation_lambda_system ] ) . now let us turn to the specific nonlinear interaction described by the same type of the hamiltonian as in the examples above and in the appendix b. we will consider the effective hamiltonian in the master equation ( [ master_markov ] ) @xmath51 in the following general form @xmath52 it describes the cross - kerr and self - kerr interaction with nonlinear coefficients @xmath53 of two modes ( or mode superpositions ) @xmath6 and @xmath7 . to solve the problem described by the master equation ( [ master_markov ] ) with the hamiltonian ( [ h_0_kerr ] ) , we adopt a simple and illustrative ` thermofield ' notation @xcite . essentially , instead of a density matrix acting on some space @xmath54 , say , @xmath55 , where @xmath56 is the fock state with @xmath57 photons in @xmath54 , we consider a state vector @xmath58 in an extended space @xmath59 , where @xmath60 is the fock state with @xmath61 photons in @xmath62 . so when the mode operators @xmath63 and @xmath64 in the master equation ( [ master_markov ] ) act on the density matrix from the left , one introduces operators @xmath65 and @xmath66 is such a manner that @xmath67 obviously , operators @xmath63 and @xmath64 commute with @xmath68 and @xmath66 . an action of the superoperator @xmath69 on the density matrix can be represented in the thermofield notation as @xmath70 also , the commutator of any function of the operators @xmath63 and @xmath64 , e.g. , the hamiltonian @xmath71 is re - written as : @xmath72\longrightarrow \mathbf{h}_0(t)|\rho\rangle=\left(h_0(t;a , a^{\dagger } ) -h_0(t;{\tilde a},{\tilde a}^{\dagger})\right)|\rho\rangle.\ ] ] with help of these notations eq . ( [ master_markov ] ) can be represented in the ` hamiltonian ' form as @xmath73 its solution is then of the form @xmath74 the advantage of using thermofield notation over more traditional algebraic manipulation with superoperators is that in many situations ( and , particularly , ones of our interest ) it enables to simplify , make more illustrative and less cumbersome finding the solution ( [ general solution ] ) and estimation of time - dependent matrix elements . in particular , it allows to represent in a simple form a factorization of the superoperator @xmath75 into multipliers with easily estimated actions on the number states @xcite . to illustrate this , let us consider a simple problem of modal loss in a single mode described by the equation @xmath76 the key to solving this equation lies in the observation that the operators @xmath77 generate the su(1,1 ) algebra with the casimir invariant @xmath78 . using the disentangling theorem @xcite for this group , we arrive at the simple result @xmath79 \exp[(1-{{\rm e}}^{-\gamma_at})a_-]|\rho(0)\rangle . \label{solution example}\end{aligned}\ ] ] generally , the dynamics described by the solution ( [ solution example ] ) leads to transforming initially pure states into mixtures . however , for a coherent initial state with the amplitude @xmath80 , @xmath81 equation ( [ solution example ] ) gives the following result : @xmath82 returning now to the effective hamiltonian ( [ h_0_kerr ] ) that describes cross - kerr and self - kerr interaction of two modes ( or modal superpositions ) @xmath6 and @xmath7 , we write it in the thermofield notation as follows : @xmath83 operators in eq . ( [ h_0_kerr_general ] ) are @xmath84 , @xmath85 . for completely uncorrelated reservoirs of different modes ( i.e. @xmath86 ) the master equation ( [ master_markov_term ] ) with the hamiltonian @xmath26 given by eq . ( [ h_0_kerr_general ] ) can be solved exactly using the following factorization @xcite : @xmath87t\right\ } \\ \times \exp[({{\rm i}}p_1-\gamma_1)a_3^{(1)}t ] \exp[({{\rm i}}p_2-\gamma_2)a_3^{(2)}t ] \exp[\gamma_{1-}a_-^{(1 ) } ] \exp[\gamma_{2-}a_-^{(2 ) } ] \label{ham}\end{aligned}\ ] ] where superoperators @xmath88 are defined similarly as in eq . ( [ algebra definition ] ) and @xmath89}{{{\rm i}}p_k-\gamma_k } , \quad p_k=2\sum\limits_{l=1}^2\chi_{kl}a_0^{(l)}.\end{aligned}\ ] ] it is useful to note that in eq . ( [ ham ] ) all multipliers apart from two last ones are diagonal in the number - state basis . also , it is easy to see that @xmath90=0 $ ] , operators @xmath91 are diagonal in the number - state basis , and operators @xmath92 are simply products of annihilation operators . thus , eq . ( [ ham ] ) provides for simple analytic solutions both for coherent initial states of interacting modes . the solution ( [ ham ] ) for uncorrelated reservoirs can be straightforwardly generalized for some special cases of correlated reservoirs and the hamiltonian ( [ h_0_kerr ] ) . for example , let us consider the problem without dephasing , @xmath93 , and introduce the rotated mode operators @xmath94 as @xmath95 where @xmath96 . then for the non - unitary part of the master equation ( [ master_markov_term ] ) one has @xmath97 where @xmath98 , @xmath99 . obviously , if the transformation ( [ rotation ] ) leaves the form of the hamiltonian ( [ h_0_kerr ] ) invariant , one can derive an exact solution in the way described in this section . in this section we consider specific effects of the uncorrelated losses in the process of the cross - kerr nonlinear interaction . notably , the losses in such a nonlinear process can lead to loss - mediated correlations between the modes , as seen from eq . ( [ ham ] ) . these intermodal correlations modify the effect of losses on the quantum state generated in the cross - kerr interaction with respect to the result , which one would intuitively expect treating the generation and loss separately , i.e. subjecting to loss a state that has been produced without losses . consider a particular problem of generating an entangled two - mode state from initially uncorrelated coherent states . we assume that the nonlinearity is given purely by the cross - kerr interaction , i.e. we put @xmath100 in the hamiltonian ( [ h_0_kerr ] ) . producing entangled states this way is important in a number of schemes of quantum computation and communication using continuous variables . we assume modes @xmath6 and @xmath7 to be initially in coherent states with amplitudes @xmath101 and @xmath102 , respectively . as was pointed in the previous section , for this choice of initial states , the solution given by eq . ( [ ham ] ) has a simple form : @xmath103t}2\right]\exp\left\{{{\rm i}}\chi t ( a_1^\dagger a_1a_2^\dagger a_2-\tilde a_1^\dagger\tilde a_1\tilde a_2^\dagger\tilde a_2 ) -\frac{\gamma_1t}2(a_1^\dagger a_1+\tilde a_1^\dagger\tilde a_1 ) -\frac{\gamma_2t}2(a_2^\dagger a_2+\tilde a_2^\dagger\tilde a_2)\right\}\\ \times \exp\left[\frac{\gamma_1\left({{\rm e}}^{{{\rm i}}\chi t(a_2^\dagger a_2-\tilde a_2^\dagger\tilde a_2)-\gamma_1t}-1\right)}{{{\rm i}}\chi ( a_2^\dagger a_2-\tilde a_2^\dagger\tilde a_2)-\gamma_1}\,|\alpha_1|^2\right ] \exp\left[\frac{\gamma_2\left({{\rm e}}^{{{\rm i}}\chi t(a_1^\dagger a_1-\tilde a_1^\dagger\tilde a_1)-\gamma_2t}-1\right)}{{{\rm i}}\chi ( a_1^\dagger a_1-\tilde a_1^\dagger\tilde a_1)-\gamma_2}|\alpha_2|^2\right ] |\alpha_1\rangle|\tilde\alpha_1^*\rangle|\alpha_2\rangle|\tilde\alpha_2^*\rangle . \quad \label{evolution}\end{aligned}\ ] ] a distinctive feature of the solution ( [ evolution ] ) is that the exponential terms contain only operators diagonal in the number - state basis . thus , in this basis the solution ( [ evolution ] ) turns into @xmath104t\right\ } } \\ & & \times\exp{\left\{{{\rm i}}\chi t ( km - ln ) -\frac{\gamma_1t}{2}(k+l)-\frac{\gamma_2t}2(m+n)+f^{(2)}_{mn}(t)+f^{(1)}_{kl}(t)\right\ } } \label{evolution_fock}\end{aligned}\ ] ] where @xmath105 now let us analyze the solution ( [ evolution_fock ] ) in more detail and consider for the moment the case of no dephasing ( @xmath106 ) . in this case the expression for the purity of the state given by the solution ( [ evolution_fock ] ) is quite similar in structure to this solution itself : @xmath107 to describe effects of simultaneous damping ( photon loss ) and cross - kerr nonlinearity on the generated state , consider first a ` short - time regime ' where the following relations hold for all the photon numbers @xmath108 ( @xmath109 ) that have a non - negligible probability of occurring in the state @xmath110 ( @xmath111 ) : @xmath112 then expressions ( [ f - functions ] ) can be expanded as @xmath113^j\over(j+1 ) ! } , \quad f^{(1)}_{kl}(t)= \gamma_2t|\alpha_2|^2\sum\limits_{j=0}^{\infty}{[{{\rm i}}\chi t(k - l)-\gamma_2t]^j\over(j+1)!}. \label{f - functions_expand}\end{aligned}\ ] ] the first three terms of these expansions already describe quite well typical effects produced by simultaneous damping and cross - kerr nonlinearity . for example , for @xmath114 in the limit of small times one has @xmath115-{{{\rm i}}\over6}\gamma_1\chi t^2(m - n)\right).\ ] ] obviously , first three terms in the round brackets of this expressions do not lead to breaking of the purity of the state . retaining only them in expansion renders unity value for the right - hand side of eq.([evolution_purity ] ) . indeed , assuming @xmath116 one obtains the time - dependent density matrix formally coinciding with the result for no photon loss @xcite : @xmath117^k[\alpha_1^*(t)]^l}{\sqrt{k!l ! } } \overline{|\alpha_2^*(t){{\rm e}}^{{{\rm i}}\chi lt}}\rangle . \label{rho_no_dephasing}\end{aligned}\ ] ] here time - dependent amplitudes do not depend on the numbers @xmath108 : @xmath118 the state given by eqs . ( [ evolution_fock ] ) remains negligibly affected by losses if @xmath119 . thus , the considered scheme of non - classical state generation is quite robust with respect to photon loss ( in drastic difference with propagation losses of already generated cat - state where the off - diagonal terms @xmath120 will decay with the rates proportional to @xmath121 ) . in addition , it is interesting to note , that one might be able to satisfy conditions @xmath119 in schemes involving dispersive atom - field interactions in qed where it is possible to restrict losses to the photon loss of cavity modes @xcite . moreover , further in this work we consider ways to circumvent an influence of losses by making them correlated . remarkably , a purity of the generated bimodal state can be preserved not only in the case of small losses , but also for large loss . indeed , in the limits of large losses one can consider @xmath122 and @xmath123 as small quantities and expand functions ( [ f - functions ] ) in the following manner : @xmath124 + \chi^2(k - l)^2\left[\left({t^2\over2}+{t\over\gamma_2}+{1\over\gamma_2 ^ 2}\right){{\rm e}}^{-\gamma_2t}- { 1\over\gamma_2 ^ 2}\right ] , \\ \nonumber |\alpha_1|^{-2}f^{(2)}_{mn}(t)\approx ( 1-{{\rm e}}^{-\gamma_1t})-{{\rm i}}\chi(m - n)\left[\left(t+{1\over\gamma_1}\right){{\rm e}}^{-\gamma_1t}-{1\over\gamma_1}\right]+ \chi^2(m - n)^2\left[\left({t^2\over2}+{t\over\gamma_1}+{1\over\gamma_1 ^ 2}\right){{\rm e}}^{-\gamma_1t}- { 1\over\gamma_1 ^ 2}\right ] . \label{tailor}\end{aligned}\ ] ] note that this approximation holds for arbitrary interaction times . if the interaction time is sufficiently large to fulfill the conditions @xmath125 then the state given by eqs . ( [ evolution_fock ] ) is practically pure . under the conditions ( [ long_time_conditions ] ) this state is of the form described by eq . ( [ rho_no_dephasing ] ) with the time - dependent amplitudes given by @xmath126\right\ } , \\ \alpha_2(t)=\alpha_2\exp\left\{-\frac{\gamma_2t}{2}-{{\rm i}}\chi|\alpha_1|^2 \left[t{{\rm e}}^{-\gamma_1t}-\gamma_1^{-1}(1-{{\rm e}}^{-\gamma_1t})\right]\right\}. \label{functions}\end{aligned}\ ] ] clearly , purity of the resulting state in the long time - limit is precisely a consequence of a strong photon loss . in this way strong photon loss paradoxically suppresses state mixing predicted by the general solution ( [ evolution_fock ] ) . there is another interesting feature that distinguishes the losses occurring in the process of kerr interaction from the losses that take place after the interaction . in particular , the non - gaussian state generated by the cross - kerr interaction in the scheme discussed in @xcite can retain its non - classical features even for the loss level , which would completely eliminate any such features in case of free propagation of the state . to be specific , a typical signature of non - classicality of a quantum state is the fact that its wigner function is negative in some regions of the phase space @xcite . we will show that for a 50% photon loss occurring in the scheme @xcite during the cross - kerr interaction , the wigner function of the output state retains its negativity while the same photon loss occurring after the interaction would make the wigner function necessarily positive . of the non - classical non - gaussian state of ref . @xcite exhibits strong negativity . wigner function is viewed along the imaginary axis of the @xmath127 plane . , width=377 ] for about @xmath128 photon loss in mode @xmath63 corresponding to @xmath129 . the negative region of the wigner function is now negligible , but still present , and the amplitude of the state is clearly damped . , width=377 ] in the scheme @xcite for generating non - classical states , two coherent states in modes @xmath130 ( @xmath131 in notations of @xcite ) interact via cross - kerr effect in a non - linear medium and subsequently the @xmath132-quadrature of mode @xmath7 is measured . the resulting state of mode @xmath6 exhibits wigner function with negative regions ( see fig . [ wigner0 ] ) and a characteristic crescent ( or banana ) shape . suppose mode @xmath6 is subject to losses during the cross - kerr interaction , i.e. , @xmath133 in eq . ( [ evolution ] ) while we assume @xmath134 . the photon loss in mode @xmath6 is given by the reduction of the coherent amplitude described by eq . ( [ functions1 ] ) . consider the situation when the mean photon loss is @xmath128 . this corresponds to @xmath135 . the plot of the corresponding wigner function is shown in fig . [ wigner07 ] which shows clearly that although the negative region of the wigner function is surpressed , it is still present . now compare this with the situation when losses are introduced to mode @xmath6 after the lossless cross - kerr interaction has taken place . such losses are equivalent to mixing mode @xmath6 with the vacuum state on a beam splitter ( bs ) and discarding one bs output . a 50% loss corresponds to a 50/50 bs . it is known @xcite that for such a balanced beam splitter the wigner function of one bs output can be expressed as a scaled husimi q - function of the input state : @xmath136 the q - function of a state @xmath137 is defined as @xmath138 and is clearly non - negative for all coherent state amplitudes @xmath80 , i.e. , in the whole phase space of the mode . therefore negative regions of the wigner function can not survive losses larger than 50% if these occur during propagation of the generated state . hence losses that take place in the process of state generation via the cross - kerr interaction are less harmful to the non - classicality of the output state than the same level of loss after the interaction . as can be seen from the solution ( [ evolution_fock ] ) , an influence of dephasing into independent reservoirs can be profoundly destructive . dephasing leads to diminishing of the off - diagonal elements in the number - state basis with rates proportional to the difference of these numbers . so for the coherent state with the amplitude @xmath139 , a condition @xmath140 should be fulfilled for the interaction time @xmath141 to consider the influence of dephasing negligible . in the recently discussed qed schemes for generating non - linearity ( including eit - like ones ) , dephasing is usually disregarded without being estimated ( see , for example , @xcite ) . however , since the rate of losses increases with increasing intensity of the coherent states used in the generation process , more caution is required with respect to the dephasing . for schemes involving large cross - kerr nonlinearity in the solid - state structures ( such as , for example , photonic crystals ) emitter - mediated dephasing could be a major source of state decoherence and a reason for failure of schemes involving initial coherent states with large number of photons . however , remarkably , if the cross - kerr nonlinear interaction scheme is designed in such a way that losses due to dephasing are correlated , then it may be possible to avoid their destructive effects . this is the subject of the next section . it is well established , that the states of quantum systems can be correlated and even entangled through interaction with the common reservoir . this phenomenon can occur even in absence of any direct interaction between systems . in in appendix b3 we give an example of such a phenomenon for a scheme of generating the cross - kerr nonlinearity via dispersive interaction of modes with emitters . there a beam - splitting action of the common reservoir is considered , and it is shown how such a reservoir can produce a stationary entangled state of two modes . here we focus our attention on another important possibility : namely , on a way to neutralize a destructive influence of losses in the process of generation by rendering these losses correlated and exploit a correlating effect of the reservoir . let us consider an example of the realistic scheme to produce the large cross - kerr nonlinearity described in section [ lambda - system ] ( see eq . ( [ hamiltonian_lambda_system ] ) and the text thereafter ) . we consider the case of @xmath142 so the master equation ( [ master_lambda_model ] ) now transforms as @xmath143+{1\over2}\left[(\gamma_1-\gamma_{12 } ) \mathcal{l}(a_1)+(\gamma_2-\gamma_{12})\mathcal{l}(a_2)\right]\rho(t)+ { \gamma_{12}\over2}\mathcal{l}(a_1+a_2)\rho(t ) . \label{master_lambda_model1}\end{aligned}\ ] ] in absence of decoherence ( i.e. , @xmath144 ) , the scheme described by eq . ( [ master_lambda_model ] ) is able to generate entangled superpositions of schrdinger - cat states from initial coherent states of modes @xmath6 and @xmath7 ( for similar schemes see , for example , ) . it is easy to see that for @xmath145 the scheme produces an entangled superposition of coherent states from a pair of initially uncorrelated coherent states , which reads as @xmath146 in the presence of correlated reservoirs one can find a solution of eq . ( [ master_lambda_model1 ] ) noticing that under the rotation ( [ rotation ] ) the hamiltonian part of eq . ( [ master_lambda_model1 ] ) remains invariant . thus , performing the rotation one obtains @xmath147+{1\over2}\left[{\bar\gamma}_1 \mathcal{l}(b_1)+{\bar\gamma}_2\mathcal{l}(b_2)\right]\rho(t ) \label{master_lambda_model2}\end{aligned}\ ] ] with the exact solution given by eq . ( [ ham ] ) . also , one immediately sees that for the completely correlated reservoirs ( i.e. @xmath49 ) , the mode @xmath148 is not affected by the loss , as @xmath149 then . naturally , a cat state can be generated in this mode . thus , in the limit of large loss , @xmath150 , and for @xmath151 it follows from eq . ( [ master_lambda_model2 ] ) that @xmath152 where @xmath153 , @xmath154 . we also assume that @xmath155 . from eq . ( [ lambda_dissipation_exact ] ) , we derive a conclusion that the only effect of completely correlated loss is a reduction of amplitudes of the coherent states forming the superposition ( [ lambda_nodissipation_exact ] ) . thus , we have seen that by making losses completely correlated one can completely avoid decoherence caused by these losses . of course , in practice one can hardly have completely correlated reservoirs due to presence of additional uncorrelated loss ( such as modal losses due to coupling to additional reservoirs etc . ) . nevertheless , designing the scheme such as to have predominantly correlated losses might greatly enhance its robustness in production of non - classical states . we illustrate this with the simple example of the conditioned cat - state generation from the solution of eq . ( [ master_lambda_model ] ) . if the rotated mode @xmath156 impinges on the detector , in the case of no signal on the detector the rotated mode @xmath148 is ( up to the normalization factor ) in the state : @xmath157 , @xmath158 , and @xmath159 in fig . [ fig2 ] one can see examples of the wigner function of the state ( [ evolution_fock2 ] ) . for no loss ( fig . [ fig2](a ) ) and @xmath145 the state ( [ evolution_fock2 ] ) is a usual schrdinger cat state with the pronounced oscillations near the origin . large correlated loss ( @xmath160 ) changes the size of the cat and rotates it ( fig . [ fig2](b ) ) , but otherwise leaves it intact . lower correlated loss distorts the cat ( fig . [ fig2](d ) ) due to influence of additional mixing between modes in the interaction process ( as it follows from eq . ( [ evolution_fock2 ] ) ) . nevertheless , the state is strongly non - classical . uncorrelated loss with the same rate eliminates the non - classicality outright ( fig . [ fig2](c ) ) . correlated loss allows for the non - classicality to survive ( fig . [ fig2](f ) ) even if the uncorrelated loss with the rate equal to difference between individual rates and the correlation rate ( e.g. , @xmath161 ) destroys the non - classicality completely ( fig . [ fig2](e ) ) . in what presented here , we followed the quest to find ways to impair the decoherence processes in quantum state generation and manipulation . for the particular class of nonlinear interaction processes , we have found two striking examples of loss dynamics , for which the losses themselves counteract decoherence : loss - mediated correlations between the interacting modes and losses to the correlated reservoirs . the latter result in strongly correlated modal loss . these correlated losses influence the dynamics of the modes undergoing the cross - kerr nonlinear interaction in a completely different way than losses into independent reservoirs , the aspect of quantum nonlinear dynamics to large extent unexplored so far . thus , remarkably , designing the schemes for the generation of the kerr nonlinearity in such a way that losses in this process are correlated , one can greatly diminish their destructive impact and even exploit them for entanglement generation ( see also appendix b ) . as to the origin of this effect , if both modes , for example , interact with the same emitter transition , emitter losses are likely to lead to the correlated loss of both of these modes . note , that in this case a significant modal loss might occur even in the case when the emitter subject to losses stays in superposition of metastable levels with negligibly small probability to occupy higher , decaying levels . for more details and examples on the origin of the correlated losses and their entangling effect see appendix b. turning to the other aforementioned unexpected aspect of the quantum dynamics , we have demonstrated that losses in the nonlinear process of state generation affect the quantum state in quite a different way to the propagation losses of already generated nonclassical state . this is mainly due to the loss - mediated correlations between the modes participating in the cross - kerr interaction . in addition , losses through coupling to the correlated reservoirs ( correlated loss ) further enhance the difference between the decoherence processes _ during _ and _ after _ the state generation . in particular , non - classicality seems to be more robust with respect to the generation loss than to the propagation loss . we discussed an example of generation loss exceeding @xmath128 with negative values of the wigner function preserved , whereas the propagation loss exceeding @xmath128 renders the wigner function completely positive . the authors gratefully acknowledge funding from the eu project fp7 - 212008 compas ( n. k. and t.t . ) , equind project of 6fp ist-034368 , ffr of belarus project f08p-131 and fapesp , brazil , project 2008/57657 - 3 ( d.m . ) . also , d.m . would like to thank the school of physics and astronomy , university of st . andrews , for their hospitality . we start from the general effective hamiltonian @xmath0 of eq . ( [ ham_ger ] , [ vloss ] ) describing both self- and cross - interaction and interaction of modes with reservoirs responsible for losses ( see section [ master_section ] ) : @xmath162 setting @xmath17 for sake of simplicity , and using a time - convolutionless projection operator technique , one obtains the following master equation @xcite : @xmath163 - \\ & & \int\limits_0^{t}{\,{\rm d}}\tau \left\{\langle v_{\rm loss}(t)v_{\rm loss}(\tau)\rangle_{\rm r}\rho(t)+\rho(t)\langle v_{\rm loss}(\tau)v_{\rm loss}(t)\rangle_{\rm r } - \langle v_{\rm loss}(\tau)\rho(t)v_{\rm loss}(t)\rangle_{\rm r}- \langle v_{\rm loss}(t)\rho(t)v_{\rm loss}(\tau)\rangle_{\rm r}\right\}\qquad \label{master_gener}\end{aligned}\ ] ] where @xmath18 denotes the density matrix of the system averaged over states of all reservoirs ( and over possible stochastic variables , too ) , and @xmath164 denotes averaging over all reservoirs . reservoir modes are assumed to be initially in the vacuum states . further , we consider reservoirs of different types to be independent ; correlation functions @xmath165 , @xmath166 are taken to be zero . we consider the reservoirs of the same kind to be mutually correlated , i.e. , for markovian dephasing reservoirs we assume @xmath167 here @xmath168 is the delta - function and the rates @xmath169 , are real and non - negative . the cross - correlation parameter @xmath170 is taken to be real . for the photon loss reservoir , we assume that the term @xmath171 in the hamiltonian ( [ ham_ger ] ) preserves the total number of photons , i.e. only non - zero correlation functions are @xmath172 also here , for simplicity , the rates @xmath173 are assumed to be real and non - negative and the cross - correlation parameter @xmath174 to be real . for self - loss rates , @xmath175 , @xmath176 , and ` cross'-loss rates , @xmath174 , @xmath170 , the following relations hold @xmath177 under the assumptions made above , the master equation of eq . ( [ master_markov ] ) is obtained from eq . ( [ master_gener ] ) . we will illustrate the process of appearance of a _ correlated modal photon loss _ in a process of off - resonant interaction between a mode and an emitter with a simple example . consider two modes of the same frequency interacting off - resonantly with just a single two - level system ( tls ) , see fig . [ fig1](a ) . in the rotating - wave approximation and in the interaction picture with respect to the free hamiltonian of the reservoir , in the frame rotating with the tls transition frequency @xmath178 , one has the following hamiltonian describing the problem : @xmath179 here @xmath180 and @xmath181 is the mode frequency ; @xmath40 are interaction constants for the corresponding modes ; @xmath182 and @xmath183 are pauli operators for the tls , @xmath184 , @xmath185 , @xmath186 ; vectors @xmath56 , @xmath187 describe the lower and the upper tls levels , correspondingly . the reservoir operator @xmath188 describes the tls energy loss . we assume this reservoir to be markovian and the following relations hold : @xmath189 where brackets @xmath190 denote an averaging over the reservoir . here we are assuming that losses are weak ( i.e. , @xmath191 ) . we adopt the usual conditions for an adiabatic elimination of the emitter , i.e. the tls starts at the lower level , and the detuning @xmath192 between the mode frequency and the tls transition frequency is much larger than @xmath40 . thus , the tls upper level remains practically unpopulated . changing to the interaction picture with respect to the part of the hamiltonian ( [ jkmodel_ham_spon ] ) corresponding to the absence of the tls - field interaction , we get the following interaction hamiltonian @xmath193 with the bosonic annihilation operator for the collective mode @xmath194 and @xmath195 . a formal solution for @xmath196 without losses up to the third - order terms can be approximated as @xmath197x^{\dagger}(t_1 ) , \label{jkmodel_solution_sigma_spon}\end{aligned}\ ] ] where @xmath198 and @xmath199 , if one takes into account the fact that the modal dynamics is very slow on the scale of the tls dynamics and @xmath200 can be considered as practically constant . from eq . ( [ jkmodel_solution_sigma_spon ] ) , neglecting small and rapidly oscillating terms , after averaging over the atomic variables one arrives to the following effective interaction hamiltonian @xmath201 . \label{jkmodel_ham_v_app_spon}\end{aligned}\ ] ] deriving the master equation in the standard manner , one obtains an equation describing the correlated photon losses @xmath202+{\bar\gamma } \mathcal{l}(c)\rho(t ) , \label{master_jkmodel_spon}\end{aligned}\ ] ] where @xmath203 , @xmath204 and @xmath205 . effectively , eq . ( [ master_jkmodel_spon ] ) describe both coupling between modes and their interaction with the same reservoir . both these interaction might lead to the entanglement between modes . as seen in section v , even in the absence of direct intermodal coupling ( i.e. for @xmath206 , @xmath207 ) an interaction of modes with the reservoir entangles these modes . the analysis made above can be readily generalized to different schemes of cross - kerr nonlinearity generation through interaction of two modes with the same ensemble of emitters . in subsection [ lambda - system ] we devise the procedure for the scheme of the giant cross - kerr nonlinearity generation suggested in ref . @xcite . [ [ an - example - dispersive - two - mode - jaynes - cummings - model - with - dephasing ] ] an example : dispersive two - mode jaynes - cummings model with dephasing ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ here we illustrate an appearance of _ correlated modal dephasing _ with the example of the jaynes - cummings model ( fig . [ fig1 ] ) considered in the previous subsection . we model an influence of the dephasing reservoir as a stochastic fluctuation of the tls transition frequency . in the rotating - wave approximation the problem is described by the following hamiltonian : @xmath208\sigma_z+ \left[\sigma^+(g_1a_1+g_2a_2)+(g_1a_1^{\dagger}+g_2a_2^{\dagger})\sigma^-\right ] , \label{jkmodel_ham}\end{aligned}\ ] ] where @xmath209 is random process describing a small rapid stochastic modulation due to non - radiative interaction with surroundings . for simplicity we take @xmath209 to be just a white noise satisfying the following relations @xmath210 where @xmath211 denotes classical averaging . we consider the case of the weak loss , @xmath212 . as before , we assume that the conditions for adiabatic elimination of the emitter hold . in the interaction picture with respect to the part of the hamiltonian ( [ jkmodel_ham ] ) corresponding to the absence of the tls - field interaction , we have the following interaction hamiltonian @xmath213 a formal solution for @xmath196 in this interaction picture can be approximated as @xmath214c^{\dagger}(t_1)f^*(t_1 ) , \label{jkmodel_solution_sigma_def}\end{aligned}\ ] ] where @xmath215 averaging over dephasing noise and neglecting small terms , we get the following master equation using the standard technique implemented to derive eq . ( [ master_gener ] ) : @xmath218+{\bar d } \mathcal{l}(c^{\dagger}c)\rho(t ) , \label{master_jkmodel}\end{aligned}\ ] ] where @xmath219 the calculations of the coefficients in ( [ master_jkmodel_coefficients ] ) are carried out using the following property : @xmath220 and the fact that the detuning @xmath192 is assumed to be large , @xmath221 . so , one can see that dephasing of the atom leads to appearance of the correlated modal dephasing practically in the same manner as atomic population losses lead to the correlated modal loss considered in the previous subsection . also , modal dephasing occurs notwithstanding the fact that the upper atomic level remains practically unpopulated . finally , we demonstrate that the scheme described by the master equation ( [ master_jkmodel_spon ] ) can effectively produce entanglement between the modes . in fact , this scheme can act as a kind of lossy beam - splitter even in the absence of intermodal interaction in unitary part of eq . ( [ master_jkmodel_spon ] ) . indeed , let us consider a completely uncorrelated single - photon initial state @xmath222 , and the initial density matrix @xmath223 . in the zero- and single - photon subspaces one can assume the following orthonormal basis : @xmath224 is not affected by the losses described by eq . ( [ master_jkmodel_spon ] ) because @xmath225 . also , the following relations are satisfied : @xmath226 thus the system of equations for the density matrix elements can be obtained from eq . ( [ master_jkmodel_spon ] ) : @xmath227\rho_{+-}(t ) , \\ \rho_{-v}(t)=\rho_{-v}(0 ) , \quad { { \,{\rm d}}\over { \,{\rm d}}t}\rho_{+v}(t)=-[{\bar\gamma}+{{\rm i}}(\delta w+\chi)]\rho_{+v}(t ) . \label{spon_system}\end{aligned}\ ] ] the solution ( [ spon_system ] ) describes an emergence of entanglement form the initially uncorrelated state of both modes ( single photon in the mode @xmath228 and vacuum of the mode @xmath229 ) . figure [ fig5 ] depicts a measure of entanglement , a negativity as given in ref . @xcite @xmath230 where @xmath231 is the density matrix @xmath137 partially transposed with respect to the first mode . non - zero value of the negativity means that the state is the entangled one . it can be seen that the decay rate into the common reservoir does not affect the finally reached entanglement . this rate affect only time during which a stationary state is reached . it follows from this system of equations that the initial uncorrelated state @xmath232 under the action of the correlated modal loss asymptotically turns into @xmath233 the state ( [ spon_entangled_state ] ) is entangled for an arbitrary @xmath234 . however , the maximal degree of asymptotic entanglement is reached when @xmath235 , and with increasing of difference between @xmath236 and @xmath237 the asymptotic entanglement decreases ( figure [ fig5 ] ) . note that the state ( [ spon_entangled_state ] ) is influenced neither by the cross - kerr interaction of the modes nor by the linear excitation exchange , and one can set both @xmath238 . the same type of state is produced by a correlated dephasing described by eq . ( [ master_jkmodel ] ) . effectively , the correlation of reservoirs allows for existence of decoherence - free subspaces to which the two - mode state eventually evolves @xcite . entangling through the common reservoir with appearance of the long - living state similar to the one described by eq . ( [ spon_entangled_state ] ) might occur for emitters and collective reservoir modes near the band - edge in photonic crystals @xcite . d. k. armani , t. j. kippenberg , s. m. spillane and k. j. vahala , nature * 421 * , 925 ( 2003 ) ; t. j. kippenberg , s.m . spillane , and k. j. vahala , phys . rev . lett . * 93 * 083904 ( 2004 ) ; s. m. spillane , t. j. kippenberg , and k. j. vahala , k. w. goh , e. wilcut , and h. j. kimble , phys . rev . a*71 * 013817 ( 2005 ) .

real networks contain unreliable components ; in critical infrastructures and technological networks some links may become non - operational owing to disasters or accidents , and in natural networks this might occur owing to pathologies . although the robustness and resilience of networks have been extensively investigated over the past decade @xcite , controllability methods for complex networks that can robustly manage structural changes have not been developed . the existing research is limited to recent studies of network controllability under node attack and cascading failures using maximum matching @xcite , the discussion of quantitative measures of network robustness to investigate the effect of edge removal on the number of controllable nodes without a comprehensive theoretical analysis @xcite and studies on multi - agent systems under simultaneous failure of links and agents @xcite . note that the problem of how the number of driver nodes change as function of removal fraction of edges @xcite and our question of how to design complex networks with structurally robust control feature are drastically different . while the former are heavily relying on percolation and cascading failure techniques well - studied over the past decade , our work studies the minimum number of driver nodes to control the entire network against arbitrary single or multiple edge failures . here , we introduce the concept of structurally robust control of complex networks . to provide a concrete example , we adopt the minimum dominating set ( mds ) model because it has been widely applied to the control of engineering systems , such as mobile ad hoc networks ( manet ) , transportation routing , computer communication networks @xcite , design of swapped networks for constructing large parallel and distributed systems @xcite as well as the investigation of social influence propagation @xcite . robust control theory emerged in the late 1970s and is based on linear , time - invariant transfer functions . the controller is designed to change the system s model dynamics until it reaches a certain degree of uncertainty or disturbance . thus , the system is designed to be robust or stable against the presence of bounded modelling errors . to address disturbances , techniques such as single - input , single - output ( siso ) feedback and h - infinity loop - shaping were developed to avoid dynamic trajectories that deviate when disturbances enter the system @xcite . although modern robust control engineering has been successfully applied to numerous real - world problems , such as stability in aircrafts and satellites and the efficiency of power , manufacturing , and chemical plants , it has not yet been applied to the problem of structural changes on a large network . a set of nodes @xmath0 in a graph @xmath1 is a dominating set ( ds ) if every node in @xmath1 is either an element of @xmath0 or adjacent to an element of @xmath0 . then , the mds approach states that a network is made structurally controllable by selecting an mds ( driver set ) because each dominated node has its own control signal @xcite ( see fig . [ fig : schema]a ) . recently , molnr et al . further studied the size of an mds by exhaustively comparing several types of artificial scale - free networks using a greedy algorithm @xcite . interestingly , wuchty demonstrated the applicability of the mds approach @xcite to the controllability of protein interaction networks and showed that the minimum dominating set of proteins were enriched with essential , cancer - related and virus - targeted genes @xcite . whereas each element is controlled by at least one node in @xmath1 ( @xmath2=1 ) ( or is covered by itself ) in an mds , the novel robust mds ( rmds ) approach states that each node must be covered by itself or at least two nodes in @xmath1 ( @xmath2=2 ) ( see fig . [ fig : schema]cd ) . the analytical results and computer simulations demonstrate that a robust configuration ( @xmath2=2 , @xmath3=2 ) and non - robust configuration ( @xmath2=1 , @xmath3=1 ) of a scale - free network with minimum degree @xmath3 require the same order of driver nodes . the robust configuration guarantees that the system remains controllable even under arbitrary single or multiple link failure . this finding has remarkable implications for designing technical and natural systems that can still operate in the presence of unavailable or damaged links because the implementation of such a robust system in a large network does not change the order of the required controllers in a conventional system without robustness capability . as a byproduct of this research , our results also demonstrate that the minimum degree @xmath3 in a network plays an important role in network controllability and significantly affects the size of the mds . in particular , for @xmath4 , the order of the size of an mds changes if the minimum degree changes , unveiling another tool to reduce the number of driver nodes . these theoretical findings are confirmed by computer simulations and an analysis of real - world undirected , directed and bipartite networks . in addition , the mds approach is extended to address probabilistic network domination when we consider the probability of link transmission failure . the derived mathematical tools allow us to identify optimal controllability configurations in real biological systems , by mapping the synaptic unrealiability distribution experimentally observed in rat brains @xcite to the most well - known and recently updated neural network model for _ c. elegans _ organism @xcite . * the mds model and the robust domination configuration . * ( a ) the network is structurally controllable by selecting a minimum dominating set ( mds ) because each dominated node has its own control signal . a maximum matching approach needs three driver nodes @xmath5,@xmath6 , and @xmath7 , assuming a matching link from @xmath5 to @xmath8 . in contrast , the mds only requires one node , @xmath5 . the labels @xmath9 , @xmath10 and @xmath11 indicate control . ( b ) example of a dominating set ( ds ) in a graph @xmath1 . a set of nodes @xmath0 ( filled ) in a graph @xmath1 is a dominating set if every node in @xmath1 is either an element of @xmath0 or adjacent to an element of @xmath0 . ( c ) the graph @xmath1 allows for an mds with cover @xmath2=1 composed of only three nodes . ( d ) example of a robust minimum dominating set ( rmds ) ( i.e. , an mds with cover @xmath2=2 ) . in an rmds , each node in @xmath1 is either an element of @xmath0 or adjacent to at least two elements of @xmath0 . ( e ) each edge ( @xmath12 , @xmath13 ) has a failure probability of @xmath14 highlighted in the figure . in a probabilistic minimum dominating set ( pmds ) , the approach shown here , each node must be covered by multiple nodes in the mds so that the probability that at least one edge is active is at least @xmath15 . for instance , @xmath16 must cover itself because the link ( @xmath17 ) is unreliable with a failure probability of 0.8 , which is higher than @xmath15 . ] the concept of structural controllability was first introduced by lin @xcite for single - input systems and it was quickly extended to multi - input systems @xcite . the maximum matching ( mm ) algorithm identifies the minimum number of nodes to control the entire network , by providing a mapping between structural controllability and network structure @xcite . however , there are several striking differences with mds approach : ( 1 ) by using mm approach , the fraction of driver nodes tends to be minimized in random networks . the mds does not necessarily give a minimum number of driver nodes in the sense of mm approach . however , mds gives less number of driver nodes in many cases , including scale - free networks in which hubs are present @xcite . for example , consider the star graph ( all nodes but one node are leaves ) with @xmath18 leaves . then , mm approach needs @xmath18 - 1 driver nodes whereas mds approach needs only one driver node . ( 2 ) the mm approach is based on linear systems whereas mds approach does not even need structural controllability : it is enough to assume that a node is controllable if it is directly connected to a driver node . this represents one of the unique features of the mds model because it suggests that can be applied to a certain kind of nonlinear and/or discrete models . however , these striking advantages have a price . ( 1 ) the set of driver nodes in mds is @xmath19 in scale - free networks with @xmath20 . ( 2 ) each edge has to be controlled independently . however , even in the case of @xmath19 , a relatively small number of drivers is required in most cases . above all , the proposed concept of structurally robust control is an algorithmic - independent framework . therefore , engineering applications of structurally robust control may flexibly give preference to one algorithm over another . in the following , we present analytically derived predictions for the minimum number of drivers using the mds controllability approach by considering specific cases for the degree exponent @xmath22 and the minimum degree @xmath3 . then , the robust control is also analysed by considering the number of drivers @xmath2 required to cover each node . we first assume that the minimum degree of an undirected graph @xmath23 with @xmath18 nodes and a degree distribution that follows a power - law @xmath24 is @xmath3 . we then use a standard mean - field approach that assumes a continuum approximation for the degree @xmath25 , so that it becomes a continuous real variable @xcite . we also note that there are some discussions on degree cut - off @xcite because our analysis assumes that there exist high - degree nodes . however , we do not introduce such degree cut - off because we are performing a kind of mean - field analysis . it is to be noted that scale - free networks are a kind of random networks and thus we can have a node with even degree @xmath26 with very small probability if @xmath27 @xcite . in the mean - field analysis , such rare cases are taken into account . however , discussions of degree cut - off are based on average case analysis , and there does not exist a consensus cut - off value . therefore , we do not introduce degree cut - off in our analysis . the results of computer simulation support that our analysis is appropriate . note also that each node with degree more than 1 must be covered by @xmath2 nodes ( not by @xmath2 edges ) in our ilp formulation and thus the effect of multiple edges is eliminated in computer simulation . as it has been shown in the field of complex network science , the analysis and classification of networks in terms of their degree distribution is a key feature to understand the complex behavior of complex systems . in particular , the scale - free topology fundamentally changes the system s behavior , with broad implications from spreading processes on networks ( like for example the spread of infectious diseases ) to cascading failures @xcite . it is therefore appropriate to examine the controllability problem in networks governed by power - law degree distributions . first we assume that the minimum degree @xmath3 is 2 in an undirected graph @xmath23 , where @xmath28 is a set of @xmath18 nodes , and @xmath29 is a set of edges connecting nodes in @xmath28 . from the following equation : @xmath30 we have @xmath31 . let @xmath32 be the set of nodes with degree between @xmath33 and @xmath18 . then , the number of nodes in @xmath32 ( denoted by @xmath34 ) is @xmath35_{n^{\beta}}^{n } = - 2^{\gamma-1 } n \cdot \left ( n^{1-\gamma } - n^{\beta(1-\gamma ) } \right ) \approx 2^{\gamma-1 } n^{1+\beta(1-\gamma)}.\ ] ] let @xmath36 be the number of edges in @xmath23 . then , @xmath36 is given by @xmath37 where the factor 2 in @xmath38 comes from the fact that each edge is counted by two nodes . the number of edges that are connected to at least one node in @xmath32 ( i.e. , the number of edges covered by @xmath32 ) is lower bounded by @xmath39 it should be noted that @xmath40 gives a lower bound and the number of edges covered by @xmath32 may be much larger because this estimate considers the case where both endpoints of these edges are in @xmath32 . the probability that an arbitrary edge is not covered by @xmath32 is upper bounded by @xmath41 then , the probability that a node with degree @xmath25 does not have any edge connected to @xmath32 is upper bounded by @xmath42 which is also upper bounded by @xmath43 because the minimum degree is assumed to be 2 . therefore , the number of nodes ( denoted by @xmath44 ) not covered by @xmath32 is @xmath45 since we can have a dominating set if we merge these nodes with @xmath32 , the number of nodes in an mds is upper bounded by @xmath46 . to minimize the order of @xmath47 . we let @xmath48 which results in @xmath49 by using this @xmath50 , an upper bound of the size of an mds is estimated as @xmath51 we can see that this order is smaller than that of our previous result on @xmath3=1 @xcite @xmath52 in particular , the above takes the minimum order @xmath53 when @xmath54 for @xmath3=1 whereas the new bound for @xmath3=2 takes the minimum order @xmath55 when @xmath56 ( see fig . this difference comes from the fact that a node @xmath57 is regarded as not covered by @xmath32 if one specific edge connected to @xmath57 is not covered by @xmath32 in an existing analysis @xcite whereas a node @xmath57 is regarded as not covered by @xmath32 if no edge connected to @xmath57 is not covered by @xmath32 in this analysis . we can extend the above result for the case where the minimum degree is @xmath3 , by replacing @xmath58 with @xmath59 then , we have @xmath60 by using this @xmath50 , an upper bound of the size of mds is estimated as @xmath61 this order of the mds size that scales as @xmath62 takes the minimum value when @xmath63 it is to be noted that although @xmath64 depends on @xmath3 , it does not affect the order of the mds size . the scaling exponent @xmath65 for the order of the mds size is shown as a function of the degree exponent @xmath22 in fig . this is our first main result and demonstrates that for scale - free networks with @xmath21 , the order of the mds size changes ( the exponent @xmath65 changes in functional form of @xmath62 ) when the minimum degree increases . the dependence of the degree exponent @xmath66 that minimises the mds size on the minimum degree @xmath3 is also shown in fig . 2c . the results demonstrate that a higher minimum degree makes it easier to control scale - free networks with @xmath21 ( fig . 2d ) . next , we show the results for the robust domination ( rmds ) ( fig . for an undirected graph @xmath23 and a positive integer @xmath2 , @xmath67 is called a @xmath2-robust dominating set if each node @xmath68 satisfies the following : either @xmath69 or @xmath57 is connected to @xmath2 or more nodes in @xmath70 . here , we provide an upper bound of the size of the minimum @xmath2-robust dominating set . note that an rds is a special case of a generalized dominating set @xcite , which has been studied in the context of the computational complexity . however , it has not been investigated from the perspective of complex networks . we consider the case of @xmath2-robust domination in which the minimum degree of @xmath23 is @xmath3 , where @xmath2 and @xmath3 are constants such that @xmath71 . as in section ii.a , let @xmath32 be the set of nodes with degree between @xmath33 and @xmath18 . then , the probability that a node is not covered by @xmath2 or more nodes in @xmath32 is bounded by @xmath72 where we do not include the factor of @xmath73 because we consider an upper bound . this number is further bounded by @xmath74 for sufficiently large @xmath18 . therefore , the number of nodes not covered by @xmath32 is @xmath75 where we let @xmath76 . it is to be noted that a constant factor is ignored here because we use @xmath77-notation . as before , by balancing the size of @xmath32 and the number of non - covered nodes , we have @xmath78 by using this @xmath50 , an upper bound of the size of rmds is estimated as @xmath79 this is our second and most important finding . this result suggests that the case of an rmds with minimum cover @xmath2 and minimum degree @xmath3 corresponds to the case of an mds with the minimum degree @xmath80 . for example , the case of an rmds with @xmath81 ( i.e. , the case where each node ( with a degree of at least 2 ) must be covered twice , and the minimum degree @xmath3 is 2 ) corresponds to the case of an mds with @xmath82 . in all theoretical analyses in sections ii.a and ii.b , we assume that multi - edges ( i.e. , multiple edges between the same pair of nodes ) are allowed because it is known that there does not exist a network strictly following a power - law distribution with @xmath83 if multi - edges are not allowed @xcite . however , even if multi - edges between the same pairs are replaced by single edges after generating a power - law network with multi - edges , the results should hold if the cover parameter @xmath2 is 1 because mds is only concerned with existence of an edge from each node not in mds to a node in mds . if @xmath84 , we need to consider the possibility that some of @xmath2 edges are connected to a node @xmath57 are multi - edges because such @xmath57 may not be dominated by @xmath2 nodes . we will show that such a factor can be ignored in many cases if we discuss the order of the size of mds . of course , the resulting network does not strictly follow a power - law distribution if multi - edges are replaced by single edges . however , because any network with @xmath4 can not strictly follow a power - law distribution , our assumption seems reasonable . here we note that eqs . ( 4)(5 ) are identical to eqs . ( 1)(2 ) if we replace @xmath29 by @xmath3 . it suggests that the case of rmds with the minimum cover @xmath2 and the minimum degree @xmath3 corresponds to the case of mds with the minimum degree @xmath80 . for example , the case of rmds with @xmath81 corresponds to the case of mds with @xmath82 . in the above , we implicitly assumed that all @xmath2 edges are connected to different nodes in @xmath32 . however , we need to consider the possibility that some of @xmath2 edges are connected to the same node in @xmath32 because we allow multi - edges in theoretical analyses . suppose that @xmath2 ( or more ) edges from @xmath57 are connected to @xmath32 . since the number of edges connected to a node of degree @xmath25 in @xmath32 is @xmath85 , there exist @xmath86 nodes of degree @xmath25 in @xmath32 , and there exist @xmath87 edges connected to @xmath32 , the probability that @xmath2 edges contain at least one common endpoint in @xmath32 is @xmath88 since there exist @xmath19 nodes covered by @xmath32 , the number of nodes not covered by @xmath2 different nodes would be @xmath89 if the exponent @xmath90 is smaller than that in eq . ( 6 ) , this factor does not affect the order of eq . ( 6 ) . for @xmath91 , it is true for @xmath92 . however , if @xmath93 , it is true only for @xmath94 ( i.e. , @xmath95 ) . therefore , we need to be careful if we consider the case of @xmath96 and @xmath93 . [ [ analysis - of - lower - bound ] ] analysis of lower bound + + + + + + + + + + + + + + + + + + + + + + + first we consider a lower bound . let @xmath3 be the minimum degree . from @xmath98 , we have @xmath99 . for @xmath100 , @xmath101 denotes the set of edges between @xmath0 and @xmath102 . ( i.e. , @xmath103 ) . here we assume without loss of generality that @xmath104 because we are only interested in cases where @xmath105 is small compared with @xmath18 . the following property is trivial @xmath106 let @xmath0 be the set of nodes whose degree is greater than or equal to @xmath107 . we estimate the size of @xmath101 as follows . @xmath108 if @xmath0 is a dominating set , the last term should be no less than @xmath109 . therefore , the following inequality should be satisfied : @xmath110 by solving this inequality , we have @xmath111^{1/(\gamma-2 ) } ~=~ \left [ 2 d^{\gamma-1 } \cdot \left({\frac { \gamma-1}{\gamma-2 } } \right ) \right]^{1/(\gamma-2)}.\end{aligned}\ ] ] then , the size of @xmath0 is estimated as @xmath112^{-{\frac { \gamma-1 } { \gamma-2 } } } \cdot n. \label{eq : lb}\end{aligned}\ ] ] we extend the above analysis to @xmath2-domination ( i.e. , each node must be covered by @xmath2 or more edges ) . in this case , ( [ eq : nodom ] ) should be replaced by @xmath113 then , ineq . ( [ eq : domcond ] ) is also replaced by @xmath114 finally , we have @xmath115^{-{\frac { \gamma-1 } { \gamma-2 } } } \cdot n ~=~ \left ( { \frac c { 2 d^{\gamma-1 } } } \right ) ^{\frac { \gamma-1}{\gamma-2 } } \cdot \left [ \left({\frac { \gamma-1}{\gamma-2 } } \right ) \right]^{-{\frac { \gamma-1 } { \gamma-2 } } } \cdot n . \label{eq : lbc}\end{aligned}\ ] ] for example , consider the case of @xmath116 and @xmath117 for fixed @xmath3 . in this case , 2-domination requires @xmath118 times larger mds . [ [ analysis - of - upper - bound ] ] analysis of upper bound + + + + + + + + + + + + + + + + + + + + + + + next we consider an upper bound . as in the above , we have @xmath99 . let @xmath32 be the set of nodes with degree between @xmath119 and @xmath18 . then , the size of @xmath32 , @xmath34 , is estimated as @xmath120_{b}^{n } = - n d^{\gamma-1 } \cdot \left ( n^{1-\gamma } - b^{1-\gamma } \right ) \approx n \cdot { \frac { d^{\gamma-1 } } { b^{\gamma-1}}}.\ ] ] as in section [ sec : gsmall ] , let @xmath36 and @xmath40 be the number of edges in @xmath23 and the number of edges connected to @xmath32 , respectively . in addition , let @xmath3 denote the minimum degree . then , we have @xmath121 the probability that an arbitrary edge is covered by @xmath32 is @xmath122 thus , a lower bound of the probability that an arbitrary node is not covered by @xmath2 or more edges is estimated as @xmath123^k \left [ 1- \left ( { \frac d b } \right)^{\gamma-2 } \right]^{d - k } ~=~ \sum_{k=0}^{c-1 } \left ( \begin{array}{l } d \\ k \end{array } \right ) \left [ \left ( { \frac d b } \right)^{\gamma-2 } \right]^k \left [ 1- \left ( { \frac d b } \right)^{\gamma-2 } \right]^{d - k } .\end{aligned}\ ] ] since it is very difficult to consider a general pair @xmath124 , we consider the case of @xmath95 . then , this probability is simplified into @xmath125 therefore , an upper bound @xmath126 of the size of mds is estimated as @xmath127 .\end{aligned}\ ] ] it is to be noted that this number does not give a meaningful bound for many @xmath128 . for example , if @xmath129 and @xmath130 , @xmath131 holds . by solving @xmath132 , we see that @xmath126 takes the minimum value @xmath133\end{aligned}\ ] ] at @xmath134 . it is interesting to note that this minimum value does not depend on the minimum degree @xmath3 . if @xmath135 ( i.e. , original mds ) , this minimum value is simplified into @xmath136.\end{aligned}\ ] ] it is also interesting to consider the case of @xmath129 and @xmath137 . in this case , @xmath126 is given by @xmath138 ^ 2 \right\ } .\end{aligned}\ ] ] although it is difficult to analytically derive its minimum , we can estimate it by numerical computation . 3 compares upper bounds for @xmath139 and @xmath140 and lower bounds for @xmath139 . this figure shows that the upper bound becomes smaller as @xmath3 increases in the case of @xmath129 . and lower bounds for @xmath139.,width=453 ] although , as shown in fig . 4 , the gap between the derived lower and upper bounds is large ( especially for larger @xmath2 and @xmath3 ) , these are not trivial . for example , suppose that there exist two nodes with degree no less than @xmath26 . then , the size of rmds for @xmath116 is 2 , which is @xmath141 ( much less than @xmath19 ) . the @xmath142 lower bound suggests that such a case seldom occurs in random scale - free networks . let @xmath23 be an undirected graph , where @xmath28 and @xmath29 are sets of nodes and edges , respectively . we begin with ilp ( integer linear programming)-formulation for computation of an mds @xcite . from @xmath23 , we construct the following ilp instance : @xmath143 then , the set @xmath144 clearly gives an mds . it is known that , in contrast to the bipartite matching @xcite , the mds problem is np - hard . therefore , it is reasonable to use ilp . suppose that each node must be covered twice except degree 1 and 0 nodes . then , we can formulate this _ robust dominating set _ problem for @xmath116 ( i.e. , each node ( with degree greater than 2 ) is either in mds or is covered by at least two nodes in mds , where each node with degree 1 is either in mds or is covered by at least one node in mds ) as follows . because it is impossible to cover each degree 1 node by two edges , we have introduced this exceptional handling of degree 1 nodes . however , if the minimum degree is 2 or more , we need not consider this exceptional case . @xmath145 however , * glpsol * ( gnu linear programming solver ) executable could not solve this problem in reasonable cpu time . so , we strengthen the condition so that each node with degree greater than 2 is covered by at least two nodes in mds even if the node belongs to mds . then , the resulting ip becomes as follows . @xmath146 it is to be noted that the solution obtained by the above ilp also satisfies the conditions of the original formulation . therefore , the solution obtained by this ilp also gives a robust dominating set although it is not necessarily minimum . we can also consider a variant of mds in which weight @xmath147 is assigned for each edge and each node @xmath148 must be covered by edges with total weight @xmath149 . then , this variant can be formulated as @xmath150 for the mds ( @xmath2=1 ) and rmds ( @xmath2=2 ) configurations computed in real - world and simulated networks , the optimal solution was calculated using glpsol solver ( http://www.gnu.org/software/glpk ) . the gnu linear programming kit ( glpk ) supplies a software package intended for solving large - scale linear programming ( lp ) , mixed integer programming ( mip ) , and other related problems . in our problem , after translating the mathematical problem into an ilp problem , the input model is solved using gnu linear programming solver ( glpsol ) executable . for the probabilistic mds ( pmds ) , to be shown later , the optimal solution for the ilp - formulation was calculated using the ibm ilog cplex optimizer studio ver.12.02 . as the glpk , it is a software package that allows to solve large - scale mathematical optimization problems . the computation of the pmds is more intensive than that of mds and rmds , therefore we used cplex because it performed faster than glpk to find the optimal solution . we employ the following algorithm to construct unipartite scale - free networks of size @xmath18 , in which the degree distribution of @xmath28 ( a set of nodes ) follows @xmath151 under the constraint that the minimum and maximum degrees are @xmath3 and @xmath18 , respectively . for given @xmath152 we generate a random unipartite network in the following way . * for each node @xmath68 , generate half edges @xmath153 ( @xmath154 is a virtual node ) according to the degree distribution @xmath155 under the constraint of the minimum degree @xmath3 and the maximum degree @xmath18 , where @xmath64 is selected so that the number of nodes in @xmath28 is almost @xmath18 . * repeat the following until there are almost no remaining half edges : randomly select non - connected @xmath156 and @xmath157 such that @xmath158 and then connect @xmath57 and @xmath159 . the probabilistic mds , to be introduced later , was computed using generated samples of synthetic scale - free networks with a variety of scaling exponent @xmath22 and average degree @xmath160 values using the havel - hakimi algorithm with random ( monte - carlo ) edge swaps ( hmc ) @xcite . to confirm the theoretical predictions shown above , we constructed artificial scale - free networks with a variety of degree exponents @xmath22 and minimum degree @xmath3=1 and @xmath3=2 . an ensemble of scale - free networks was constructed for each network size up to 10,000 nodes , and the mean value together with standard error of the mean ( s.e.m . ) for mds size with @xmath2=1 and @xmath2=2 were computed . for @xmath21 , the theoretical results predict the same order of mds size ( the same exponent @xmath65 in the scaling function @xmath62 ) for configurations ( @xmath3=2 , @xmath2=2 ) and ( @xmath3=1 , @xmath2=1 ) ( see eq . in contrast , the results predict a different scaling for the configuration ( @xmath3=2 , @xmath2=1 ) , as shown by eq . 3 . fig . [ fig : combined_sim ] presents the simulation results for @xmath21 , which agree with the analytical predictions . for @xmath161 , the analytical computations predict the same scaling functional form @xmath62 with @xmath162 for all the three configurations . the simulation results agree with this prediction with high accuracy ( see figs . 6 and 7 ) . we used the concepts and mathematical tools presented above to investigate the robust control of several real networks . the experimental data analysis includes undirected , directed and bipartite networks from biological and socio - technical systems ( see tables 1 , 2 and 3 ) . we first present the results for undirected networks and show that the mds density for @xmath2=1 increases with increasing @xmath22 . the computation of the robust mds density ( @xmath2=2 ) exhibits a similar dependency , as predicted by eq . interestingly , the mds ratio for @xmath2=2 and @xmath2=1 differs , on average , by a factor of 2 or less ( fig . 9d ) , in agreement with the theoretical predictions shown in fig . 4c . when overlapping the real data and the predictions from eqs . 9 and 10 for the lower and upper bounds , respectively , for networks with @xmath161 , we see that the real data are always within the theoretical boundaries ( fig . 10 ) . the mds size for both @xmath2=1 and @xmath2=2 scales linearly with @xmath18 ( see fig . [ fig:2x3]a ) , which is in agreement with the theoretical predictions shown in eqs . 9 and 10 for @xmath161 and the computer simulations ( see figs . note that fig . 8 displays the mds fraction , and fig . [ fig:2x3]a represents the mds size . to investigate the influence of the frequency of nodes with degree 1 and 2 ( @xmath163 and @xmath164 ) on robust control , we computed mds1 and mds2 versus @xmath165 and @xmath166 . we then calculated the ratio of @xmath167 versus @xmath165 and @xmath166 . the results indicate that a small @xmath165 and large @xmath166 tend to be associated with a small mds density ( see fig . the ratio of mds2/mds1 is less than 2 in most cases . the analysis for directed networks included an internet peer - to - peer ( p2p ) network , the transcriptional regulatory network for _ e. coli _ from the regulon database , a set of food webs from different ecosystems , u.s . political blogs and the chemical synapse network for _ c. elegans_. the results demonstrate that mds1 and mds2 densities increase with increasing @xmath168 ( figs . 12a - b ) and @xmath169 ( fig . 13a - b ) , which is in agreement with the dependence found for undirected networks . in addition , the mds sizes for @xmath2=1 and @xmath2=2 scale linearly with @xmath18 ( fig . moreover , as in the undirected case , the mds ratio between @xmath2=1 and @xmath2=2 is almost always less than two , with only one exception ( see fig . moreover , less than 50@xmath170 of nodes are needed to control the network in both the typical ( @xmath2=1 ) and robust ( @xmath2=2 ) control configurations ( see figs . 12c - d and 13c - d ) . we define a bipartite graph as @xmath171 , where @xmath172 is a set of top nodes , @xmath173 is a set of bottom nodes , and @xmath29 is a set of edges ( @xmath174 ) . in our analysis , the directions of the edges are considered from @xmath172 to @xmath173 . therefore , the set of driver nodes will be a subset of @xmath172 , where nodes in @xmath172 need not be covered . the computation of an mds of a bipartite network is equivalent to the computation of a minimum set cover . although it is an np - hard problem , we have verified that the optimal solution is obtained in networks with power - law distributions of up to approximately 110,000 nodes within a few seconds . the computation was formalised as the following integer linear programming ( ilp ) problem @xmath175 the above mentioned ilp can be extended for computation of an rmds in bipartite networks . it is formalized as @xmath176 where @xmath177 indicates the degree of node @xmath148 . it should be noted that for any node @xmath178 with degree 1 , it is not possible to cover @xmath148 twice and thus we must relax the condition for these nodes . we employ the following algorithm to construct bipartite scale - free networks , in which the degree distributions of @xmath172 and @xmath173 follow @xmath179 and @xmath180 , respectively . here , we consider @xmath181 and @xmath182 . the maximum degree for the nodes in @xmath172 corresponds to @xmath183 . then , for given @xmath183 , @xmath184 , @xmath185 we generate a random bipartite network in the following way . * for each node @xmath186 , generate half edges @xmath156 ( @xmath154 is a virtual node ) according to the degree distribution @xmath187 where @xmath188 is selected so that the number of nodes in @xmath172 is almost @xmath183 . * for each node @xmath189 , generate half edges @xmath190 ( @xmath191 is a virtual node ) according to the degree distribution @xmath192 where @xmath193 is selected so that the number of @xmath194s is equal to the number of @xmath195s . * randomly connect @xmath196s and @xmath194s in a one - to - one manner . it is to be noted that @xmath197 ( the number of nodes of @xmath173 ) is determined automatically in step 2 to satisfy the condition on edge numbers . we collected a set of 10 real - world bipartite networks corresponding to socio - technical ( fig . 14a ) and biological systems ( fig . 14b ) . we then formalised and computed the mds for the @xmath2=1 and @xmath2=2 configurations . although the mds with @xmath2=2 is always larger than the mds with @xmath2=1 , the difference is proportionally very small in most cases . 14ab also illustrates that biological systems tend to require a larger mds size than socio - technical systems . the computer simulation of ensembles of bipartite scale - free networks with a variety of degree exponents also demonstrates that @xmath2=1 , @xmath3=1 and @xmath2=2 , @xmath3=2 can control the network with a similar fraction of nodes . therefore , robustly controlling a bipartite network requires a similar fraction of nodes as the typical , non - robust system ( see figs . in some real networks , each link has a probability of failing , which leads to the probabilistic concept of robust control ( pmds ) . for example , experimental analyses on neural networks have confirmed the unreliability of central synaptic transmission in rat brains @xcite . the mean transmission failure probability was found to be @xmath198=0.71 , with a range of 0.3 to 0.95 ( @xmath199=0.24 ) . in this work , we used the well - studied _ c. elegans _ neural network to investigate probabilistic robust control . to investigate this type of systems from a theoretical perspective using the robust mds approach , we assume that each edge @xmath200 has the probability of failure @xmath201 ( see fig . [ fig : schema]e ) . we require that each node is covered by multiple nodes in an mds so that the probability that at least one edge is active is at least @xmath202 . let @xmath0 be a ds then , @xmath0 must satisfy @xmath203 as we will show later , this problem can be also formalized and solved using integer linear programming ( ilp ) . first , we consider the case of @xmath82 ( i.e. , the minimum degree is 1 ) and @xmath204 . let @xmath32 be an mds for @xmath23 for the non - probabilistic version . let @xmath205 be a set of degree 1 nodes each of which does not belong to @xmath32 but is dominated by a node in @xmath32 . let @xmath206 be the only edge connecting to @xmath207 . we can observe : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ if @xmath208 , @xmath57 must be covered by itself . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ therefore , all nodes in @xmath205 should be added to @xmath32 ( in a probabilistic version ) when @xmath209 . therefore , it is expected that the mds size increases approximately from @xmath210 to @xmath211 at around @xmath212 . for example , consider the case of @xmath213 . then , there should be great increase of the mds size at @xmath214 . it shows good agreement with the simulation result ( see figs . 16 - 17 ) . we can extend the above analysis to the case of @xmath137 ( i.e. , the minimum degree is 2 ) and @xmath204 . let @xmath32 be an mds for @xmath23 for the non - probabilistic version . in this case , we consider two types of nodes of degree 2 : * @xmath57 has one edge connecting to a node in @xmath32 , * @xmath57 has two edges connecting to nodes in @xmath32 , where each node does not belong to @xmath32 but is dominated by a node in @xmath32 . let @xmath215 and @xmath216 be the sets of type ( a ) and type ( b ) nodes , respectively . then , nodes in @xmath215 should be added to @xmath32 if @xmath217 . on the other hand , nodes in @xmath216 should be added to @xmath32 if @xmath218 . therefore , it is expected that the mds size increases approximately from @xmath210 to @xmath219 at around @xmath212 and from @xmath219 to @xmath220 at around @xmath221 . in the case of @xmath213 , these two threshold values are @xmath222 and @xmath223 , in good agreement with the simulation result . next , we consider the case of @xmath82 and @xmath224 . as in the above , let @xmath32 be an mds for @xmath23 for the non - probabilistic version , and let @xmath205 be a set of degree 1 nodes each of which does not belong to @xmath32 but is dominated by a node in @xmath32 . let @xmath225 be the only edge connecting to @xmath207 . let @xmath226 be the failure probability of this edge @xmath227 , where @xmath228 . we can observe : _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ if @xmath229 , @xmath57 must be covered by itself . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ we define @xmath230 by @xmath231 therefore , all nodes in @xmath230 should be added to a ds ( in a probabilistic version ) where @xmath232 . here , the size of @xmath230 is estimated as latexmath:[\ ] ] again , the fraction is given by @xmath255 . the simulation results with 1,000 nodes for @xmath3=1 , @xmath3=2 with @xmath244 configurations are shown in figs . we also compared the theoretical results for the case @xmath137 with those from the simulations performed on scale - free networks with @xmath137 and @xmath130 . the plots show a similar overall tendendy although the inflection point is more noticiable in the theoretical curve ( see fig . it is worth noticing that the theoretical values @xmath256 and @xmath257 are scaled so that these take almost the same values as the simulated ones at the beginning and ending points ( i.e. , so that the values take between 0.3 and 0.85 instead of between 0.0 and 1.0 ) since it is assumed in theoretical analysis that all nodes are of degree 2 and the effects of the other nodes are ignored ( note also that degree 2 nodes occupy a major portion of nodes in the case of @xmath137 ) . this comparison result suggests that theoretical analysis captures some tendency even if nodes with degree more than 2 are ignored . we assume that each edge @xmath200 has the probability of failure @xmath201 . we want each node be covered by multiple nodes in mds so that the probability that at least one edge is active is at least @xmath202 . let @xmath0 be a dominating set . then , we require @xmath0 to satisfy @xmath258 then , we have @xmath259 then , we have @xmath260 where @xmath177 indicates the degree of node @xmath148 . recent reconstructions of the _ c. elegans _ neural network have significantly updated the wiring diagram of the somatic nervous system . the new reconstruction includes original data from white et al . @xcite , hall and russel @xcite and adds new information . in particular , 3,000 synaptic contacts , including gap junctions , chemical synapses and neuromuscular juctions were updated or added to the latest network version @xcite . as as a result , the large - scale structure of the network has significantly changed with respect to that of white et al . here , we focus on the connectivity of gap junction and chemical synapse networks of _ c. elegans _ neurons . the channels that provide electrical coupling between neurons are called gap junctions . in contrast , chemical synapses use neurotransmitters to link neurons . because these network are biologically different , they are treated independently , as done in @xcite . although it might be possible that gap junctions could conduct current in only one direction , this feature has not been observed or confirmed yet in _ c. elegans therefore , this network was considered as undirected network . the chemical synapses , in contrast , contains directionality capability , a feature that has been confirmed using micrographs @xcite . the analyzed gap junction network consisted of 279 neurons and 514 gap junction connections . the giant connected component is composed of 248 neurons and two smaller components of 2 and 3 neurons . after removing the 26 isolated neurons , we performed our analysis using 253 neurons and 514 connections . the statistical analysis revealed a power - law distribution for the degree distribution with a characteristic degree exponent of @xmath22=3.14 @xcite . the chemical synapse network consisted of 279 neurons and 2,194 directed connections . the statistical analysis showed that the in - degree ( out - degree ) distribution followed a power - law with degree exponent @xmath169=3.17 ( @xmath168=4.22 ) , respectively . these results contrast with analyses done using the dataset from white et al . @xcite , which reported an exponential decay for the degree distribution @xcite . experimental analyses on neural networks have confirmed the unreliability of central synaptic transmission in rat brains @xcite . the mean transmission failure probability was found to be @xmath198=0.71 , with a range from 0.3 to 0.95 ( @xmath199=0.24 ) . in this work , we used the most well - studied neural network corresponding to the _ c. elegans _ ( chemical synapse and gap junction ) to investigate probabilistic robust control . a visual representation of experimental neural gap junction ( undirected ) for _ c. elegans _ is shown in fig . 19 . a transmission failure probability distribution similar to that observed in rat brains was mapped on the links of these networks , making a fraction of them unreliable . the results of the analyses are described in figs . 16 - 17 for computer simulations and figs . 20 - 22 for real neural gap junction and chemical synapse networks and suggest that the presence of variance of the failure probability @xmath199 does not significantly affect the fraction of driver nodes . in contrast , it is strongly affected by both the minimum degree @xmath3 and the average failure probability @xmath198 . this biological example of unreliable links suggests that theoretical results and simulations on probabilistic robust control analysis may have an impact on understanding and controlling at will real - world systems with unreliable components . we have introduced the concept of structurally robust control of complex networks and have used the mds model , which is widely applied in engineering problems , to illustrate an example of robust complex network controllability . counterintuitively , the developed analytical tools , computer simulations and real - world network analyses demonstrate that robust control in a large network does not change the order of required driver nodes compared to a conventional system without such robust capability . when using an mds with @xmath2=1 , @xmath3=1 , the system can easily become uncontrollable if only one power or communication line fails during major natural disasters . in contrast , in the rmds framework ( @xmath2=2 , @xmath3=2 ) the system remains controllable even under arbitrary single or multiple link failure . therefore , both configurations require exactly the same order of controllers . engineering and biological systems could benefit from these findings . in addition , the order of the mds changes for @xmath4 by changing the minimum degree @xmath3 ( e.g. , constructing real networks with degree @xmath261 ) , unveiling another tool to decrease the number of driver nodes . because some real networks have unreliable links , we have extended our framework to probabilistic robust control ( pmds ) and have successfully applied the developed analytical tools to real neural networks of _ c. elegans _ with unreliable synaptic transmission . with the forthcoming comprehensive map of neural connections in the human brain @xcite , the presented method could offer new avenues to examine the brain s large - scale structure , to address synaptic reliability and to stimulate large fractions of the brain by interacting only with relatively few components . the proposed concept of structurally robust control of complex networks could also be investigated using a different algorithmic framework . as discussed above , we selected the mds model because it has already found applications in real engineering systems . however , the concept could also be mathematically formalised and implemented using , for example , the maximum matching model @xcite . in this case , additional computations would be needed to investigate the order of drivers in an optimal robust control configuration ; therefore , this analysis is left for future work . in addition , the presented method can also address the simultaneous failure of multiple links . the aim of the rmds ( @xmath2=2 ) framework is to construct a system that remains controllable even if an arbitrary link is damaged . however , the developed analytical tools also allow us to design a system with a more robust configuration ( @xmath2=3 ) or ( @xmath2=4 ) so that the network is still controllable even in case of arbitrary failure of pair or triplet of links , respectively . the emerging picture for probabilistic failure or malfunction of transportation and transmission lines in real - world complex infrastructures , socio - technical networks and biological networks emphasizes the importance and role of the presented robust ds approach for controllability . the proposed framework and tools offer a new direction for understanding the linkage between controllability and robustness in complex networks , with implications from engineering to biological systems . was partially supported by mext , japan ( grant - in - aid 25330351 ) and t.a . was partially supported by mext , japan ( grant - in - aid 26540125 ) . this work was also partially supported by research collaboration projects by institute for chemical research , kyoto university . = 1 and @xmath2=2 as shown in legend . note that social networks tend to be controlled with a smaller fraction of nodes compared to biological networks . ( c - d ) computation of the mds in synthetic bipartite networks generated with the model shown in section iv.c with several values of degree exponents @xmath184 and @xmath185 and three configurations ( @xmath2=1 , @xmath3=1 ) , ( @xmath2=2 , @xmath3=1 ) and ( @xmath2=2 , @xmath3=2 ) . note that the minimum degree @xmath137 almost completely compensates the increasing of the mds size from the robust control @xmath2=2.,width=604 ] = 2 and @xmath244 . * in order to estimate the fraction of degree-2 nodes to be added to a ds , it is enough to consider the regions ( a ) and ( b ) for the cases of @xmath248 and @xmath253 , respectively.,width=529 ] .*the real undirected networks analysed in this work . * type , name and description of each undirected network . we used the discrete maximum - likelihood fitting method to estimate the degree exponent @xmath22 from the cumulative degree distribution of each network @xcite . the standard error of @xmath22 is derived from the width of the maximum likelihood . the same method was used to estimate the degree exponent in the directed and bipartite networks shown in tables 2 and 3 . [ cols="<,<,<,<",options="header " , ]

molecular dynamics ( md ) simulations of macromolecules such as proteins and polymers have become increasingly common as forcefields have improved and computers have become more powerful . of particular interest is the integration of md simulations with nuclear overhauser effect spectroscopy ( noesy ) , and variants thereof , which directly probe the nuclear dipole - dipole interactions , and which have proven to yield unique information about the structure and dynamics of macromolecules @xcite . an important approach in the interpretation of non - rigid macromolecules has been the separation of the fast internal dynamics which are accessible by md simulations , from the much slower dynamics related to rotation of the entire macromolecule . a popular technique for this separation has been the model - free lipari - szabo approach @xcite , which yields important information about the structure and dynamics of the non - rigid branches and internal motions of the macromolecules . the integration of md simulations and nmr measurements presented here also requires state of the art in both computation and measurement . on the measurement side , such studies require careful de - oxygenation of the fluid for accurate measurements of the intrinsic nmr relaxation times @xcite , as well as diffusion experiments free of convection and hardware artifacts @xcite . on the computational side , with current empirical potential ( forcefield ) models it is now possible to readily access @xmath10 ns time scales that are of interest in understanding relaxation processes in confined fluids . but care is needed in ensuring the potential models are reasonable and the simulations are well - crafted . one emerging area of interest for integrating md simulations with nmr measurement is in porous organic media such as kerogen , where the phase - behavior and transport properties of hydrocarbons and water are greatly affected by confinement in the organic nanometer - sized pores . studies of fluid confinement in organic nano - pores is of great interest in characterizing organic - shale reservoirs such as gas - shale and tight - oil shale @xcite . for example , md simulations are well suited for investigating the effects of wettability alterations in organic matter @xcite , or the effects of restricted diffusion and pore connectivity of organic nano - pores . recent md simulations of fluids confined in graphitic nano - pores have shown promising results for molecular interactions between surfaces and fluids such as hydrocarbons and water @xcite . recent nmr measurements of hydrocarbons and water confined in the organic nano - pores of isolated kerogen indicate that @xmath0h@xmath0h dipole - dipole interactions play a key role in surface relaxivity , pore - size analysis , and fluid typing @xcite . however , to the best of the authors knowledge , no comprehensive studies have been performed which integrate md simulations with nmr measurements of hydrocarbons and water , either in the bulk phase or under confinement in porous media , even though the techniques exist to generate molecular models of kerogen for md simulations @xcite . another emerging area of interest is in using md simulations to characterize the nmr relaxation in complex multi - component fluids such as heavy crude - oils @xcite , or model systems for crude - oils such as polymer - solvent mixes @xcite . in such cases the lipari - szabo model is a good approach for separating the slow dynamics related to viscosity from the fast dynamics related to local non - rigid molecular motions . the objective in such cases is to obtain a more fundamental understanding of the nmr relaxation and diffusion of crude - oils , thereby helping to characterize petroleum fluids . the first step in embarking in such studies is to successfully integrate md simulations and nmr measurements of relaxation and diffusion for bulk hydrocarbons and bulk water . this report presents such a task , and validates the forcefields used in md simulations for bulk hydrocarbons and water , without any adjustable parameters in the interpretation of the simulation data . the md simulations are then used to distinguish intramolecular from intermolecular nmr relaxation . our studies provide new insight into the two relaxation mechanisms , which is not straightforward to do from nmr measurements alone @xcite . a comprehensive investigation into the two relaxation mechanisms is required to improve the characterization of macromolecules such as proteins and polymers , where typically only the intramolecular interaction is considered @xcite , without any substantial justification . the rest of the article is organized as follows . the methodology behind the md simulations of nmr relaxation and diffusion are presented in section [ sc : method ] . this is followed by results and discussions in section [ sc : results ] , including correlation between simulations and measurements , intramolecular versus intermolecular interactions , and , analysis using traditional hard - sphere models . conclusions are presented in section [ sc : conc ] . this section presents details of the molecular simulation procedure , the nmr relaxation analysis , and the diffusion analysis . the molecular simulations were performed using namd @xcite version 2.11 . the bulk alkanes were modeled using the charmm general force field ( cgenff ) @xcite and bulk water was modeled using the tip4p/2005 model @xcite . in the case of water , the relaxation behavior ( see below ) using the more refined amoeba polarizable model @xcite was in excellent agreement with predictions based on tip4p/2005 ( data not shown ) . given the typically greater challenge in modeling water , we thus expect the relaxation behavior in alkanes modeled with a polarizable model and the modern cgenff non - polarizable model to be comparable . thus for computational economy , in this study we used non - polarizable models throughout . ( the good agreement between simulations and experiments supports this suggestion . ) the simulation cells for the hydrocarbons were constructed as follows . we first construct a single molecule using the avogadro program @xcite . we than create @xmath11 copies of this molecule and pack it into a cube of volume @xmath12 using the packmol program @xcite . the volume is chosen such that the number density @xmath13 corresponds to the experimentally determined number density at 293.15 k ( 20@xmath14c ) . all the subsequent simulations are performed at this number density and temperature due to the availability of density @xcite and viscosity @xcite data at that temperature . this initial system was energy minimized to relieve any energetic strain due to limitations in the packing procedure . for the water simulation , we used an existing equilibrated simulation cell for the present study . the alkane simulations typically comprised between 1660 ( @xmath1-c@xmath2h@xmath3 ) and 2125 ( @xmath1-c@xmath4h@xmath5 ) carbon atoms , while the water system comprised 512 water molecules . for the molecular dynamics simulations , the equations of motion were propagated using the verlet algorithm with a time step of 1.0 fs . ( the shake @xcite procedure was used to constrain the geometry of water molecules , but the c h bond in the alkane was not similarly constrained . ) we first equilibrate the simulation cell for 2 ns at a temperature of 293.15 k. the temperature was controlled by reassigning velocities ( obtained from a maxwell - boltzmann distribution ) every 250 fs . the lennard - jones interactions were terminated at 14.00 ( 11 for water ) by smoothly switching to zero starting at 13.00 ( 10 for water ) . electrostatic interactions were treated with the particle mesh ewald method with a grid spacing of 0.5 ; the real - space contributions to the electrostatic interaction were cutoff at 14.00 . for water the equilibration was for 0.25 ns . we started the production run of 2 ns time from the end - point of the equilibration run . since the dynamics of relaxation are of first interest here , the production phase was performed at constant density and energy ( nve ensemble ) , i.e. the system was not thermostated . the average temperature of the system is within 2 k of the target ( 293.15 k ) for all the alkanes except @xmath1-c@xmath15h@xmath3 ( pentane ) for which the deviation was about 3 k. as a further check of the simulation , we also monitored the average energy of the system . all the simulations showed excellent energy conservation , with the standard error of the mean energy relative to the mean energy being less than @xmath16 . in the production phase , configurations were saved every 0.1 ps for the hydrocarbons and 0.125 ps for water . the theory of nmr relaxation in liquids is very well established @xcite , and for brevity we consider only the essential details . the autocorrelation function @xmath17 of fluctuating magnetic dipole - dipole interactions is central to the development of nmr relaxation theory in liquids . this autocorrelation is also well suited for computation using md simulations @xcite . using the convention in the text by mcconnell @xcite , @xmath18 ( in units of s@xmath19 ) is given by @xmath20 where @xmath21 is the lag time in the autocorrelation , @xmath22 is the vacuum permeability , @xmath23 is the reduced planck constant , and @xmath24 is the gyromagnetic ratio . @xmath25 , where @xmath26 is the polar angle ( see fig . [ fg : hept ] ) and @xmath27 the azimuthal angle . the ensemble average is performed by a double summation over spin pairs @xmath28 and @xmath29 of the nuclear ensemble , with @xmath30 . the subscripts @xmath31 and @xmath32 correspond to the partial ensembles @xmath33 and @xmath34 for intramolecular or intermolecular @xmath0h@xmath0h dipole - dipole interactions , respectively . this is illustrated in fig . [ fg : hept ] , -c@xmath35h@xmath36 ( heptane ) molecules , with hydrogen ( white ) and carbon ( black ) atoms . a @xmath0h nuclear spin is shown interacting with a neighboring @xmath0h nuclear spin on the same heptane molecule , known as the intramolecular dipole - dipole interaction ( red ) which is characterized by the rotation correlation time @xmath37 of a single molecule . the same @xmath0h nuclear spin is also shown interacting with a @xmath0h nuclear spin from a different heptane molecule , known as the intermolecular dipole - dipole interaction ( blue ) which is characterized by the translational correlation time @xmath38 between molecules . axes illustrate the distance @xmath39 between nuclear spins * i@xmath40 * and * i@xmath41 * with respect to the applied static magnetic - field direction @xmath42 , and corresponding polar angle @xmath43 . ] where the intramolecular interaction corresponds to @xmath44 in the case of heptane ( @xmath1-c@xmath35h@xmath36 ) , while @xmath34 corresponds to @xmath0h in all other molecules . the partition into partial ensembles @xmath33 and @xmath34 is a straight - forward book - keeping step in md simulations , and leads to a clear separation of @xmath45 and @xmath46 . the brackets in eq . [ eq : gmrty ] refer to the average over recorded time @xmath47 during the production run : @xmath48 where @xmath47 is recorded in 0.1 ps increments , with a total of @xmath4920000 for hydrocarbons and @xmath50 for water . for spherically symmetric systems such as the bulk fluids in question , the autocorrelation function @xmath51 is independent of the order @xmath52 , which amounts to saying that the direction of the applied magnetic field @xmath42 is arbitrary . for simplicity , we study the @xmath53 harmonic @xmath54 @xcite by md simulations . thus the final expression for the autocorrelation used here is : @xmath55 from the stored simulation trajectory , both an ensemble average over @xmath33 or @xmath34 and a time average over @xmath47 were performed in eq . [ eq : gmrt ] . to economize on computational time used in the analysis , for all the systems , only a subset of molecules was sampled for constructing the above average . for a chosen molecule , a hydrogen is first selected and its separation ( @xmath56 ) from all other hydrogen nuclei on the chosen molecule is recorded ( intramolecular ) , as well as the separation from all other hydrogen nuclei in the rest of the fluid ( intermolecular ) . such a record is then constructed for all time points @xmath47 . from the record of @xmath57 and @xmath58 ( fig . [ fg : hept ] ) , we calculated @xmath59 at different lag times @xmath21 using fast fourier transforms for computational efficiency ( the code is available upon request ) . the computed lag time ranged from 0 ps @xmath60 150 ps in increments of 0.1 ps for hydrocarbons and 0 ps @xmath60 50 ps for water in increments of 0.125 ps . for intramolecular interactions using eq . [ eq : gmrt ] and ( b ) autocorrelation function @xmath61 for intermolecular interactions using eq . [ eq : gmrt ] , for bulk alkanes ( solid lines , arrow shows increasing carbon number ) and water ( blue line ) , at @xmath62c . @xmath63-axis has been normalized by zero time value @xmath64 , and @xmath65-axis has been normalized by correlation time @xmath66 ( eq . [ eq : taurt ] ) . the relaxation behavior based on the traditional hard - sphere model ( eq . [ eq : gmodel ] ) is indicated by the black dashed line.,title="fig : " ] for intramolecular interactions using eq . [ eq : gmrt ] and ( b ) autocorrelation function @xmath61 for intermolecular interactions using eq . [ eq : gmrt ] , for bulk alkanes ( solid lines , arrow shows increasing carbon number ) and water ( blue line ) , at @xmath62c . @xmath63-axis has been normalized by zero time value @xmath64 , and @xmath65-axis has been normalized by correlation time @xmath66 ( eq . [ eq : taurt ] ) . the relaxation behavior based on the traditional hard - sphere model ( eq . [ eq : gmodel ] ) is indicated by the black dashed line.,title="fig : " ] the simulation results for @xmath59 are shown in fig . [ fg : gt ] , where both @xmath65 and @xmath63 axes have been normalized for better comparison between the different fluids . the parameters used for the normalization are detailed below . a significant parameter in the analysis is the autocorrelation at @xmath67 , which is given by the following expression : @xmath68 nmr relaxation theory states a simple relation between @xmath64 and the second - moment @xmath69 of the dipole - dipole interaction as such @xcite : @xmath70 which leads to the final expression : @xmath71 the second - moment is significant because it determines the strength of the relaxation mechanism , and it is a strong function ( @xmath72 ) of the nuclear spin separation @xmath56 . when the fluctuations in @xmath39 and @xmath43 are uncorrelated , the angular term in the numerator can be averaged out , resulting in the following approximate expression : @xmath73 md simulations indicate that eq . [ eq : smrtuncorr ] holds within @xmath74 for both intramolecular and intermolecular interactions , indicating that the ensemble averages of @xmath39 and @xmath43 are indeed uncorrelated . this approximation will not necessarily hold for fluids under confinement such as in nano - porous media , where complications from residual dipolar - coupling may come into play @xcite . in such cases the full expression in eq . [ eq : smrt ] must be used . once the autocorrelation function @xmath59 is determined from md simulations , the spectral density of the local magnetic - field fluctuations is determined by fourier transform : @xmath75 given that @xmath59 is real ( @xmath53 ) and an even function of @xmath21 . @xmath76 ( in units of s@xmath77 ) is defined using the non - unitary angular - frequency form of the fourier transform . the relaxation times @xmath78 and @xmath79 are then determined by @xmath76 at specific angular frequencies given by @xcite @xmath80 where @xmath81 is the larmor frequency , and @xmath82 mhz / t is the @xmath0h nuclear gyro - magnetic ratio . the relaxation rates are additive , giving the following final expression : @xmath83 which are the measurable quantities in the laboratory . the most significant parameter in the description of nmr relaxation in liquids is the correlation time @xmath66 , which is defined by the following expressions @xcite : @xmath84 in other words , @xmath66 is defined as the normalized area under the autocorrelation functions @xmath59 . [ eq : taurt ] also shows the simple relation to the spectral density @xmath85 at @xmath86 using eq . [ eq : fourierrtcos ] . as shown below in section [ sc : results ] , the correlation times @xmath66 for the bulk fluids are in the fast motion regime , i.e. @xmath87 and @xmath88 , where @xmath89 mhz for the nmr relaxation measurements @xcite used in this report . in the fast motion regime , the spectral density @xmath76 becomes independent of @xmath90 for @xmath91 , and equal to the @xmath86 component as such : @xmath92 from eqs . [ eq : t12r ] and [ eq : jrtmotional ] we obtain the following expression for the fast motion regime : @xmath93 in other words , @xmath94 and @xmath95 , which is characteristic for low viscosity fluids . from eqs . [ eq : t12rtmotional ] and [ eq : t12sum ] we thus obtain @xmath96 which is the final expression used to predict the nmr relaxation times from simulation results . as shown below , the md simulation results are compared with measured ( de - oxygenated ) @xmath97 data from shikhov _ et al . the nmr measurements were acquired at @xmath98 mhz using a cpmg echo train with an echo spacing of @xmath99 0.2 ms . the samples were carefully de - oxygenated by the freeze - pump - thaw technique in order to remove unwanted ( in the present case ) nmr relaxation from paramagnetic o@xmath100 molecules in solution . an independent computation of translational diffusion @xmath9 was performed from md simulations . the mean square displacement was computed from the center - of - mass of the molecule and plotted against diffusion evolution time @xmath101 ps ( fig . [ fg : diff ] ) . versus time @xmath21 for ( a ) water and @xmath1-c@xmath15h@xmath3 @xmath102 @xmath1-c@xmath103h@xmath104 ( arrow shows increasing carbon number ) , and ( b ) @xmath1-c@xmath105h@xmath106 @xmath102 @xmath1-c@xmath4h@xmath5 , at @xmath62c . solid lines show fitting region used to obtain translational diffusion coefficient @xmath9 from eq . [ eq : dsimulation ] for @xmath107 ps , and dashed lines show early time regime not used in the fit.,title="fig : " ] versus time @xmath21 for ( a ) water and @xmath1-c@xmath15h@xmath3 @xmath102 @xmath1-c@xmath103h@xmath104 ( arrow shows increasing carbon number ) , and ( b ) @xmath1-c@xmath105h@xmath106 @xmath102 @xmath1-c@xmath4h@xmath5 , at @xmath62c . solid lines show fitting region used to obtain translational diffusion coefficient @xmath9 from eq . [ eq : dsimulation ] for @xmath107 ps , and dashed lines show early time regime not used in the fit.,title="fig : " ] at long - times ( @xmath21 ) , the slope of the linear diffusive regime ( fig . [ fg : diff ] ) gives the translational self - diffusion coefficient @xmath9 according to einstein s relation : @xmath108 in the linear regression procedure , the early ballistic regime and part of the linear regime is excluded to obtain a robust estimate of @xmath9 . as shown below , the md simulation results are compared with measured @xmath9 data acquired using @xmath0h nmr diffusivity , where diffusion was encoded using uni - polar pulsed - field - gradient stimulated - echo experiments @xcite with diffusion evolution times of the order @xmath109 ms . due to an incomplete range of alkanes @xmath1-c@xmath2h@xmath3 @xmath102 @xmath1-c@xmath4h@xmath5 from a single measurement source , the @xmath9 measurements were taken from shikhov _ et al . _ @xcite at @xmath89 mhz , and from tofts _ et al . _ @xcite at @xmath110 mhz . this section presents the results of the nmr relaxation and diffusion , followed by analysis using hard - sphere models . . [ eq : t12motional ] ) from md simulations on the @xmath63-axis versus ( de - oxygenated ) @xmath8 measurements taken from shikhov _ et al . _ @xcite on the @xmath65-axis , for alkanes and water , showing a very strong correlation of @xmath111 and a small deviation of @xmath112 ( eq . [ eq : nrmsd ] ) . ( b ) cross - plot of predicted translational diffusion coefficient @xmath9 ( eq . [ eq : dsimulation ] ) from md simulations on the @xmath63-axis versus @xmath9 measurements taken from shikhov _ et al . _ @xcite ( open symbols ) and taken from tofts _ et al . _ @xcite ( closed symbols ) on the @xmath65-axis , for alkanes and water , showing a strong correlation of @xmath113 and a deviation of @xmath114 . data are at @xmath62c ( or @xmath115c ) , and are listed in table [ tb:1 ] . carbon number increases from right - to - left in ( a ) and ( b).,title="fig : " ] ( eq . [ eq : t12motional ] ) from md simulations on the @xmath63-axis versus ( de - oxygenated ) @xmath8 measurements taken from shikhov _ et al . _ @xcite on the @xmath65-axis , for alkanes and water , showing a very strong correlation of @xmath111 and a small deviation of @xmath112 ( eq . [ eq : nrmsd ] ) . ( b ) cross - plot of predicted translational diffusion coefficient @xmath9 ( eq . [ eq : dsimulation ] ) from md simulations on the @xmath63-axis versus @xmath9 measurements taken from shikhov _ et al . _ @xcite ( open symbols ) and taken from tofts _ et al . _ @xcite ( closed symbols ) on the @xmath65-axis , for alkanes and water , showing a strong correlation of @xmath113 and a deviation of @xmath114 . data are at @xmath62c ( or @xmath115c ) , and are listed in table [ tb:1 ] . carbon number increases from right - to - left in ( a ) and ( b).,title="fig : " ] 18cmccccccccccccc & @xmath8 & @xmath6 & @xmath116 & @xmath117 & @xmath37 & @xmath38 & @xmath118 & @xmath119 & @xmath9 & @xmath9 & @xmath9 + @xmath120 & & de - oxy & & & & & & /2@xmath121 & /2@xmath121 & ( @xmath122 & ( @xmath122 & ( @xmath122 + @xmath120 & & ( s ) & ( s ) & ( s ) & ( s ) & ( ps ) & ( ps ) & ( khz ) & ( khz ) & m@xmath123/s ) & m@xmath123/s ) & m@xmath123/s ) + @xmath120 & & ref . @xcite & md & md & md & md & md & md & md & ref . @xcite & ref . @xcite & md + @xmath120 & & 22.6@xmath14c & 20@xmath14c & 20@xmath14c & 20@xmath14c & 20@xmath14c & 20@xmath14c & 20@xmath14c & 20@xmath14c & 22.6@xmath14c & 20@xmath14c & 20@xmath14c + & 3.28 & 3.09 & 5.22 & 7.56 & 2.70 & 3.99 & 23.22 & 15.87 & 2.15 & 2.03 & 1.87 + & 13.44 & 14.34 & 24.26 & 35.06 & 0.77 & 2.32 & 20.17 & 9.66 & & & 3.85 + & 9.16 & 10.31 & 16.38 & 27.81 & 1.15 & 2.89 & 20.05 & 9.72 & 3.70 & & 2.88 + & 6.78 & 7.46 & 11.13 & 22.63 & 1.71 & 3.55 & 19.99 & 9.73 & 2.99 & & 2.24 + & 4.99 & 5.40 & 7.64 & 18.47 & 2.50 & 4.35 & 19.93 & 9.73 & 2.08 & 2.14 & 1.66 + & 3.68 & 4.13 & 5.67 & 15.22 & 3.40 & 5.24 & 19.87 & 9.76 & & 1.63 & 1.31 + & 2.86 & 3.13 & 4.14 & 12.72 & 4.66 & 6.30 & 19.83 & 9.74 & 1.20 & 1.26 & 1.01 + & 2.23 & 2.47 & 3.20 & 10.76 & 6.05 & 7.45 & 19.82 & 9.74 & & 1.01 & 0.80 + & 1.78 & 1.79 & 2.35 & 7.55 & 8.27 & 10.62 & 19.80 & 9.73 & 0.82 & 0.78 & 0.61 + & 1.44 & 1.41 & 1.80 & 6.50 & 10.79 & 12.42 & 19.76 & 9.70 & & 0.63 & 0.51 + & 1.19 & 1.25 & 1.58 & 5.95 & 12.40 & 13.66 & 19.71 & 9.66 & 0.56 & 0.49 & 0.44 + & 0.98 & 0.96 & 1.20 & 4.94 & 16.29 & 16.37 & 19.75 & 9.69 & & 0.40 & 0.34 + & 0.83 & 0.84 & 1.04 & 4.44 & 18.96 & 18.24 & 19.66 & 9.69 & 0.38 & 0.34 & 0.29 + & 0.71 & 0.72 & 0.87 & 4.12 & 22.41 & 20.05 & 19.79 & 9.59 & & & 0.25 + the results and analysis leading up to eq . [ eq : t12motional ] do not have any adjustable parameters in the interpretation of the simulation data . as shown in fig . [ fg : xplot](a ) , the md simulation results are very strongly correlated with measurement ( @xmath111 ) , and an estimate of the deviation can be established using the normalized root - mean - square deviation @xmath124 , defined as @xmath125 where @xmath126 are the md simulation results , @xmath127 are the measurements , and @xmath11 are the number of data points . the data in fig . [ fg : xplot](a ) indicate a small @xmath128 , which could come from uncertainties in the simulations and/or the measurements . the slight difference in temperature between simulation ( 20@xmath14c ) and measurement ( 22.6@xmath14c ) is unlikely to be the cause of the deviation , since this would result in a systematic difference for all the data . the experimental challenge is also to fully de - oxygenate the samples ; imperfect de - oxygenation would cause the measured @xmath8 to be systematically lower than @xmath6 from simulation , however this is also not found to be the case . on the computational side , the most likely source of uncertainties is from the truncation of @xmath59 before @xmath129 . this has the effect of underestimating @xmath66 ( eq . [ eq : taurt ] ) and therefore overestimating @xmath6 ( eq . [ eq : t12motional ] ) . the resulting correlation times @xmath66 are listed in table [ tb:1 ] , and plotted in fig . [ fg : tausm](a ) . for the alkanes , both @xmath66 increase by over an order of magnitude with increasing chain length ( i.e. carbon number ) . @xmath130 is also found to be @xmath131 shorter than @xmath132 at low carbon number , and are found to merge at high carbon number . the @xmath66 values for water are found to be comparable to octane ( @xmath1-c@xmath133h@xmath134 ) . the resulting second - moments @xmath69 are listed in table [ tb:1 ] and plotted in fig . [ fg : tausm](b ) . the data are presented as the square - root @xmath135 ( @xmath136 ) in units of khz . for the alkanes , the intramolecular @xmath137 khz are found to be @xmath138 larger than the intermolecular @xmath139 khz . the values for intermolecular contribution to the second moment for water are @xmath140 larger than alkanes , which is expected given the closer packing of water due to hydrogen bonding . ( eq . [ eq : taurt ] ) , intermolecular correlation time @xmath38 ( eq . [ eq : taurt ] ) , and intermolecular diffusion correlation time @xmath141 , for water ( w ) and alkanes ( carbon number c5@xmath102c17 ) . ( b ) md simulation result for square - root of intramolecular second moment @xmath118 ( eq . [ eq : smrt ] ) and square - root of intermolecular second moment @xmath119 ( eq . [ eq : smrt ] ) , for water ( w ) and alkanes ( c5@xmath102c17 ) . data are at @xmath62c , and are listed in table [ tb:1].,title="fig : " ] ( eq . [ eq : taurt ] ) , intermolecular correlation time @xmath38 ( eq . [ eq : taurt ] ) , and intermolecular diffusion correlation time @xmath141 , for water ( w ) and alkanes ( carbon number c5@xmath102c17 ) . ( b ) md simulation result for square - root of intramolecular second moment @xmath118 ( eq . [ eq : smrt ] ) and square - root of intermolecular second moment @xmath119 ( eq . [ eq : smrt ] ) , for water ( w ) and alkanes ( c5@xmath102c17 ) . data are at @xmath62c , and are listed in table [ tb:1].,title="fig : " ] as shown in fig . [ fg : xplot](a ) , the md simulations provide an accurate method of predicting the total nmr relaxation times @xmath6 . the next natural step is to use the md simulations to obtain new insight into intramolecular @xmath142 versus intermolecular @xmath143 relaxation . the first important insight is the relative strength of the two mechanisms . [ fg : ratios](a ) shows the ratio @xmath144 for each fluid , where a large ratio indicates that the intramolecular relaxation ( short @xmath142 ) dominates over intermolecular ( long @xmath143 ) . both pentane and water indicate that @xmath145 , however the ratio steadily increases with increasing chain length . the data suggests that larger molecules become increasingly dominated by intramolecular interactions ( i.e. rotational motions ) versus intermolecular interactions ( i.e. translational motions ) . this is an important finding for simulating nmr relaxation in macromolecules such as polymers and proteins , where , perhaps without much justification , the common practice is to only consider intramolecular dipole - dipole interactions . the md simulation results for @xmath9 are listed in table [ tb:1 ] and plotted in fig . [ fg : xplot](b ) against measurements . the nmr diffusion measurement is ( in principle ) independent of the applied magnetic field @xmath146 , provided no internal gradients exist , which is the case for bulk liquids . the nmr diffusion measurement is also ( in principle ) independent of whether the samples are air - saturated or de - oxygented . the air - saturated @xmath9 data were chosen from @xcite due to the larger range of alkanes ( @xmath1-c@xmath147h@xmath148 @xmath102 @xmath1-c@xmath36h@xmath149 ) measured . one potential complication in measuring nmr diffusion for bulk liquids is convection due to temperature gradients in the sample , which would tend to overestimate @xmath9 . however , these measurements were taken at a stable ambient temperature and therefore should not contain such complications . while the md simulations and measurements are strongly correlated ( @xmath113 ) , there is a systematic deviation of @xmath150 ( eq . [ eq : nrmsd ] ) . both the md simulations and the measurements have statistical uncertainties associated with them , however it is evident from fig . [ fg : xplot](b ) that the deviation @xmath150 is dominated by a systematic offset , where md simulations appear lower than measurement . it should be noted however that the agreement is remarkable given the huge difference in diffusion evolution time between md simulation @xmath151 ps and nmr measurement @xmath152 ms . put otherwise , the translational diffusion coefficient is compared over 9 orders of magnitude in diffusion evolution time , and found to be constant within @xmath153 . this could be viewed as a testament to einstein s theory of brownian motion ( eq . [ eq : dsimulation ] ) . in all of the results summarized in table [ tb:2 ] , there are no models or adjustable parameters used in going from md simulation results to predictions of nmr relaxation and diffusion . however in order to gain further insight into the molecular dynamics of bulk hydrocarbons and water , a physical model is required . the intramolecular dipole - dipole interaction @xcite is traditionally derived using the rotational diffusion equation of rank 2 for hard - spheres of radius @xmath154 , and is parameterized by the rotational diffusion coefficient @xmath155 . the parameter @xmath155 is simply related to rotational correlation time @xmath37 , which is defined as the average time it takes the molecule to rotate by 1 radian . the intermolecular dipole - dipole interaction @xcite is traditionally derived using the brownian motion model where the diffusion propagator is derived for hard - spheres of radius @xmath156 , and is parameterized by the translational diffusion coefficient @xmath9 . the parameter @xmath9 is simply related to the diffusion correlation time @xmath157 , which is defined as the average time it takes the molecule to diffuse by one hard - core diameter @xmath158 @xcite . for hard - spheres , the diffusion coefficients ( and correlation times ) can be related to the bulk properties of the fluid , namely the viscosity @xmath159 and absolute temperature @xmath32 , using the traditional stokes - einstein relation @xcite : @xmath160 note that in order to relate the stokes - einstein diffusion correlation time @xmath157 to the nmr derived translational correlation time @xmath38 ( eq . [ eq : taurt ] ) , a factor of @xmath161 is required , i.e. @xmath162 @xcite . an expression for the two autocorrelation functions are then derived as such : @xmath163 where both expressions reduce to @xmath64 ( eq . [ eq : g0sm ] ) at @xmath67 . [ eq : t12rtmotional ] ) of intermolecular relaxation time to intramolecular relaxation time , for water ( w ) and alkanes ( carbon number c5@xmath102c17 ) . ( b ) md simulation result for ratio @xmath164 of intermolecular diffusion correlation time ( @xmath162 ) to intramolecular correlation time ( @xmath37 ) , along with predicted value @xmath165 @xcite from stokes - einstein relation ( eq . [ eq : taumodel ] ) , for water ( w ) and alkanes ( c5@xmath102c17 ) . data are at @xmath62c , and can be inferred from table [ tb:1].,title="fig : " ] ( eq . [ eq : t12rtmotional ] ) of intermolecular relaxation time to intramolecular relaxation time , for water ( w ) and alkanes ( carbon number c5@xmath102c17 ) . ( b ) md simulation result for ratio @xmath164 of intermolecular diffusion correlation time ( @xmath162 ) to intramolecular correlation time ( @xmath37 ) , along with predicted value @xmath165 @xcite from stokes - einstein relation ( eq . [ eq : taumodel ] ) , for water ( w ) and alkanes ( c5@xmath102c17 ) . data are at @xmath62c , and can be inferred from table [ tb:1].,title="fig : " ] the model for the intramolecular autocorrelation @xmath45 is plotted in fig . [ fg : gt](a ) as the dashed black lines . the model @xmath45 is an exponential decay , which looks like a straight - line decay on the semilog-@xmath63 plot . it is clear that none of the intramolecular @xmath166 resemble an exponential decay , but appear to be more stretched " in nature . what is more remarkable from fig [ fg : gt](a ) is that even the intramolecular autocorrelation @xmath166 for water is _ not _ exponential . @xmath166 for water only involves two @xmath167 spins , and therefore can not contain cross - correlation effects such as predicted for three or more spins @xcite . in the present case , cross - correlation effects originate from interference between different interaction mechanisms , such as a dipole - dipole interaction between two spins and a dipole - dipole interaction between a different pair of spins @xcite . given that cross - correlation can not exist for intramolecular @xmath166 in water indicates that the observed disagreement with eq . [ eq : gmodel ] is due to the non - spherical nature of the molecules , and not cross - correlation effects . it also should be clarified that a stretched " autocorrelation function @xmath166 observed in fig . [ fg : gt](a ) does _ not _ imply a stretched " or multi - exponential relaxation @xmath116 . all of the measured @xmath8 @xcite and @xmath168 @xcite data are consistent with single - exponential decays . the nature of the stretched " correlation function @xmath166 is simply a reflection that the underlying interaction is not well modeled by hard spheres , and that a more sophisticated theory is required . similar observations and conclusions can be made of the intermolecular @xmath61 in fig . [ fg : gt](b ) , where the alkanes appear to be more stretched " in nature than the model , and increasingly so with larger chain - lengths . @xmath61 for water appears more consistent with the model , but still shows deviations . one informative way to quantify deviations from the hard - sphere model is to plot the ratio @xmath169 , which according to eq . [ eq : gmodel ] should be a constant @xmath170 @xcite if the stokes - einstein radius for rotation @xmath154 equals the stokes - einstein radius for translation @xmath156 . the ratio plotted in fig . [ fg : ratios](b ) clearly shows that lower alkanes tend towards the expectations of the model , while the higher alkanes increasingly depart from the model . this finding again confirms that the lower alkanes are more spherical " than the higher alkanes , but surprisingly water , which on the basis of the molecular structure should be considered closest to being spherical , " has relaxation times similar to octane . this emphasizes that correct account of intermolecular interactions , crucially short - range hydrogen bonding in the case of water , is necessary in capturing the relaxation behavior . this adds an additional level of complexity to the analysis , but one that is captured by the md simulations . more discussions of the stokes - einstein radii are given below . for completeness , the fourier transform ( eq . [ eq : fourierrtcos ] ) of the autocorrelation functions eq . [ eq : gmodel ] leads to the following spectral densities for the model @xcite : @xmath171 both expressions correctly reduce to @xmath85 ( eq . [ eq : jrtmotional ] ) in the fast motion regime @xmath172 and @xmath173 . the origin of the @xmath162 factor comes from normalizing the @xmath174 integral in eq . [ eq : jmodel ] , while keeping the 2/3 pre - factor the same as @xmath175 @xcite . in the hard - sphere model , the intramolecular interaction is based on a rotating hard sphere where the dipole - dipole distances @xmath39 are assumed to be rigid , i.e. @xmath176 are time independent . this simplifies the expression for the second moment @xmath177 in eq . [ eq : smrtuncorr ] to the following : @xmath178 where @xmath177 is consistent with previous definitions @xcite . the full md simulation for @xmath177 with non - rigid bonds ( eq . [ eq : smrtuncorr ] ) are slightly higher @xmath179 than with rigid bonds ( eq . [ eq : smmodel ] ) . the hard - sphere model for the intermolecular interaction is based on diffusing hard spheres with a hard - core distance @xmath158 . the model yields the above expression for @xmath180 in terms of @xmath156 and the density of @xmath0h spins defined as @xmath181 . the following expression is used @xmath182 , where @xmath183 is the number of @xmath0h per formula unit , @xmath184 is avogadro s number , @xmath185 is the molar mass , and @xmath186 is the density from table [ tb:2 ] . it is informative to compare the various sets of molecular radii that result from md simulations and the stokes - einstein relation in eq . [ eq : taumodel ] . as shown in fig . [ fg : ratios](b ) , the predicted ratio of translational diffusion to rotational correlation times are @xmath187 , which implies that @xmath188 . using eq . [ eq : taumodel ] and viscosity @xmath159 data in table [ tb:1 ] results in the stokes - einstein radii @xmath154 and @xmath156 , which are listed in table [ tb:2 ] and plotted in fig . [ fg : radius ] . while @xmath189 in the case of pentane , the rotational radius @xmath154 increases mildly with carbon number , while @xmath156 mildly _ decreases _ with increasing carbon number . these trends indicate that the molecules become less spherical " with increasing chain length . 8.5cmcccccccc & @xmath159 & @xmath186 & @xmath154 & @xmath156 & @xmath156 & @xmath190 + @xmath120 & & ( cp ) & ( g/@xmath191 ) & @xmath192 & @xmath192 & @xmath192 & @xmath192 + @xmath120 & & ref . @xcite & ref . @xcite & eq . [ eq : taumodel ] & eq . [ eq : taumodel ] & eq . [ eq : smmodel ] & md + & 1.00 & 0.998 & 1.38 & 1.02 & 0.97 & 1.74 + & 0.24 & 0.626 & 1.47 & 1.38 & 1.32 & 1.76 + & 0.31 & 0.659 & 1.53 & 1.35 & 1.32 & 2.20 + & 0.42 & 0.684 & 1.58 & 1.32 & 1.33 & 2.26 + & 0.55 & 0.702 & 1.64 & 1.29 & 1.34 & 2.50 + & 0.72 & 0.718 & 1.66 & 1.25 & 1.34 & 2.75 + & 0.93 & 0.731 & 1.69 & 1.22 & 1.35 & 2.97 + & 1.19 & 0.740 & 1.70 & 1.19 & 1.35 & 3.21 + & 1.51 & 0.750 & 1.74 & 1.24 & 1.35 & 3.43 + & 1.89 & 0.756 & 1.77 & 1.21 & 1.36 & 3.66 + & 2.34 & 0.760 & 1.72 & 1.16 & 1.36 & 3.88 + & 2.87 & 0.769 & 1.76 & 1.15 & 1.36 & 4.11 + & 3.48 & 0.773 & 1.74 & 1.12 & 1.36 & 4.28 + & 4.21 & 0.778 & 1.73 & 1.09 & 1.37 & 4.51 + another prediction from the hard - sphere model is @xmath156 according to @xmath193 defined in eq . [ eq : smmodel ] . using the values of @xmath181 and @xmath193 in table [ tb:1 ] results in a prediction for @xmath156 , which is listed in table [ tb:2 ] and plotted in fig . [ fg : radius ] . since both @xmath181 and @xmath193 show little variation with carbon number , the resulting @xmath156 is also found to be roughly independent of carbon number . in the case of pentane , a consistent value of @xmath156 according to @xmath194 and @xmath193 confirm that the model is more accurate for more spherical " molecules . from eq . [ eq : taumodel ] ( using @xmath194 , with @xmath159 listed in table [ tb:2 ] ) ) , stokes - einstein radius for translational motion @xmath156 from eq . [ eq : taumodel ] ( using @xmath194 ) , @xmath156 prediction from second moment @xmath180 from eq . [ eq : smmodel ] ( using @xmath186 from table [ tb:1 ] ) , and @xmath190 from md simulations ( eq . [ eq : rg ] ) , for water ( w ) and alkanes ( carbon number c5@xmath102c17 ) . data are at @xmath62c , and are listed in table [ tb:2 ] . ] the radius of gyration of the molecule , defined by @xmath195 provides an independent estimate of the size of the molecule ; for a spherical molecule the @xmath190 is expected to be a reasonable estimate of the hydrodynamic radius . here @xmath190 is calculated by sampling a single molecule from the simulation and averaging over the entire trajectory . as expected , increasing the sample size of the simulation did not change the results . the simulation results for @xmath190 are listed in table [ tb:2 ] and plotted in fig . [ fg : radius ] . the results in fig . [ fg : radius ] clearly show that while @xmath190 increases linearly with chain length , both @xmath154 and @xmath156 show only mild variations with increasing chain length . this suggests that the nmr related quantities @xmath154 and @xmath156 remain more local in nature than the structure - based @xmath190 radius . in the case of water , the pattern @xmath196 again emphasizes that it is best to not regard water as a spherical object , as required by the traditional hard - sphere model . md simulation techniques for predicting nmr relaxation and diffusion in bulk hydrocarbons and water are validated against nmr measurements , without any adjustable parameters in the interpretation of the simulation data . md simulations reveal new insight about the intramolecular versus intermolecular nmr relaxation in bulk fluids , which were not easily accessible before . the simulations quantify the relative strength of the two relaxation mechanisms , indicating that intramolecular increasingly dominates over intermolecular relaxation with increasing molecular chain - length ( i.e. increasing carbon number ) . this validates the common practice of only considering intramolecular dipole - dipole interactions for simulations of macromolecules such as proteins and polymers . with increasing chain length , the simulations indicate increasing departure from the traditional hard - sphere models for the time dependence in autocorrelation functions of both intramolecular and intermolecular dipole - dipole interactions . the md simulations indicate that the stokes - einstein radius for rotational and translational motion are comparable for small molecules , but increasingly diverge with increasing chain length , indicating a departure from spherical " molecules for the higher alkanes . furthermore , with increasing chain length the radius of gyration is found to increase more rapidly than the nmr derived stokes - einstein radius , indicative of the more local nature of the nmr measurement . the validation of md simulations for predicting nmr relaxation and diffusion of bulk hydrocarbons and water opens up new opportunities for investigating more complex scenarios , such as the effects of nanometer confinement on the phase - behavior and transport properties of hydrocarbons and water in organic porous - media such as kerogen . it also opens up opportunities for studying nmr relaxation in complex fluids such as crude - oils or model systems such as polymer - solvent mixes . we gratefully acknowledge funding from the rice university consortium on processes in porous media . we gratefully acknowledge the national energy research scientific computing center , which is supported by the office of science of the u.s . department of energy [ de - ac02 - 05ch11231 ] , for hpc time and support . we also gratefully acknowledge the texas advanced computing center ( tacc ) at the university of texas at austin ( url : http://www.tacc.utexas.edu ) for providing hpc resources ; arjun valiya parambathu , jinlu liu , and zeliang chen for their assistance ; and , igor shikhov and christoph h. arns for sharing their experimental nmr data . 40ifxundefined [ 1 ] ifx#1 ifnum [ 1 ] # 1firstoftwo secondoftwo ifx [ 1 ] # 1firstoftwo secondoftwo `` `` # 1''''@noop [ 0]secondoftwosanitize@url [ 0 ] + 12$12 & 12#1212_12%12@startlink[1]@endlink[0]@bib@innerbibempty @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) `` , '' ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) `` , '' ( , ) chap . , `` , '' ( , ) chap . , _ _ , ( , , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) `` , '' ( , ) `` , '' ( , ) `` , '' ( , ) `` , '' ( , ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( ) @noop * * , ( )

the quantum vacuum is a fundamental concept in theoretical physics and its properties has been widely investigated in the literature of quantum gravity and string theory . the theory of quantum fields in curved spacetime which is believed to be the low energy limit of the ultimate theory of quantum gravity , has predicted famous quantum effects in the presence of gravity . in general , due to the lack of the global symmetries in the spacetime manifold , quantum effect considerations in curved spacetime are mainly limited to the analysis of the local quantities such as the vacuum expectation value of the energy - momentum tensor , i.e. @xmath0 , in some point @xmath1 . in fact , the most famous results of the semi - classical theory of gravity like the hawking radiation and the particle production in the expanding universe has been achieved from the analysis of the @xmath2 in the related curved backgrounds . the most famous vacuum state effect is the casimir effect . an important aspect in the researches around the casimir effect in curved spacetime is that the characteristics of the vacuum state are apparently dependent to the geometry of the background spacetime . we are also motivated to see explicitly such a dependency in this paper . furthermore , computation of the energy , i.e. @xmath3 , has also been done for large number of problems in various spacetimes and sometimes @xcite has helped us to confirm the validity of the principle of correspondence in the context of the casimir effect . finding the total gravitational force on a set of two conducting casimir plates @xcite is a typical example . so , due to the importance of the stress - energy tensor , we will consider the @xmath2 for the plates in a general weak background . the casimir effect arises when there is a boundary in our problem and it predicts a force between two uncharged conducting metals in the presence of a quantum field . the effect has been measured to a great accuracy @xcite . we use the zero - point energy approach here although it is possible to find the casimir force and energy without any reference to the zero - point energy @xcite . we may also have the casimir effect without having a boundary at all . in fact , some non - trivial topologies in curved spacetime do the same job as a boundary does@xcite,@xcite . the casimir energy in curved spacetime has been also analysed by many authors ( @xcite and references there in ) . recently , a casimir apparatus consisting of two ideal conducting parallel plates in the weak field limit of the kerr and the horava - lifshitz spacetimes has been studied in @xcite . a purpose of this paper is to generalize the above analysis for scalar field doing exact solution of the klein - gordon equation in a general weak gravitational field . also we extend the method for the case the electromagnetic field is present inside the plates . the organization of the paper is as follows . in section ii using the fact that the apparatus has composed of tiny pales , we will find the small deviations of the metric from flat spacetime inside the apparatus . in fact we will expand the metric up to first order in terms of the parameters of the general weak gravitational field . in section iii the klein - gordon equation will be solved exactly inside the apparatus using the metric obtained in the previous section . in section iv mode frequencies inside the apparatus will be obtained under the influence of both dirichlet and neumann boundary conditions for the scalar field . generalization to the case where the electromagnetic field is present will be done using an interesting property of the fermi spacetime in section v. computation of the energy momentum tensor for the scalar field for both neumann and dirichlet boundary conditions are done in section vi . also the electromagnetic energy density is obtained in this section . well known weak gravitational fields are examined in section vii . the electromagnetic energy density for the far field limit of the kerr spacetime and the horava - lifschitz theory of gravity are of special interest . the final section is devoted to the conclusion . in the 1 + 3 formalism of general relativity , stationary spacetime metric is defined by @xmath4 where @xmath5 is the so called gravitomagnetic potential and @xmath6 in the weak field slowly rotating limit ( @xmath7 ) , the metric ( [ eq01 ] ) is equivalent to @xmath8 the explicit form of a general weak gravitational field line element is as follows ( see ( 19.13 ) in @xcite ) @xmath9dt^2-[4\epsilon_{ijk}s^j\frac{x^k}{r^3}+o(\frac{1}{r^3})]dx^idt \\ & -\{(1+\frac{2m}{r}+\frac{3m^2}{2r^2})\delta_{ij}+gravitational\;radiation\ ; terms\;that\;die\ ; out\ ; as\ ; o(\frac{1}{r } ) \}dx^i dx^j , \end{split}\end{aligned}\ ] ] comparison between ( [ eq03 ] ) and ( [ eq04 ] ) shows that @xmath10 is the newtonian potential , @xmath11 is the gravitomagnetic potential and @xmath12 . we need to have an isotropic coordinate representation of the metric in the next sections and equations ( [ eq01]),([eq02 ] ) and ( [ eq03 ] ) shows us the way we can construct an isotropic coordinate representation of any weak gravitational field . . the boundaries are at z=0,z = l ] fig . [ fig : apparatus ] demonstrates the apparatus in a weak gravitational field . two ideal conducting plates are in a small coordinate distance @xmath13 from each other . inside the plates , the metric has a small variation relative to the flat spacetime metric . to find this variation , we adapt a rectangular coordinate system having the origin at one of the plates ( the one which is closer to the source and have coordinate distance @xmath14 from it ) and expand the metric ( [ eq04 ] ) inside the apparatus in the neighborhood of the point @xmath15 . the overall size of the apparatus is so small that we can assume @xmath16 : @xmath17 in which @xmath18 . for the case of a static spacetime , the components of the metric ( [ eq05 ] ) can be written in the following form : @xmath19 where @xmath20 are constants and use is made of @xmath21 . concerning the form of @xmath22 , the above expansion satisfies @xmath23 provided that @xmath24 i.e. in the equatorial plane . so ( [ eq05]),([eq06 ] ) are also valid for the case of the far field limit of the kerr spacetime and we will back to it in the examples in the section vi . motivated by the above discussion , in this paper we analyse the general case of the spacetime of the form @xmath25,\end{aligned}\ ] ] in which @xmath26 . to solve the klein - gordon equation , it is better to recast the metric ( [ eq07 ] ) to the known fermi metric . we use the linearized weak field regime of general relativity and change the variables with the aid of the following gauge transformation : @xmath27 which we have assumed @xmath28 to force the spatial sector of the metric ( [ eq07 ] ) to be flat . more explicitly we have : @xmath29 the metric then takes the following form up to first order in the parameters @xmath30 @xmath31 the gauge transformation , however , changes our primary problem as follows . in fact according to the last equation in ( [ eq08 ] ) the boundaries must be transformed from @xmath32 and @xmath33 in the spacetime ( [ eq07 ] ) to @xmath34 and @xmath35 in the spacetime ( [ eq10 ] ) . another change that the gauge transformation brings into the problem is the rescaling of time in ( [ eq09 ] ) by the factor @xmath36 . this corresponds , in turn , to dividing the mode frequencies @xmath37 by @xmath36 because of the presence of the factor @xmath38 in our solution of the klein - gordon equation in the next section . so time must be re - inverted after solving the klein - gordon equation when we obtain the mode frequencies . _ the net effect of the rescaling of time is that the final mode frequencies must be multiplied by the factor @xmath36_. we drop the dashes @xmath39 on @xmath40 from now on using the new boundary conditions instead . the massless klein - gordon equation is @xmath41= 0 \;\;\;\;\;\;,\;\;\ ; \mathfrak{g}\equiv { \rm det}g^{\mu\nu}.\end{aligned}\ ] ] since the spacetime is spatially flat we assume the following form for the solution @xmath42 where @xmath43 is normalization constant determined through the commutation relations @xmath44 the scalar product is defined as @xmath45^{\frac{1}{2}}n^{\mu } d\sigma.\end{aligned}\ ] ] in which @xmath46 and @xmath47 spans the space between the plates . under the above assumptions equation ( [ eq11 ] ) reads @xmath48 where @xmath49 denotes derivation with respect to z and @xmath50 . another variable change @xmath51 yields @xmath52 appearance of factor 3 in front of the second term , introduces a significant simplification when we change the variable to @xmath53 . it recasts ( [ eq16 ] ) into @xmath54 a simple reparametrization of this last equation via @xmath55 end up with @xmath56 this is the known kummer s differential equation @xmath57 in which @xmath58 . kummer s differential equation is not suitable for next considerations and we transform it via @xmath59 to another known form called wittaker s differential equation : @xmath60 in which in our case @xmath61 . equation ( [ eq20 ] ) has two independent set of solutions @xmath62 , @xmath63 and their asymptotic behaviour is as follows @xcite @xmath64 @xmath65 in general , @xmath63 is the acceptable physical solution which is finite at infinity . in the problem under consideration we must choose a linear combination of the two , due to the fact that both are finite in between the plates . the exact mode functions are : @xmath66 e^{-i\omega t - ik_xx -ik_yy}\;\;\;\;\;\;\;\;\\end{aligned}\ ] ] the asymptotic form of ( [ eq23 ] ) for small value of @xmath67 can be written as follows ( see appendix a ) @xmath68 in which @xmath69 . in the next section we use this asymptotic form to extract mode frequencies for both neumann and dirichlet boundary conditions on the plates . we put the approximation @xmath70 into ( [ eq24 ] ) . from the dirichlet boundary condition @xmath71 we have @xmath72 and from @xmath73 we have : @xmath74 after careful expansion of ( [ eq25 ] ) , the mode frequencies proved to satisfy the following relation : @xmath75\}=\sqrt{k_\perp^2+(\frac{n\pi}{l+\gamma_1 l+ \frac{1}{2 } \lambda_1 l^2})^2}\end{aligned}\ ] ] note that the factor @xmath76 is nothing but the _ proper distance _ between the plates and so we have @xmath77 in which @xmath78 denotes proper ( or the corresponding flat space ) mode frequencies in the local lorentz frame of an observer comoving with the plates . as stated in the previous section , the final mode frequencies will be obtained by multiplication of the factor @xmath36 due to rescaling of time during the gauge transformation ( [ eq09 ] ) . so we have the final mode frequencies inside the casimir apparatus for the spacetime ( [ eq07 ] ) : @xmath79 mode frequencies are influenced only by @xmath80 component of the metric from the point of view of a proper observer . the neumann boundary condition @xmath81 imposed on ( [ eq24 ] ) gives @xmath82 : @xmath83 another neuman boundary condition @xmath84 end up with @xmath85 which in turn result in ( [ eq26 ] ) and ( [ eq27 ] ) again . the electromagnetic field has two physical degrees of freedom and it is known that in the rindler spacetime the electromagnetic field can be represented in terms of two scalar fields satisfying the klein - gordon equation separately @xcite . the photon propagator and the energy - momentum tensor of the electromagnetic field in a weak gravitational of the fermi spacetime has also been obtained in @xcite . however in @xcite , the computations has been done through a lengthy and cumbersome method of the green functions . in a paper by the author @xcite , it is shown that the energy density ( i.e. the 0 - 0 component of the energy - momentum tensor ) of the electromagnetic field in fermi spacetime is exactly the some of the energy density of the two scalar fields mentioned in @xcite . the method was done without any address to the green function method frequently used in the literature . here , we briefly review the relationship between the two scalar fields and the electromagnetic field in fermi spacetime . the spin one vector field in curved spacetime in the lorentz gauge satisfies @xmath86 it can be shown that in fermi metric , the ricci tensor satisfies @xmath87 and the second term of the wave equation ( [ eq31 ] ) must be ignored . furthermore , because the metric ( [ eq10 ] ) is spatially flat , the lorentz gauge in ( [ eq31 ] ) can be broken into two independent parts @xcite : @xmath88 in which @xmath89 and @xmath90 we know also that both @xmath91 and @xmath92 satisfy the klein - gordon eqaution separately(see the appendix in @xcite ) . boundary condition for the electric field on the plates is @xmath93 and @xmath94 which in turn can be recast into boundary conditions on @xmath91 and @xmath92 according to equation ( [ eq33 ] ) . in @xcite it has been shown that boundary conditions for the electric and magnetic fields return a dirichlet boundary condition on @xmath92 and a neumann boundary condition on @xmath91 . we have shown in previous section that both of this conditions will be ended up to a same frequency shift . as a result mode frequencies in ( [ eq28 ] ) are also valid for the electromagnetic field . this section has three subsections . in the first subsection , the electromagnetic energy - momentum tensor will be represented in terms of the energy momentum of the scalar fields mentioned in ( [ eq33 ] ) . in the other two sections the energy momentum tensor of the scalar and vector fields will be calculated . the vacuum expectation value of the quantum energy - momentum tensor is defined as @xmath95.\end{aligned}\ ] ] the classical energy - momentum tensors for the scalar and vector fields are @xmath96 , \\ t_{\mu\nu}^{gauge } & = -c^*_{;\mu}c_{;\nu}-c^*_{;\nu}c_{;\mu}-g_{\mu\nu}g^{\lambda\theta}c_{;\lambda}c_{;\theta } , \\ t_{\mu\nu}^{maxwell } & = \frac{1}{4}g_{\mu\nu}f^{\lambda\theta}f_{\lambda\theta}-f_{\mu^\theta}f_{\theta\nu}. \end{split}\end{aligned}\ ] ] as is well - known , in the quantum level , the contributions of ghost and gauge fields in the electromagnetic energy - momentum tensor cancel each other @xcite and we only concern the maxwell sector of the energy - momentum tensor . expansion of @xmath97 in terms of scalar fields @xmath91 and @xmath92 shows that ( see appendix b ) @xmath98 in which @xmath99 . using @xmath100 and ( [ eq32 ] ) we have : @xmath101 in which we have used the general form of the wave function ( [ eq12 ] ) . quadratic products of fields @xmath102 and @xmath103 produce terms like @xmath104 which have no contribution when the expectation value is taken because of the fact that @xmath91 and @xmath92 are not correlated and belongs to independent hilbert spaces . we calculate @xmath105 component of the energy - momentum tensor for the electromagnetic field : @xmath106 @xmath107 \ } \\ & = k_\perp^2[<0|t_{00}^{\phi}|0>+<0|t_{00}^{\psi}|0 > ] \end{split}\end{aligned}\ ] ] after a lengthy but straightforward calculation , other components of the energy - momentum tensor in both sides are related to each other as follows : @xmath108 note that @xmath109 can be absorbed into @xmath43 . this section is devoted to the calculation of the energy density for the casimir apparatus via the direct method without any reference to the traditional green function method . using the approximations @xmath110 we expand the wave function ( [ eq24 ] ) and find up to first order in @xmath111 : @xmath112 sin(\sqrt{b}z)+\frac{a}{4\sqrt{b}}z^2 cos(\sqrt{b}z)\ } \;\;\ , dirichlet \\ z(z ) & = z_0 \{[1-(\frac{\lambda}{2}+\frac{a}{4b})z ] cos(\sqrt{b}z)-\frac{a}{4\sqrt{b}}z^2 sin(\sqrt{b}z)\ } \;\;\ , newmann \end{split}\end{aligned}\ ] ] @xmath113 can be absorbed also in @xmath114 . the energy density is defined by @xmath115 where @xmath116 is the lapse vector normal to the hypersurface @xmath117 i.e. @xmath118 . the mean energy density so has the following form : @xmath119 calculating the factor @xmath120 for both wave functions in ( [ eq45 ] ) results in @xmath121 \\ & + \frac{a}{4\sqrt{b}}[\frac{z}{2b}sin(2\sqrt{b}z ) + ( \frac{1}{2b}-z^2)\frac{cos(2\sqrt{b}z)}{2\sqrt{b } } ] \}|_\alpha^\beta , \\ f_2 = & \{\frac{z}{2}-\frac{sin(2\sqrt{b}z)}{4\sqrt{b}}-(\lambda+\frac{a}{2b})[\frac{z^2}{4}-\frac{z sin(2\sqrt{b}z)}{4\sqrt{b}}-\frac { cos(2\sqrt{b}z)}{8b } ] \\ & + \frac{a}{4\sqrt{b}}[\frac{z}{2b}sin(2\sqrt{b}z)+(\frac{1}{2b}-z^2)\frac{cos(2\sqrt{b}z)}{2\sqrt{b } } ] \}|_\alpha^\beta , \\ f_3 = & \ { b[\frac{z}{2}+\frac{sin(2\sqrt{b}z)}{4\sqrt{b}}+(\frac{a}{2b}-\lambda)[\frac{z^2}{4}+\frac{z sin(2\sqrt{b}z)}{4\sqrt{b}}+\frac { cos(2\sqrt{b}z)}{8b } ] ] \\ & -\frac{a\sqrt{b}}{4 } [ \frac{z}{2b}sin(2\sqrt{b}z)+(\frac{1}{2b}-z^2)\frac{cos(2\sqrt{b}z)}{2\sqrt{b } } ] + ( \frac{\lambda}{4}+\frac{a}{8b } ) cos(2\sqrt{b}z)\}|_\alpha^\beta . \\ \end{split}\end{aligned}\ ] ] @xmath122 \\ & -\frac{a}{4\sqrt{b}}[\frac{z}{2b}sin(2\sqrt{b}z ) + ( \frac{1}{2b}-z^2)\frac{cos(2\sqrt{b}z)}{2\sqrt{b } } ] \}|_\alpha^\beta , \\ f_2 = & \{\frac{z}{2}+\frac{sin(2\sqrt{b}z)}{4\sqrt{b}}-(\lambda+\frac{a}{2b})[\frac{z^2}{4}+\frac{z sin(2\sqrt{b}z)}{4\sqrt{b}}+\frac { cos(2\sqrt{b}z)}{8b } ] \\ & -\frac{a}{4\sqrt{b}}[\frac{z}{2b}sin(2\sqrt{b}z)+(\frac{1}{2b}-z^2)\frac{cos(2\sqrt{b}z)}{2\sqrt{b } } ] \}|_\alpha^\beta , \\ f_3 = & \ { b[\frac{z}{2}-\frac{sin(2\sqrt{b}z)}{4\sqrt{b}}+(\frac{a}{2b}-\lambda)[\frac{z^2}{4}-\frac{z sin(2\sqrt{b}z)}{4\sqrt{b}}-\frac { cos(2\sqrt{b}z)}{8b } ] ] \\ & + \frac{a\sqrt{b}}{4 } [ \frac{z}{2b}sin(2\sqrt{b}z)+(\frac{1}{2b}-z^2)\frac{cos(2\sqrt{b}z)}{2\sqrt{b } } ] -(\frac{\lambda}{4}+\frac{a}{8b } ) cos(2\sqrt{b}z)\}|_\alpha^\beta . \end{split}\end{aligned}\ ] ] simplification of the terms like @xmath123 and @xmath124 is possible using the fact that in boundaries the wave function ( [ eq24 ] ) must vanish and : @xmath125 in which @xmath126 takes one of two boundary values @xmath32 and @xmath127 . equation ( [ eq47 ] ) result in @xmath128 and the final result is as follows @xmath129(\beta^2-\alpha^2 ) \}.\end{aligned}\ ] ] exactly the same @xmath130 is obtained for dirichlet boundary condition . the constant @xmath43 was defined in ( [ eq07 ] ) and can be determined simply as : @xmath131 sin^2(\sqrt{b}z)+\frac{a}{4\sqrt{b}}z^2 sin(2\sqrt{b}z)\}dz\}^{-1 } \;\ , dirichlet \\ c^2 & = \frac{1}{2(2\pi)^2\omega } \{\int_0^l \{[1-(2\lambda+\frac{a}{2b})z ] cos^2(\sqrt{b}z)-\frac{a}{4\sqrt{b}}z^2 sin(2\sqrt{b}z)\}dz\}^{-1 } \;\ , neumann \end{split}\end{aligned}\ ] ] just like @xmath130 , the constant @xmath43 has also the same form for both boundary conditions up to first order in @xmath111 @xmath132 the final result for the energy - density after all is : @xmath133 in which @xmath134 is the corresponding flat spacetime casimir energy density @xmath135 @xcite . up to now , we have shown that for neumann and dirichlet boundary conditions there exists the relation @xmath136 between flat and curved energy density contents of the casimir apparatus in the weak spacetime of the metric ( [ eq07 ] ) . using the first equation of ( [ eq41 ] ) it is evident that the same shift in the energy density obtained in ( [ eq52 ] ) holds also for the electromagnetic field : @xmath137 in the next section we will analyse some well - known weak gravitational fields and find the parameters @xmath138 in each case . the problem of the divergences near a perfect generic conductor first studied systematically by deutsch and candelas @xcite . they found that the energy - momentum tensor near the surface behaves like : @xmath139 where @xmath140 are constants and @xmath141 is the distance from the surface of the ideal boundary . in case where there is a conformal invariance in the action @xmath142 . the divergences originates from the unphysical nature of classical ideal conductor boundary conditions . it has been shown @xcite that we can remove the infinities of the total energy ( and not the energy - momentum tensor ) of the plates using the zeta function regularization unless for the case that the zeta function has poles itself . on the other hand , the cut off regularization method suggests the removal of the divergences via _ ad hoc _ although in this method it still remains a logarithmic ambiguity @xcite in the energy density . for the imperfect conductors ( more realistic boundary conditions ) we can easily remove the divergences introducing some suitable cut off frequencies although the boundary effect may become quit large ( but finite)@xcite . now the question is that what happens when we go to curved spacetime ? . does the surface divergences of the energy - momentum tensor are ignorable in the semi - classical einstein s equations ? . the answer is negative according to @xcite . in @xcite however , the authors has found a way to get ride of the surface divergences of the einstein s field equations for the case the boundary is a parallelepiped . they have used a suitable cut off along with the so - called estrada - kanwal distribution theory of asymptotics to regularize / renormalize the infinities and show that the energy - momentum tensor near a plane boundary , as a source , converges to a consistent theory when the cutoff is removed . remarkably , the process of curing the divergences in curved spacetime has been continued by milton et al @xcite where they have shown that most of the energy between the plates is restored near the surfaces and a part of it resides exactly on the plates . they finally have shown that this energies responds to gravity just like any other finite energy following the newtonian relation @xmath143 . this is expectable as this large energies are simply a part of the total energy of the system . we show here that , in our case , the divergences do not present in the first order of approximation that we have used here and they appear only in the higher orders of approximation . to do so , we write down the explicit structure of possible divergences in the einstein s field equations along the lines depicted in @xcite . the one - loop effective action @xmath144 for the semi - classical theory of gravity is : @xmath145 in which @xmath146 is the dewitt - schwinger - feynmann s propagator . using the dewitt - schwinger representation of the action , the asymptotic expansion of @xmath147 is as follows : @xmath148 in which @xmath149 is a length scale to fix the dimensional issues . the potentially divergent part of the effective action ( the first three terms ) @xcite reads : @xmath150 \ } [ \frac{1}{n(n-2)}4m^2a_0-\frac{1}{(n-2)}2m^2a_1+a_2 ] \\. \end{split}\end{aligned}\ ] ] in which @xmath151 is the mass of the scalar field(m=0 in our case ) . the coefficients @xmath152 are : @xmath153 in which @xmath154 is the riemann s curvature tensor . the total gravitational lagrangian density can be shown to have the form ( see ( 6.49 ) in @xcite ) : @xmath155 \ } a_2 , \\ & a=\frac{4m^4}{(4\pi)^{\frac{n}{2 } } n(n-2 ) } \{\frac{1}{n-4}+\frac{1}{2}[\gamma+\ln{\frac{m^2}{\mu^2 } } ] \ } \\ & b=\frac{m^2}{(4\pi)^{3\frac{n}{2 } } ( n-2 ) } \{\frac{1}{n-4}+\frac{1}{2}[\gamma+\ln{\frac{m^2}{\mu^2 } } ] \}. \\ \end{split}\end{aligned}\ ] ] however , in the massless case of ours , the only non - vanishing potentially ultraviolent term is the one related to @xmath156 in ( [ eq52 ] ) and @xmath157 in ( [ eq52 ] ) vanish ( see ( 6.101 ) in @xcite ) . our calculation for the metric ( [ eq07 ] ) shows that @xmath156 is of second order of approximation : @xmath158 so the total lagrangian density ( [ eq59 ] ) reduces to the the standard bare density @xmath159 . in conclusion , the potentially divergent term @xmath156 vanishes within the first order of approximation in this section number of spacetimes are investigated and the parameters appeared in ( [ eq28]),([eq52]),([eq53 ] ) will be found . to this end , we will try to find their weak field form according to ( [ eq03]),([eq04 ] ) and ( [ eq07 ] ) . recently bezerra et al @xcite studied the apparatus for the scalar fields in the weak field limit of the kerr spacetime in the equatorial plane . in that work the apparatus co - rotates with the local angular velocity of the spacetime i.e. the measurements had been assumed to be done in the point of view of a zero angular momentum observer(zamo ) . they found the metric inside the apparatus through some two stage successive approximation method as follows : @xmath160\;\;\ ; 0-th\ ; order\ ; of\ ; approximation , \\ ds^2 \approx & ( 1 + 2b\lambda z)dt^2 - ( 1 -2\lambda z)[dx^2+dy^2+dz^2]\;\;\ ; first\;order\;of\;approximation . \end{split}\end{aligned}\ ] ] in which @xmath161 , @xmath162 and @xmath163 . @xmath164 is the angular momentum per mass and @xmath165 the local angular velocity of the kerr spacetime . the authors set @xmath166 for the first order of approximations . this does not works because @xmath111 is related to @xmath167 through @xmath168 . putting one of them equal to zero forces the other one to vanish also . evidently , in the first order of approximations we must keep both @xmath167 and @xmath169 and the metric must be written as follows instead of ( [ eq60 ] ) : @xmath170\;\;\ ; 0-th\ ; order\ ; of\ ; approximation , \\ ds^2 \approx & [ 1 + 2b(\phi_0+\gamma z)]dt^2 - [ 1 -2(\phi_0+\gamma z)][dx^2+dy^2+dz^2]\;\;\ ; first\ ; order\ ; of\ ; approximation . \end{split}\end{aligned}\ ] ] taking the above comment and the metric ( [ eq61 ] ) into account , we find @xmath171 and arrive at the following relations for the e__lectromagnetic casimir energy density _ _ and the mode frequencies : @xmath172 , \\ < 0|t_{00}^{maxwell}|0 > & = -\frac{\pi^2}{720 l^4}[1+b\phi_0+b\frac{\gamma l_p}{2 } ] . \end{split}\end{aligned}\ ] ] the scalar field casimir energy density inside the apparatus which has been sketched in eq.(32 ) in @xcite must also be corrected as follows : @xmath173 , \\ < 0|t_{00}^{scalar}|0 > & = -\frac{\pi^2}{1440 l^4}[1+b\phi_0+b\frac{\gamma l_p}{2 } ] . \end{split}\end{aligned}\ ] ] _ far field limit of the schwarzschild spacetime _ is also covered by setting @xmath174 the fermi spacetime is described by the following metric @xmath175 the importance of this metric is that it is traditionally recognized as the spacetime of a static accelerating observer near the surface of the source of a constant gravitational field @xcite . comparison of this metric to the general metric in ( [ eq07 ] ) gives @xmath176 : @xmath177 , \\ < 0|t_{00}^{maxwell}|0 > & = -\frac{\pi^2}{720 l^4}[1+\frac{a l_p}{2 } ] , \\ < 0|t_{00}^{scalar}|0 > & = -\frac{\pi^2}{1440 l^4}[1+\frac{a l_p}{2 } ] . \end{split}\end{aligned}\ ] ] the above energy densities are exactly the results ( 5.2 ) in @xcite , ( 3.4 ) in @xcite and ( 5.4 ) in @xcite . the hoava - lifshitz ( hl ) gravity is a renormalizable theory of gravity that is invariant under the lifshitz scaling transformation @xmath178 . this transformations manifestly break the space and time covariance . the anisotropy between space and time , in turn , may affects the casimir effect as well . it is interesting to investigate the vacuum characteristics of the theory . recently , the effect of the hl theory on the casimir energy of the apparatus has been studied in @xcite . the authors recommended to set a constraint on spacetime anisotropies in such a way that the casimir energy modifications remain within the experimental bounds . recently in @xcite , @xcite the same problem considered in curved spacetime in the context of a spherical symmetric solution of the hl theory . finite temperature casimir energy in spacetime ( [ eq07 ] ) has been analysed by the author in @xcite and the following weak field limit for the hl theory has been calculated : @xmath179 + [ -2\frac{\widehat{m}}{r^2}-\frac{3\widehat{m}^2}{2r^3}+\frac{\widehat{m}^2}{2\widehat{\omega } r^5}]z\ } dt^2 \\ & -\{1+[-\frac{\widehat{m}}{r}+\frac{\widehat{m}^2}{4r^2}+\frac{\widehat{m}^2}{4\widehat{\omega}r^4 } ] + [ \frac{\widehat{m}}{r^2}-\frac{\widehat{m}^2}{4r^3}-\frac{\widehat{m}^2}{4\widehat{\omega}r^5}]z \ } ( dx^2+dy^2+dz^2 ) \end{split}\end{aligned}\ ] ] where @xmath180 . this spacetime is the weak field limit of the park s spherical symmetric solution to the ir limit of the hl theory _ in the presence of a cosmological constant@xcite _ @xcite . @xmath181 may have the same role as the cosmological constant but not necessarily being a small parameter and @xmath37 is a constant frequently used to regulate the uv limit of the hl theory . comparison between ( [ eq66 ] ) and the general metric ( [ eq07 ] ) shows : @xmath182 another spacetime , which is a solution to the hl theory _ without cosmological constant _ is the kehagias - sfetsos ( ks ) solution . this spacetime has been discussed in @xcite to obtain the energy density of the apparatus and is as follows : @xmath183 where @xmath184 . @xmath185 is a radial coordinate and @xmath37 is the free parameter of the hl theory . putting @xmath186 in the park s solution recover the ks solution and so the coefficients in equation ( [ eq67 ] ) are also valid for the ks solution . we analysed the energy density of a casimir apparatus consisting of two nearby conducting parallel plates in a general weak gravitational field . the metric in the equation ( [ eq07 ] ) denotes the deviation of the weak gravitational field from flat spacetime inside the apparatus . we transformed the metric ( [ eq07 ] ) through a gauge transformation into the fermi metric and then solved the klein - gordon equation exactly . the mode frequencies were found for the scalar field inside the apparatus for both neumann and dirichlet boundary conditions in terms of the weak gravitational field parameters @xmath138 . this result was shown to be valid also for the electromagnetic field in section v. the energy density of the apparatus was found for both scalar and electromagnetic fields in terms of the weak field parameters . some examples of weak gravitational fields were analysed in section vii . specially the electromagnetic energy density and mode frequencies in the far field limit of the kerr spacetime in its equatorial plane were obtained . the weak field limit of the hoava - lifshitz gravity with a cosmological constant was also investigated and the weak field parameters were sketched . consistency of the results with the literature was checked by considering the fermi metric . in this section we find an explicit and simple asymptotic form for the wave function . as is apparent we need to have an asymptotic expansion with both argument and first parameter being large . whittaker functions have such an expansion in terms of airy functions @xcite : @xmath187\sum_{n=0}^{\infty}\frac{a_n(\zeta)}{(4\kappa)^{2n}}+\frac{ai^{'}[(4\kappa)^{\frac{2}{3}}\zeta]}{(4\kappa)^{\frac{4}{3}}}\sum_{n=0}^{\infty}\frac{b_n(\zeta)}{(4\kappa)^{2n}}\}\end{aligned}\ ] ] where @xmath188 is defined as @xmath189 and in our case @xmath190 and so @xmath191 if @xmath192 then both the argument and the first parameter go to infinity . as @xmath193 the second term in the bracket is of order @xmath194 and must be ignored to stay within the first order of expansion in @xmath111 . also the summation in the first term reduces to only @xmath195 term and @xmath196 . @xmath197\end{aligned}\ ] ] the argument of the above airy function can be written as @xmath198\}^{\frac{2}{3}}\equiv - v(x ) \;\;\ , \;\;\ v\rightarrow\infty\end{aligned}\ ] ] the airy function with large argument is as follows ( section 9.7 from @xcite ) @xmath199 as @xmath200 , the @xmath201-part and @xmath202 in the @xmath203-part must be ignored also . the observation that @xmath204 and @xmath205 moves the situation forward significantly : @xmath206 a same process can be applied to the other whittaker s function @xmath62 except that according to @xcite instead of the a argument @xmath207 in ( [ a5 ] ) we have @xmath208 . this difference changes the above process as follows : @xmath209 @xmath210 as a result the total wave function in ( [ eq22 ] ) when @xmath211 reads @xmath212 where @xmath213 and @xmath157 came from ( [ eq22 ] ) and the following relation has been used : @xmath214 in this section we find @xmath215 . other components will be find by similar techniques . we do the computations in the fermi spacetime i.e. @xmath216 . greece indices run from 0 to 3 and latins from 1 to 3 . from ( [ eq36 ] ) we have @xmath217 the first term has the following form : @xmath218 where we have used @xmath219 . the second term in ( [ b1 ] ) simplifies as follows : @xmath220b^bb^n \\ & \;\;\;\;\;\;\;\;\;\;\;= ... =-g^{ij}b^2+b^ib^j . \end{split}\end{aligned}\ ] ] based on ( [ b1])-([b3 ] ) we find @xmath221 after lowering the indices we have finally : @xmath222 now we find @xmath223 . from ( [ eq36 ] ) we have @xmath224 in which the first term already has been obtained in ( [ b2 ] ) . the second term has been also obtained in ( [ b3 ] ) and after lowering the indices again we have @xmath225 now we find @xmath223 . from ( [ eq36 ] ) we have @xmath226 from which , after lowering the indices again , we find @xmath227 the author would like to thank university of tehran for supporting this project under the grants provided by the research council . 30 k. a. milton et al . , j.phys.a 40,10935 - 10943,2007 ; s. a. fulling , k. a. milton , p. parashar , a. romeo , k.v . shajesh , j. wagner phys . d 76 , 025004 ( 2007 ) k. a. milton , k. v. shajesh , s. a. fulling , prachi parashar , phys . d 89 , 064027 ( 2014 ) . s. lamoreaux , phys . 78 , 5 ( 1996 ) ; g. bressi et al . 88 041804 ( 2002 ) . a. jaffe , phys . d 72 021301 ( 2005 ) . l. h. ford , phys . d 14 , 3304 ( 1976 ) . b. s. dewitt , phys . 19 , 295 ( 1975 ) . bezerra , h.f.mota , c.r . phys.rev.d 89,044015(2014 ) . f. sorge , phys . d 90 , 084050 ( 2014 ) . muniz , v.b . bezerra , m.s . cunha , phys . d 88 , 104035(2013 ) . muniz , v.b . bezerra , m.s . cunha , annals of physics 359 ( 2015 ) 55 - 63 . b. geyer , g. l. klimchitskaya , and v. m. mostepanenko phys . a 67 , 062102 ( 2003 ) ; m. bordag , g.l . klimchitskaya , u. mohideen , v.m . mostepanenko , advances in casimir effect , oxford university press , 2009 ; m. bordag , proceedings of the fourth workshop on quantum field theory under the influence of external conditions : the casimir effect 50 years later , world scientific , 1998 . g. bimonte et al . , phys . d , 74 ( 2006 ) 085011;phys . d 78 ( 2008 ) 024010 ; erratum ibid . d 75 , 049904(2007 ) ; erratum ibid . d 75 , 089901 ( 2007 ) ; erratum ibid . d 77 , 109903 ( 2008 ) . f. sorge , class . ( 2005 ) 5109 - 5119 . g. esposito , g. m. napolitano , l rosa , phys.rev.d 77,105011(2008 ) . g. m. napolitano , g. esposito , l rosa , phys.rev.d 78,107701(2008 ) . m. nouri - zonoz , b. nazari , phys . d 82 ( 2010 ) 044047 . borzoo nazari , m. nouri - zonoz , phys . d 85 ( 2012 ) 044060 . j. s. dowker , r. critchley , j.phy.a : math.gen . 9 535(1976 ) . d. lynden - bell , m. nouri - zonoz , rev . phys.70 , 427 ( 1998 ) l. d. landao , e. m. lifshitz , the classical theory of fields , pergamon press , 1987 ( reprint ) n.d . birrell , p.c.w . davies , quantum field theory in curved spacetime , cambridge universsity press , 1982 . p. hoava , phys . rev . d 79 , 084008 ( 2009 ) . m. park , jhep 0909 ( 2009 ) 123 . ferrari et al , mod . lett . a 29 ( 2013 ) 1350052 . misner , k. s. thorne , j. h. wheeler , gravitation , w.h . freeman and company , 1973 . p. deutsch , d. candelas , proc . a. 354 , 79 - 99 ( 1977 ) . p. deutsch , d. candelas , phys . d 20 , 3063(1979 ) 79 - 99 ( 1977 ) . r. estrada , s. a fulling , l. kaplan , k. kirsten , z. liu and k. a. milton , j. phys . a : math . theor . 41 ( 2008 ) 164055 ; r. estrada , s. a fulling , f. d mera , j. phys . a : math . theor . 45 ( 2012 ) 455402 . olde daalhuis , nist handbook of mathematical functions , cambridge university press , new york , ny , 2010 , ch.13 i.s . gradshteyn , i.m . ryzhik , table of integrals , series and products , 7th edition , elsevier inc . , 2007 milton abramowitz , irene a. stegun , handbook of mathematical functions , 10th printing , national bureau of standards us , 1972 , ch.13 w. rindler , relativity : special , gerneral and cosmological , cambridge university press , 2006 ( chapter 9 ) f. w. j. olver , asymptotics and special functions , academic press , 1974 . ( chapter 11 , ex . 7.3 ) a. erdlyi , c. a.swanson asymptotic forms of whittaker s confluent hypergeometric functions . 1957 ( 1957 ) , no . 25 , 49 pp . ( equation 10.3 ) borzoo nazari , thermal corrections to the casimir effect in a weak gravitational field , under review

there are many astrophysical systems on which the physics of neutrinos in a magnetic field plays an important role . let us recall that proto - neutron stars typically possess very strong magnetic fields . large magnetic fields @xmath3 g have been associated with the surface of supernovas @xcite and neutron stars @xcite , and fields perhaps as large as @xmath4 g with magnetars @xcite . even larger fields could exist in the star s interior . it is presumed from the scalar virial theorem @xcite that the interior field in neutron stars could be as high as @xmath5 g. a magnetic field as this ( @xmath6 g ) in the interior of a compact star will be larger in two orders than the chemical potential characterizing its quark matter density . unveiling the interconnection between the star magnetic field and its particle current flows could shed new light to the question of the star evolution . for example , it is well known that neutrinos drive supernova dynamics from beginning to end . neutrino emission and interactions play a crucial role in core collapse supernovae @xcite . their eventual emission from the proto - neutron star contains nearly all the energy released in the star explosion . neutrino luminosity , emissivity and the specific heat of the densest parts of the star are governed by charged and neutral current interactions involving matter at high densities and in the presence of strong magnetic fields . thus , a total understanding of the star cooling mechanism in a strongly magnetized medium is crucial for astrophysics . on the other hand , the explanation of large - scale magnetic fields observed in a number of galaxies , and in clusters of galaxies @xcite seems to require the existence of seed fields of primordial origin @xcite . according to several mechanisms @xcite , strong primordial fields @xmath7 g could be generated at the electroweak transition . even larger fields have been associated with superconducting magnetic strings , which would generate fields @xmath8 g in their vicinity if created after inflation @xcite . were primordial magnetic fields present in the early universe , they would have had non - trivial consequences for particle - physics cosmology . for instance , as it is well known , oscillations between neutrino flavors may change the relative abundance of neutrino species and may thereby affect primordial nucleosynthesis ( for a recent review on neutrinos in cosmology see @xcite ) . therefore , if a strong magnetic field ( @xmath9 , with @xmath10 and @xmath11 the w - boson and electron masses respectively ) modifies the neutrino energy spectrum of different flavors in different ways , a primordial magnetic field can consequently influence the oscillation process in the primeval plasma @xcite . the propagation of neutrinos in magnetized media has been previously investigated by several authors @xcite-@xcite . weak - field calculations were done for magnetized vacuum in refs . @xcite , @xcite , and at @xmath12 and @xmath13 ( @xmath14 is the electric chemical potential ) in refs . @xcite . in the @xmath13 case , as long as @xmath15 , both the field- and temperature - dependent leading contributions to the neutrino energy resulted of @xmath16-order @xcite-@xcite . in the charged plasma @xcite , the field - dependent terms were much larger , @xmath17 , but they vanished in the ( @xmath18)-limit . the weak - field results of papers @xcite-@xcite led to think that the magnetic - field effects could be significant in astrophysics , because of the field- and @xmath19-dependent terms of order @xmath20 , but were irrelevant ( @xmath21-order ) in the early universe due to its charge - symmetric character ( @xmath22 ) . hence , it has been assumed that in cosmology the main contribution to neutrino energy was the purely thermal term of order @xmath23 @xcite . however , as we will show below , a strong magnetic field ( @xmath24 ) gives rise to a new contribution to the neutrino energy that is linear in the field , independent of the chemical potential , and that is of the same order ( @xmath20 ) as the largest terms found in the weak - field charged - medium case . this new result can turn magnetic - field effects relevant for cosmology . in recent papers @xcite , @xcite , we investigated the effects of a strong magnetic field on neutrinos in magnetized vacuum ( i.e. with @xmath25 and @xmath13 ) . there , to facilitate the calculations in the strong - field limit , we extended the ritus ep - eigenfunction method of diagonalization of the green functions of spin-1/2 charged particle in electromagnetic field @xcite-@xcite , to the case of spin-1 charged particles @xcite . this formulation , which is particularly advantageous for strong - field calculations , provides an alternative method to the schwinger approach to address qft problems in electromagnetic backgrounds @xcite . the use of the ritus method resulted very convenient to study the neutrino - self - energy in magnetized media , since it allowed to diagonalize in momentum space both the electron and the w - boson green s functions in the presence of a magnetic field . ritus formalism has also been recently applied to investigate non - perturbative qft in electromagnetic backgrounds @xcite . from the above considerations it is clear that strong magnetic fields can play a significant role in a variety of astrophysical systems , and possibly also in the early universe . for these applications , the analysis has to be carried out in the presence of a medium . thus , in the present paper we extend the results obtained in papers @xcite and @xcite to include finite temperature and density , performing a detailed study of the effects of a strong magnetic field on neutrino propagation in neutral and charged media , and discussing possible applications to astrophysics and cosmology . we stress that in calculating the neutrino self - energy in a magnetized medium , we should consider , as usual , its vacuum and statistical parts . in this case the vacuum part depends on the magnetic field , and for strong fields it can make important contributions even at high temperatures @xmath26 . the reason is that the vacuum and statistical terms have different analytical behaviors , due to the lack of the statistic ultraviolet cutoff in the vacuum part . this fact gives rise to a field - dependent vacuum contribution ( @xmath20-order ) which is larger than the thermal one ( @xmath16-order ) , into the self - energy . therefore , as shown in this paper , a strong magnetic field can become more relevant than temperature for neutrino propagation in neutral media . the plan of the paper is as follows . in sec . [ selfen - gen - struc ] we consider the radiative correction to the neutrino dispersion relation in the presence of a constant magnetic field . we introduce the general covariant structure of the neutrino self - energy in the presence of an external field , and find the dispersion relation as a function of the coefficients of each independent term in the covariant structure . the general form of the found dispersion relation goes beyond any given approximation and medium characteristic and serves as a guidance for particular applications . the general expressions for the coefficients of the self - energy structures in the presence of the magnetic field are found in the one - loop approximation in sec . [ 1-loop - selfen ] . the leading behavior in the @xmath20 expansion of those coefficients are then calculated in sec . [ strong - b - selfen ] in the strong - field limit ( i.e. in the lowest landau level ( lll ) approximation ) for neutral and charged magnetized media . these results are then used in sec . [ index - ref ] to find the corresponding neutrino dispersion relations and the indexes of refraction in neutral and charged strongly magnetized media . in sec . [ numerical ] , the lll - approximation is numerically corroborated by summing in all landau levels and finding the values of the coefficients for parameter ranges corresponding to strong @xmath27 and weakly strong @xmath2 fields . possible applications to cosmology and astrophysics are discussed in sec . [ applicat ] . finally , in sec . [ conclusions ] we summarize the main outcomes of the paper and make some final remarks . the neutrino field equation of motion in a magnetized medium , including radiative corrections , is @xmath28\psi _ { l}=0 \label{1}\ ] ] where the neutrino self - energy operator @xmath29 depends on the parameters characterizing the medium , as for instance , temperature , magnetic field , particle density , etc . the operator @xmath29 is a lorentz scalar that can be formed in the spinorial space taking the contractions between the characteristic vectors and tensors of the system with all the independent elements of the dirac ring . explicit chirality reduces it to @xmath30 where @xmath31 are the chiral projector operators , and @xmath32 is a lorentz vector that can be spanned as a superposition of four basic vectors that can be formed from the characteristic tensors of the problem . in a magnetized medium , besides the neutrino four - momentum @xmath33 , we have to consider the magnetic - field strength tensor @xmath34 , to form the covariant structures @xmath35 in ( [ 3 ] ) we introduced the notations @xmath36 , and @xmath37 @xmath38 . in the covariant representation the magnetic field can be expressed as @xmath39 , where @xmath40 is the vector four - velocity of the center of mass of the magnetized medium . the presence of the magnetic field , and hence of the dimensionless magnetic field tensor @xmath41 and its dual @xmath42 , allows the covariant separation in ( [ 3 ] ) between longitudinal and transverse momentum terms that appears naturally in magnetic backgrounds @xmath43 coefficients @xmath44 , @xmath45 , @xmath46 , and @xmath47 are lorentz scalars that depend on the parameters of the theory and the approximation used . notice that finite temperature and/or density would not introduce any new term into eq . ( [ 3 ] ) . this formulation is indeed the most general one for a magnetized medium . in magnetized space , even if @xmath48 , the vector @xmath40 is present , since the presence of a constant magnetic field fixes a special lorentz frame : the rest frame ( on which @xmath49 ) where the magnetic field is defined ( @xmath50 . thus , the inclusion of a medium does not break any additional symmetry . to express @xmath51 in terms of @xmath40 only means , therefore , to make a change in the selected basis vectors when expressing @xmath51 in terms of @xmath40 in the charged medium . ] . the neutrino energy spectrum in a magnetic background can be found from eq . ( [ 1 ] ) as the nontrivial solution of @xmath52 = 0 . \label{5}\ ] ] using the covariant structure ( [ 3 ] ) , the dispersion relation ( [ 5 ] ) takes the form @xmath53 one of the main goals of this paper is to find the coefficients @xmath44 , @xmath45 , @xmath46 , and @xmath47 , in the one - loop approximation , and hence , the dispersion relations for different systems in the presence of strong magnetic - field backgrounds . let us consider the one - loop corrections to the neutrino self - energy in the presence of a constant magnetic field . to leading order in the fermi coupling constant the main field - dependent contribution to the self - energy comes from the bubble diagram ( fig.i - a ) with internal lines of virtual charged leptons and w - bosons , and from the tadpole diagram ( fig.i - b ) with virtual loop of charged leptons . in a neutral medium only the bubble diagram contributes , while in a charged medium both diagrams should be considered . however , only the bubble distinguishes between neutrino flavors , since the flavor of the internal charged lepton is directly associated to the flavor of the propagating neutrino ( i.e. the one appearing in the external legs ) . for the tadpole , because the internal boson is the neutral z - boson , the charged leptons corresponding to different families are linked to the same neutrino flavor . therefore , in regards to magnetic field effects on neutrino oscillations , the bubble contribution is the essential one . thus , because the ultimate goal of our study is to elucidate the possible effect of a strong magnetic field on neutrino flavor oscillations , henceforth we only consider the bubble diagram contribution , represented in fig.i - a , into the neutrino self - energy . the bubble contribution to the one - loop self - energy is @xmath54 where @xmath55 and @xmath56 are the green s functions of the electron and w - boson , in the presence of the magnetic field , respectively . since both virtual particles are electrically charged , the magnetic field interacts with both of them producing the landau quantization of the corresponding transverse momenta @xcite , @xcite . ( 450,80)(0,-10 ) ( 180,10)1.5 ( 120,10)1.5 ( 120,10)(180,10)(100,10)(120,10)5 ( 180,10)(200,10)5(300,50)131 ( 300,37)(300,10)35(300,37)1.5 ( 300,10)1.5 ( 280,10)(320,10)5 ( 150,0)[](@xmath44)(300,0)[](@xmath45 ) ( 150,10)(30,0,180)48.5 fig.i . one - loop contributions to the neutrino self - energy in a magnetized medium . ( a ) bubble graph - the dashed line represents the test neutrino , the solid line a charged lepton of the same family of the test neutrino , and the wiggly line the w - boson . ( b ) tadpole graph - the dashed line represents the test neutrino , the solid line a charged lepton of any species , and the zigzag line the z - boson . + the electron green s function , diagonal in momentum space at arbitrary field strength , was obtained in refs . @xcite , @xcite by ritus , using what has become an alternative method to the schwinger approach @xcite to deal with qft problems on electromagnetic backgrounds . in this method the electron green s function in configuration space is given by @xmath57 where @xmath58 , @xmath59 , and the magnetic field has been specialized in the rest frame along the @xmath60-direction ( i.e. given in the landau gauge as @xmath61 ) . the transformation functions @xmath62 in ( [ 8 ] ) play the role , in the presence of magnetic fields , of the usual fourier functions @xmath63 in the free case and are given by @xmath64 with spin matrix @xmath65 and eigenfunctions @xmath66 where @xmath67 is a normalization factor and @xmath68 denotes the parabolic cylinder functions @xcite with argument @xmath69 and positive integer index @xmath70 the integer @xmath71 in eq . ( [ 12 ] ) labels the landau levels @xmath72 . the electron momentum eigenvalue in the presence of the magnetic field , @xmath73 , is given by @xmath74 ritus technique for spin-1/2 particles was recently extended to the spin-1 particle case in refs . @xcite , @xcite . in this case the corresponding w - boson green s function can be diagonalized in momentum space as @xmath75 where @xmath76 , and the w - boson momentum eigenvalue in the presence of the magnetic field , @xmath77 , is given by the relation @xmath78 with @xmath79 denoting the w - boson landau levels . considering the mass - shell condition @xmath80 in eq . ( [ 15 ] ) , we can identify at @xmath81 the so called zero - mode problem " @xcite-@xcite . as known , at those values of the magnetic field a vacuum instability appears giving rise to w - condensation @xcite . in our calculations we restrict the magnitude of the magnetic field to @xmath82 , so to avoid the presence of any tachyonic mode . the w - boson green s function in configuration space is then @xmath83 where the transformation functions are @xmath84 ^{\gamma } \,_{\lambda } p^{-1\lambda } \,_{\mu } \label{17}\ ] ] with @xmath85 and @xmath86 ^{\gamma } \,_{\lambda } = \sum\limits_{\eta = 0,\pm 1}\mathcal{f}_{k\eta } ( x)\left [ \omega ^{(\eta ) } \right ] ^{\gamma } \,_{\lambda } \label{19}\ ] ] the matrices @xmath87 are explicitly given in terms of the w - boson spin projections @xmath88 by @xmath89 and the eigenfunctions @xmath90 are @xmath91 with @xmath92 a normalization factor similar to that appearing in ( [ 11 ] ) , and @xmath93 denoting the parabolic cylinder functions with argument @xmath94 and positive integer index @xmath95 given in terms of the w - boson landau numbers @xmath79 and spin projections @xmath88 as @xmath96 as the neutrino is an electrically neutral particle , the transformation to momentum space of its self - energy can be carried out by the usual fourier transform @xmath97 substituting with ( [ 7 ] ) , ( [ 8 ] ) and ( [ 16 ] ) in the rhs of ( [ 22 ] ) we obtain @xmath98 @xmath99 l\right\ } \label{23}\ ] ] the coefficients of the covariant expression for @xmath100 can be found using the explicit form ( [ 23 ] ) in the covariant expression @xmath101 introducing in ( [ 23 ] ) the transformation functions @xmath102 and @xmath103 , and taking into account the following properties of the spinor matrices @xmath104 @xmath105 @xmath106 where the notation @xmath107 and @xmath108 was used , we find for the coefficients appearing in ( [ 24-a ] ) the following general expressions in the one - loop approximation ( henceforth we consider @xmath109 , to simplify the notation . ) @xmath110 @xmath111 @xmath112 @xmath113 @xmath114 @xmath115 @xmath116 @xmath117 where @xmath118\varphi _ { n}(\xi /\sqrt{2})\varphi _ { l}(\rho /\sqrt{2 } ) \label{29}\ ] ] and @xmath119 are the orthonormalized harmonic oscillator wave functions , defined in terms of the hermite polynomials @xmath120 as @xmath121^{1/2}}h_{m}(x/\sqrt{2})\exp -(x^{2}/4 ) \label{30}\ ] ] expressions ( [ 25])-([28 ] ) , when substituted in ( [ 24-a ] ) , give the general formula for the one - loop neutrino self - energy in a constant magnetic field of arbitrary strength . note that , in this approach , the w - boson / magnetic - field interaction is kept in the poles of the self - energy operator through the effective momentum @xmath122 , as well as in the harmonic oscillator wave functions @xmath123 . since in the strong - field limit ( @xmath124 ) the gap between the electron landau levels is larger than the rest of the parameters entering in the electron energy , it is consistent to use the lll approximation in the electron green s function , while in the w - boson green s function , because @xmath125 we must sum in all w - boson landau levels . the neutrino self - energy in a magnetized vacuum was found within this approximation in ref . @xcite . in what follows , we extend that result to neutral and charged media , introducing finite temperature and density effects . however , since the vacuum contribution is always present in the statistical calculations of the self - energy , we summarize below the results found in ref . @xcite . in the strong field approximation ( @xmath126 ) the covariant structure of the neutrino self - energy reduces to @xmath127 that is , @xmath128 and @xmath129 are zero and @xmath130 @xmath131 after summing in @xmath79 and integrating in @xmath132 , @xmath133 , and @xmath134 we obtain @xmath135[(q_{\vert } -p_{\vert } ) ^{2}-eb+m_{w}^{2 } ] } \label{34}\ ] ] integrating in @xmath136 and considering that @xmath137 , we obtain for the neutrino self - energy in the strongly magnetized vacuum @xmath138 with coefficients given by @xmath139 notice the @xmath20-order of the leading contribution in the lll approximation . using the zero - temperature weak - field results of ref . @xcite we can identify the scalar coefficients of the general structure ( [ 3 ] ) for that case as @xmath140 if we compare the vacuum results at weak - field , eq . ( [ 36 ] ) , with those at strong field , eq . ( [ 35 ] ) , we can see that in both cases the non - zero coefficients have field - dependent leading contributions of @xmath20-order . however , as we will show in sec . [ index - ref ] , they will play very different roles in the neutrino energy spectrum , since they are associated to different self - energy structures . while the strong - field results ( [ 35 ] ) produce a magnetic - field dependence in the neutrino energy spectrum which is linear in the fermi coupling constant , the weak - field results ( [ 36 ] ) generate a smaller second - order contribution . in a neutral medium the self - energy operator can be written as @xmath141 where @xmath142 is the vacuum contribution ( given in the strong - field approximation by eqs . ( [ 34-a])-([35 ] ) ) , and @xmath143 is the statistical part which depends on temperature and magnetic field ( in a charged medium the statistical part can also depend , of course , on the chemical potential ) . our goal now is to find @xmath144 in the parameter range @xmath145 . here we can also assume that the electrons are confined to the lll . thus , performing the same calculations as in ( [ 32])-([34 ] ) we arrive to a self - energy operator @xmath146 with a structure identical to ( [ 31 ] ) , and coefficients given by an expression similar to ( [ 34 ] ) , but with the matsubara replacement , @xmath147 @xmath148 , corresponding to the discretization of the fourth component @xmath149 of the fermion momentum , obtained after introducing the wick rotation to euclidean space @xmath150 , @xmath151 being the inverse of the temperature @xmath152 . after summing in @xmath153 and extracting the vacuum part ( [ 34 ] ) , we obtain that the coefficients corresponding to the thermal contribution @xmath154 are given by @xmath155 @xmath156}f_{f}(\varepsilon _ { 1})+\frac{(q_{3}-ip_{4})(p_{4}^{2}-% \varepsilon _ { 2}^{2}+\varepsilon _ { 1}^{2})-2ip_{4}\varepsilon _ { 2}^{2}}{% \varepsilon _ { 2}[(p_{4}^{2}-\varepsilon _ { 2}^{2}+\varepsilon _ { 1}^{2})^{2}+4\varepsilon _ { 2}^{2}p_{4}^{2}]}f_{b}(\varepsilon _ { 2})\right\ } \label{40}\ ] ] where @xmath157 are the fermion and boson distribution functions respectively , with effective energies @xmath158 to do the integral in ( [ 40 ] ) we take into account the approximation @xmath159 . thus , to leading order we can neglect the contribution of the boson distribution , since it is damped by the exponential factor @xmath160 . integrating in momentum and expanding in powers of @xmath20 , we obtain , up to leading order , @xmath161 @xmath162 hence , the thermal contribution is one order smaller in powers of @xmath163 than the field - dependent vacuum contribution ( [ 35 ] ) . notice that in the strong - field approximation the thermal contribution ( [ 44 ] ) has the same @xmath16-order as the corresponding thermal contributions at weak @xcite and zero @xcite field . we should call attention to the fact that when calculating the thermal contribution to the e - w bubble by using the schwinger proper - time method @xcite , some authors @xcite have considered the following expansion @xmath164 within the w - boson green function in the presence of a constant magnetic field @xmath165 where @xmath166 is the well - known phase factor depending on the applied field @xcite , and @xmath167 is the energy - momentum transfer . based on this expansion , it has been argued that the non - local lower contribution to the bubble diagram should be of order @xmath16 , and that the local ( @xmath20)-order term only contributes at finite density . while this argument is correct when applied to the statistical part of the self - energy operator ( it explains the @xmath16 order appearing in the thermal coefficients @xmath168 and @xmath169 in ( [ 44 ] ) ) , it is not valid for the self - energy vacuum contribution . for the vacuum ( @xmath170 @xmath13 ) contribution the situation is different , since we can not neglect the internal momenta as compared with @xmath171 in the poles of the w - boson green s function , due to the lack of the ultraviolet cutoff which is present in statistics . thus , when calculating the vacuum part , the first term in ( [ 47 ] ) should be replaced by @xmath172 . such momentum dependence in the pole of the w - boson green s function makes a nontrivial contribution of ( @xmath20)-order into the non - local part of the bubble diagram . observe that the w - boson green s function obtained in our approach , eq . ( [ 14 ] ) , only has such structure indeed , and that the vacuum results at strong ( [ 35 ] ) and weak ( [ 36 ] ) fields give rise to non - local ( @xmath20)-order contributions to neutrino self - energy . we consider now a medium with a finite density of electric charge . as usual , a finite density is reflected through the shift @xmath173 in the momenta of the particles with non - zero charge associated to @xmath14 . the chemical potential , depending on the value of the charge density , plays the role of an external parameter . in stellar medium , for example , the electric charge is fixed by the net charge of the baryons , which , due to their large masses , are usually treated as a classical background . here the following comment is in order , in electroweak matter at finite density there exists the possibility to induce additional `` chemical potentials '' @xcite . these chemical potentials are nothing but dynamically generated background fields given by the average @xmath174 and @xmath175 of the zero components of the gauge fields . they are known to appear , for instance , in the presence of finite lepton / baryon density @xcite ( for recent applications related to this effect see refs . @xcite ) . in our case , a possible consequence of the condensation of such average fields could be the modification of the effective chemical potential of the w - boson , which then might be different from the electron chemical potential . nevertheless , that modification , even if present in the case under study , will have not any relevant consequence , as the w - boson contribution will be always suppressed by the factor @xmath176 . in the strong - field approximation ( i.e. after considering the lll approximation in the neutrino self - energy ( [ 24-a])-([28 ] ) ) the non - zero coefficients of the neutrino self - energy in the charged medium are given in euclidean space by @xmath177[(q_{4}^{*}-p_{4})^{2}+\varepsilon _ { 2}^{2 } ] } , \label{50}\ ] ] this expression is similar to eq . ( [ 34 ] ) after taking into account the matsubara replacement @xmath178 @xmath179 , with the additional change @xmath180 , where @xmath181 . after carrying out the temperature sum in ( [ 50 ] ) , and subtracting the vacuum part ( [ 34 ] ) , we arrive to @xmath182 @xmath183}f_{f}^{-}(\varepsilon _ { 1})+\frac{% ( q_{3}-\varepsilon _ { 1})}{\varepsilon _ { 1}[(i\varepsilon _ { 1}+p_{4})^{2}+\varepsilon _ { 2}^{2}]}f_{f}^{+}(\varepsilon _ { 1})\ ] ] @xmath184}f_{b}^{-}(\varepsilon _ { 2})+\frac{% ( q_{3}-\varepsilon _ { 2}-ip_{4})}{\varepsilon _ { 2}[(i\varepsilon _ { 2}-p_{4})^{2}+\varepsilon _ { 1}^{2}]}f_{b}^{+}(\varepsilon _ { 2})\ } \label{52}\ ] ] where @xmath185 are the fermion / anti - fermion and boson / anti - boson distributions respectively . here , we obtain an equal energy split for both distributions since both charged particles have the same electric charge and thus are characterized by the same chemical potential . assuming the approximation @xmath186 , eq . ( [ 52 ] ) is reduced in leading order to @xmath187e^{-p_{\bot } ^{2}/2eb}\delta ^{(4)}(p - p^{\prime } ) \label{54}\ ] ] where the notation @xmath188 for the electron / positron number densities in the lll @xmath189 was introduced . it is interesting to consider the situation where @xmath190 , which is a natural condition in many astrophysical environments . in this case ( [ 54 ] ) becomes @xmath191 after analytic continuation to minkowski space , we find that the leading contribution to the statistical part of the self - energy in a strongly magnetized charged medium is @xmath192 @xmath193 the apparent dependence on the momentum components @xmath194 , @xmath195 of the coefficients @xmath196 , @xmath197 is deceiving , as we can easily rewrite eqs . ( [ 57])-([58 ] ) in a more convenient way using the base formed by the four - velocity @xmath40 and the covariant magnetic field vector @xmath198 , as @xmath199 @xmath200 @xmath201from ( [ 59])-([60 ] ) we see that the leading term in the statistical part of the self - energy in the charged medium in strong magnetic field is independent of the momentum and proportional to @xmath20 , thus of larger order than the statistical contribution of the neutral case ( eq . ( [ 44 ] ) ) . comparing our result ( [ 59])-([60 ] ) with those found at weak field @xcite , we see that the coefficient of the structure @xmath202 has similar qualitative characteristics . that is , both @xmath46 coefficients are linear in @xmath14 and have the same order , @xmath20 . nevertheless , the coefficients of the structure @xmath203 significantly differ in the strong and weak cases . in the weak - field limit @xmath204 only depends on the density , thus its leading contribution is proportional to @xmath205 ( this is the characteristic term of the msw effect @xcite ) . however , for fields larger than @xmath14 , the relation @xmath206 holds , and the leading becomes proportional to @xmath207 , as in eq . ( [ 60 ] ) . thus , in the strong field case it is the field the parameter that drives the main dependence of both structures in @xmath208 . the neutrino index of refraction , defined by @xmath209 where @xmath210 and @xmath211 are the neutrino momentum and energy respectively , is a quantity that plays a significant role in neutrino flavor oscillations in a medium @xcite . a flavor - dependent index of refraction would enhance the periodical change between different flavors of neutrinos travelling through the medium . to find the index of refraction we need to know the dispersion relation @xmath212 . in the presence of a medium the energy spectra of the massless left - handed neutrinos are modified due to their weak interaction with the particles of the background . in a magnetized medium , even though the neutrinos are electrically neutral , they feel the magnetic field indirectly through their interaction with the charged particles of the medium whose properties are affected by the field . as discussed in sec.[selfen - gen - struc ] , in a magnetized medium the general form of the neutrino dispersion relation is given by eq . ( [ 5 ] ) , with the self - energy depending on the medium characteristics . as follows , we present the neutrino energy spectrum and index of refraction for magnetized neutral and charged media . in the strongly magnetized neutral medium the dispersion relation is obtained from @xmath213 = 0 , \label{62}\ ] ] with @xmath214 and @xmath215 explicitly given in eqs . ( [ 34-a])-([35 ] ) and ( [ 43])-([44 ] ) respectively . the solution of ( [ 62 ] ) in leading order in the fermi coupling constant is @xmath216 \label{63}\ ] ] in ( [ 63])@xmath217 is the angle between the direction of the neutrino momentum and that of the applied magnetic field , and the coefficient @xmath218 is explicitly given by @xmath219\exp ( -p_{\perp } ^{2}/eb ) , \label{64}\ ] ] @xmath201from ( [ 64 ] ) it is clear that in the strong - field limit the thermal correction @xmath168 to the neutrino energy is much smaller than the vacuum correction @xmath220 . as we will discuss in sec.[conclusions ] , this result can be significant for cosmological applications in case that a strong primordial magnetic field could exist in the primeval plasma . the neutrino ( @xmath221)/anti - neutrino ( @xmath222 ) energies are given by @xmath223 respectively . substituting eq . ( [ 63 ] ) into ( [ 61 ] ) we obtain for the neutrino ( antineutrino ) index of refraction in a neutral medium @xmath224 @xmath201eq . ( [ 65 ] ) implies that the index of refraction depends on the neutrino direction of motion . the order of the anisotropy is @xmath225 . maximum field effects take place for neutrinos propagating perpendicularly to the magnetic field . that is , the maximum departure of the neutrino phase velocity from the light velocity @xmath46 , occurs at @xmath226 . notice that in the neutral magnetized medium the magnetic field effect does not differentiate between neutrinos and antineutrinos . here the following comment is in order . if we consider the weak - field results ( [ 36 ] ) obtained in ref . @xcite for the neutrino self - energy , in the dispersion relation ( [ 6 ] ) we obtain @xmath227 \label{66}\ ] ] in ( [ 66 ] ) , as the energy depends on @xmath228 , we can see that the weak field produces a negligible second order correction in term of the fermi coupling constant expansion ( i.e. a @xmath16-order effect ) . it can be corroborated that the inclusion of temperature in this approximation also produces a second order correction @xcite . thus , we conclude that the strong field produces an effect qualitatively larger , which is even more important than the thermal one . we call reader s attention to the fact that the field - dependent vacuum contribution to the neutrino energy ( [ 63 ] ) has the same order in the fermi coupling constant as the ones found in a charged medium at zero @xcite and weak @xcite fields . now we consider the effect of a strong magnetic field in neutrino propagation in a charged medium . in this case the dispersion relation is @xmath229 = 0 , \label{67}\ ] ] with @xmath214 and @xmath230 given in eqs . ( [ 34-a])-([35 ] ) and ( [ 59])-([60 ] ) respectively . the solution of ( [ 67 ] ) is @xmath231p_{3}}{1 - 2a_{0}}% } . \label{67-a}\ ] ] in leading order in the fermi coupling constant , expression ( [ 67-a ] ) is approximated by @xmath232-\mathbf{\mathcal{m}.b}\pm \mathnormal{e_{0 } } \label{68}\ ] ] where @xmath233 and @xmath234 \label{70}\ ] ] in the rhs of eq . ( [ 68 ] ) , the first term is the modified neutrino kinetic energy , with the field - dependent radiative correction already found in the neutral case ( see eq . ( [ 63 ] ) ) ; the second term can be interpreted as a magnetic - moment / magnetic - field interaction energy , with @xmath235 playing the role of a neutrino effective magnetic moment ; and the third term is a rest energy induced by the magnetized charged medium . we underline that , contrary to the case of the neutrino anomalous magnetic moment found beyond the standard model by including right - chiral neutrinos @xcite , the induced magnetic moment here obtained does not require a massive neutrino . we remind that we are only considering left - chiral dirac neutrinos , so the magnetic moment in ( [ 68 ] ) can not be associated with the anomalous magnetic moment structural term @xmath236 in the self - energy . this structure is forbidden in the present case by the explicit chirality of the standard model . in the case under study we have that although the neutrino is a neutral massless particle , the charged medium can endow it , through quantum corrections , with a magnetic moment proportional to the induced neutrino rest energy . thus , we are in the presence of a peculiar magnetic moment created thanks to the particle - antiparticle unbalance of the charged medium . such a charge asymmetry permits the formation of a net virtual current that , due to the magnetic field , circulates in a plane perpendicular to the neutrino propagation and therefore produces an effective magnetic moment in the direction of the neutrino momentum . in a charged medium cp - symmetry is violated @xcite , and a common property of electroweak media with cp violation is a separation between neutrino and anti - neutrino energies . in the present case this property is manifested in the double sign of @xmath237 in ( [ 68 ] ) . although this difference depends on the magnetic field entering in @xmath238 , we stress that it depends on the magnitude of the magnetic field but not on its direction . thus , the energy anisotropy connected to the first two terms of ( [ 68 ] ) is the same for neutrinos and anti - neutrinos . the neutrino / antineutrino energy in the charged medium in the presence of a weak field has been found @xcite to be @xmath239 with @xmath240 where @xmath241 are the electron / positron number densities , and @xmath242 the electron / positron distributions . comparing the strong - field ( [ 68 ] ) and the weak - field ( [ 71 ] ) dispersion relations , we see that the field - dependent correction to the kinetic energy appearing in the strong - field approximation is absent in the leading order of the weak - field case . for a strong - field , that correction turns out to be important for neutrinos propagating perpendicular to the magnetic field . another difference between ( [ 68 ] ) and ( [ 71 ] ) , is that in the last dispersion relation the rest energy @xmath243 does not depend on the magnetic field . this is the pure medium contribution that gives rise to the well known msw effect @xcite . as discussed above , in the strong - field case , where all parameters , including @xmath14 and @xmath244 are assumed smaller than @xmath245 , such a term is negligible compared to the contribution @xmath237 which is proportional to the field . in both approximations there is an additional anisotropy related to the induced magnetic moment term , which is linear in the magnetic field and depends on the unbalance between particles and anti - particles . we should notice that this anisotropic term changes its sign when the neutrino reverses its motion . this property is crucial for a possible explanation of the peculiar high pulsar velocities @xcite . nevertheless , for the anisotropic term that modifies the kinetic energy in the strong - field approximation , the neutrino energy - momentum relation is invariant under the change of @xmath246 by @xmath247 . substituting ( [ 68 ] ) into ( [ 61 ] ) we obtain that the neutrino / anti - neutrino index of refraction in the charged medium is given by @xmath248 the result ( [ 73 ] ) implies that the index of refraction depends on the neutrino direction of motion relative to the magnetic field . moreover , neutrinos and antineutrinos have different index of refractions and therefore different phase velocities even if they move in the same direction . for neutrinos , if @xmath249 ( i.e. if there is a larger number of electrons than positrons ) the index of refraction for @xmath250 is smaller than one , so their phase velocities will be smaller than light velocity . thus , in the charged medium with strong magnetic field , neutrinos behave as massive particles . this is in agreement with the behavior they have in a dense medium , even in the absence of magnetic field @xcite . on the other hand , for anti - neutrinos the index of refraction ( the one with the positive sign in front of @xmath237 in ( [ 73 ] ) ) can be larger than one . for instance , for anti - neutrinos moving opposite to the field - line directions the index of refraction is @xmath251 thus , antineutrinos will have phase velocities larger than @xmath46 and depending on the magnetic field strength and electron density . in this regard we should mention that particles with zero rest mass can even have group velocities that exceed @xmath46 in anomalously dispersive media @xcite . moreover , in non - trivial backgrounds , particles with superluminal propagations are not intrinsically forbidden in quantum field theories . for instance , superluminal photons appears in curved spaces @xcite , casimir vacua @xcite and in qed with compactified spatial dimensions @xcite . discussions about the non - violation of microcausality by the existence of such superluminal velocities , as well as the lack of a contradiction between such a phenomenon with the bases of special theory of relativity , can be found in refs . there , it was noted that the `` front velocity , '' the one related to the index of refraction in the infinite frequency limit , is the one that can not exceed @xmath252 since it is related to the signal transmission . in the calculation of the neutrino self - energy in the strong - field limit ( @xmath253 ) done in sec . [ strong - b - selfen ] , we assumed that the main contribution to the self - energy loop diagram came from electrons in the lll , since in the strong - field limit the energy gap between electron landau levels is much larger than the electron average energy in the medium . in this section we will check the validity of these arguments with the help of numerical calculations . taking into account that the relevant physical quantity in this study is the neutrino energy , we can concentrate our analysis on the self - energy coefficients that give the largest contribution to the dispersion relation . it is easy to check , from the analytical structure of eqs . ( [ 25])-([28 ] ) , that the leading term in the expansion of the coefficients in powers of @xmath20 can be at most of order @xmath20 . keeping in mind that the only coefficients that could contribute with a @xmath20-order term in the dispersion relation ( [ 6 ] ) are @xmath44 and @xmath45 ( coefficients @xmath46 and @xmath254 appear squared in ( [ 6 ] ) and their contribution to the dispersion relation should start at least with terms of order @xmath16 ) , we can approximate the dispersion relation ( [ 6 ] ) by @xmath255 then , to validate the lll approximation we should numerically study the ratio @xmath256 where @xmath44 and @xmath45 are respectively obtained from eqs . ( [ 25 ] ) and ( [ 26 ] ) , summing in all the landau levels and taking parameter values in the strong - field region @xmath257 @xmath258 @xmath259 . the coefficients @xmath220 and @xmath260 are the corresponding values at @xmath25 in the lll . in the denominator of ( [ 74-a ] ) we are neglecting the thermal contributions @xmath168 and @xmath261 since they would make an insignificant contribution ( @xmath20-order smaller ) as compared with the vacuum ones in the parameter range @xmath262 . in sec . [ strong - b - selfen ] , we showed that @xmath263 , and that @xmath220 is given by eq . ( [ 35 ] ) . the strong - field conditions can be naturally found in many astrophysical systems like magnetars , neutron stars , etc . in cosmological applications , however , the viable primordial fields can never be much larger than the temperature , as according to the equipartition principle , the magnetic energy can only be a small fraction of the universe energy density . this argument leads to the relation @xmath264 . clearly , this is not a situation very consistent with a strong - field approximation . however , even under this condition , it is natural to expect that the thermal energy should be barely enough to induce the occupation of just a few of the lower electron landau levels , and one would not be surprised if the lll approximation still gives the leading contribution to the dispersion relation . to investigate ( [ 74-b ] ) it will be more convenient to independently study the variation range for each coefficient ratio , @xmath265 where @xmath266 ( @xmath267 ) and @xmath268 ( @xmath269 ) are respectively the vacuum and thermal contributions of each coefficient @xmath44 ( @xmath45 ) . here we subtract the ultraviolet divergent zero - field , zero - temperature parts @xmath270 @xmath271 as they can only contribute , after renormalization , with negligible zero - field terms . the validity of the lll approximation should imply that @xmath272 and @xmath273 for the range of parameters considered . in the first part of this section we examine the validity of the lll in the zero - temperature case obtaining @xmath274 , @xmath275 . in the second part , we do a similar analysis , for the thermal part of @xmath44 , finding that @xmath276 , as expected from the theoretical considerations given in sec . [ strong - b - selfen ] ( see the discussion below eq . ( [ 46 ] ) ) . as this last numerical calculation is performed for magnetic fields that can even be a few orders smaller than @xmath277 the result @xmath278 confirms the appropriateness of the lll - approximation for a rather wide range of primordial fields . the analysis of the finite temperature part of coefficient @xmath45 ( @xmath279 ) is not explicitly done in the paper , however , it is not hard to see that it gives rise to a similar result , that is , @xmath280 . if we introduce in expressions ( [ 25 ] ) and ( [ 26 ] ) the integral representations @xmath281 @xmath282 and take into account the recurrence relation @xcite @xmath283 in ( [ 26 ] ) , it is possible to rewrite coefficients @xmath243 and @xmath284 as @xmath285 @xmath286s_{\alpha } s_{\beta } \ } \label{75}\ ] ] @xmath287 @xmath288rts_{\alpha } s_{\beta } \ } \label{76}\ ] ] here , @xmath289 and @xmath290 are found doing the sum in landau levels with the help of mehler s formula @xcite @xmath291^{-1/2}\exp -(\frac{\rho ^{2}-\rho ^{\prime 2}}{2})\exp -\frac{(\rho ^{\prime } -\rho t)^{2}}{1-t^{2 } } \label{77}\ ] ] where @xmath292 , @xmath293 and @xmath294 @xmath290 is obtained from eq . ( [ 77 ] ) by replacing @xmath295 , @xmath296 and @xmath297 . integrating in @xmath298 , @xmath299 , @xmath300 and @xmath301 , and introducing the variable changes @xmath302 and @xmath303 , we find @xmath304\alpha -(k_{\vert } ^{2}+m_{w}^{2}-eb)\beta } e^{-i(\xi -\xi ^{\prime } ) p_{1}/\sqrt{eb}}\ ] ] @xmath305s_{\alpha } ^{\prime } s_{\beta } \ } \label{81}\ ] ] @xmath306\alpha -(k_{\vert } ^{2}+m_{w}^{2}-eb)\beta } \ ] ] @xmath307e^{-i(\xi -\xi ^{\prime } ) p_{1}/% \sqrt{eb}}rts_{\alpha } ^{\prime } s_{\beta } \ } \label{82}\ ] ] where @xmath308^{-1/2}\exp -(\frac{[\xi + p_{2}/\sqrt{eb% } ] ^{2}-[\xi ^{\prime } + p_{2}/\sqrt{eb}]^{2}}{2})\ ] ] @xmath309^{2}}{% 1-t^{2 } } \label{83}\ ] ] plot of @xmath310 vs. @xmath311 for field range @xmath312 and parameter values @xmath313 , @xmath314 plot of @xmath315 vs. @xmath311 for field range @xmath312 and parameter values @xmath313 , @xmath314 after wick rotating to euclidean space and doing the gaussian integrals in @xmath316 , @xmath317 , @xmath318 and @xmath319 , we finally obtain @xmath320 @xmath321 \label{84}\ ] ] @xmath322 @xmath321 \label{85}\ ] ] where we introduced the normalized parameters @xmath323 , @xmath324 , and @xmath325 . eqs . ( [ 84 ] ) and ( [ 85 ] ) are as far as we can go in the calculation of @xmath243 and @xmath128 after summing in all landau levels without using any approximation . in the zero - field limit the coefficients ( [ 84 ] ) and ( [ 85 ] ) reduce to @xmath326 where @xmath327 @xmath328 . our goal now is to investigate the validity of the lll approximation for the zero - temperature contribution of the coefficients in the parameters range : @xmath329 . with this purpose , in figs . 1 and 2 we plot , @xmath330 \label{87}\ ] ] and @xmath331 \label{88}\ ] ] respectively . here @xmath220 is the lll coefficient defined in eq . ( [ 35 ] ) . notice that primed and unprimed variables are related through : @xmath332 @xmath333 . @xmath201from fig . 1 it can be seen that @xmath334 when @xmath335 , while , from fig . 2 , we see that @xmath336 in all the range @xmath337 . this result corroborates the initial assumption that the lll approximation gives the leading contribution to the vacuum part of the neutrino self - energy at strong fields ( @xmath338 ) . let us consider now the finite temperature contributions . the finite temperature part of the coefficient @xmath243 can be found in the following way . we start from eq . ( [ 81 ] ) , perform the wick rotation to euclidean space , replace the integral in the four momentum by the temperature sum @xmath339 , ( with @xmath340 and integrate in @xmath341 to obtain @xmath342\alpha -(m_{w}^{2}-eb)\beta } e^{-i(\xi -\xi ^{\prime } ) p_{1}/\sqrt{eb}}\ ] ] @xmath343+\ ] ] @xmath344\}e^{-(\alpha + \beta ) k_{4}^{2}-2i\alpha p_{0}k_{4}}\}\ ] ] [ 90 ] after integrating in @xmath318 and @xmath319 , and introducing the elliptic theta functions representation @xcite @xmath345 \label{91}\ ] ] we find @xmath346 where @xmath347 @xmath348\alpha \exp -\widehat{m}_{w}^{2}\beta \exp -\frac{% \widehat{p}_{\bot } ^{2}}{\coth \alpha + \coth \beta } \ ] ] @xmath349+(% \widehat{p}_{0}^{2}-\widehat{p}_{3}^{2})(r^{-1}t^{-1/2}+rt^{1/2})\right]\ ] ] @xmath350 and @xmath351 @xmath352\gamma \exp -\widehat{m}_{w}^{2}(\lambda -\gamma ) \exp -\frac{% \widehat{p}_{\bot } ^{2}}{\coth \gamma + \coth ( \lambda -\gamma ) } \ ] ] @xmath353\frac{% \partial _ { \gamma } \theta _ { 3}\left ( -2\widehat{p}_{0}\widehat{t}\gamma \left/ i4\pi \widehat{t}^{2}\lambda \right . \right ) } { -2\widehat{p}_{0 } } \label{94}\ ] ] in eq . ( [ 94 ] ) we introduced the notation @xmath354 integrating ( [ 94 ] ) by parts and adding the result with ( [ 93 ] ) , we obtain @xmath355\left [ \exp -[-\widehat{p}_{0}^{2}+\widehat{m% } _ { e}^{2}]\lambda \right ] \theta _ { 3}\left ( -2\widehat{p}_{0}\widehat{t}% \lambda \left/ i4\pi \widehat{t}^{2}\lambda \right . \right)\ ] ] @xmath356\left [ \exp -\widehat{m}_{w}^{2}\lambda \right ] \theta _ { 3}\left ( 0/i4\pi \widehat{t% } ^{2}\lambda \right)\ ] ] @xmath357\alpha \exp ( \frac{\alpha ^{2}}{\alpha + \beta } ) \widehat{p}_{3}^{2}\ ] ] @xmath358 -\widehat{p}_{\vert } ^{2}\cosh ( \alpha + 2\beta ) \right.\ ] ] @xmath359 ^{2}}\right ] \left [ \widehat{p}% _ { 3}\sinh ( \alpha + 2\beta ) + \widehat{p}_{0}\cosh ( \alpha + 2\beta ) \right]\ ] ] @xmath360 \right\ } \theta _ { 3}\left ( -2\widehat{p}_{0}% \widehat{t}\alpha /\left/ i4\pi \widehat{t}^{2}(\alpha + \beta ) \right . \right ) \label{96}\ ] ] to isolate the temperature part of @xmath243 , we have to subtract the zero - temperature ( vacuum ) contribution from eq . ( [ 96 ] ) . with this aim , we use the jacobi imaginary transformation @xcite @xmath361\theta _ { 3}\left ( \frac{u}{\tau } \left/ -\frac{1}{\tau } \right . \right ) , \label{97}\ ] ] to write @xmath362}{2\widehat{t}\sqrt{\pi } \sqrt{(\alpha + \beta ) } } % \times\ ] ] @xmath363 \label{98}\ ] ] where the symbol @xmath364@xmath365 means that the term @xmath366 was taken out . then , to subtract the vacuum term from @xmath243 is equivalent to make in eq . ( [ 96 ] ) the following substitution @xmath367}{2% \widehat{t}\sqrt{\pi } \sqrt{(\alpha + \beta ) } } , \label{99}\ ] ] @xmath368}{2\widehat{t}\sqrt{\pi } \sqrt{% \lambda } } , \label{100}\ ] ] @xmath369 plot of @xmath370 vs. @xmath371 for field range @xmath372 and parameter values @xmath373 , @xmath374 , @xmath375 which leads to @xmath376\left [ \exp -[-\widehat{p}% _ { 0}^{2}+\widehat{m}_{e}^{2}]\lambda \right ] \nonumber \\ & & \left [ \theta _ { 3}\left ( -2\widehat{p}_{0}\widehat{t}\lambda / i4\pi \widehat{t}^{2}\lambda \right ) -\frac{\exp -[\widehat{p}_{0}^{2}\lambda ] } { 2% \widehat{t}\sqrt{\pi } \sqrt{\lambda } } \right ] \nonumber \\ & & -\int\limits_{0}^{\infty } d\lambda \frac{\sinh ^{-1}(\lambda ) } { \lambda } [ % \widehat{p}_{3}\sinh ( 2\lambda ) + \widehat{p}_{0}\cosh ( 2\lambda ) ] \nonumber \\ & & \times \left [ \exp -\widehat{m}_{w}^{2}\lambda \right ] \left [ \theta _ { 3}\left ( 0/i4\pi \widehat{t}^{2}\lambda \right ) -\frac{1}{2\widehat{t}% \sqrt{\pi \lambda } } \right ] \nonumber \\ & & + \int\limits_{0}^{\infty } d\alpha \int\limits_{0}^{\infty } d\beta \frac{% \sinh ^{-1}(\alpha + \beta ) } { \sqrt{\alpha + \beta } } \exp -[\widehat{m}% _ { w}^{2}\beta ] \exp -[\widehat{p}_{\vert } ^{2}+\widehat{m}_{e}^{2}]\alpha \nonumber \\ & & \times \exp ( \frac{\alpha ^{2}}{\alpha + \beta } ) \widehat{\,p}_{3}^{2}\exp -% \frac{\widehat{p}_{\bot } ^{2}}{\coth \alpha + \coth \beta } \nonumber \\ & & \times ( \left [ \frac{2\alpha \widehat{p}_{3}\,\widehat{p}_{0}}{\alpha + \beta } + 1\right ] \left [ \widehat{p}_{3}\cosh ( \alpha + 2\beta ) + \widehat{p}% _ { 0}\sinh ( \alpha + 2\beta ) \right ] -\widehat{p}_{\vert } ^{2}\cosh ( \alpha + 2\beta ) \nonumber \\ & & + \left [ ( \widehat{p}_{\vert } ^{2}+\widehat{m}_{e}^{2}-\widehat{m}_{w}^{2})-% \frac{2\alpha \widehat{p}_{3}^{2}}{\alpha + \beta } + ( \frac{1}{\sinh ^{2}\alpha } -\frac{1}{\sinh ^{2}\beta } ) \frac{\widehat{p}_{\bot } ^{2}}{% \left [ \coth \alpha + \coth \beta \right ] ^{2}}\right ] \nonumber \\ & & \times \left [ \widehat{p}_{0}\sinh ( x+2y)+\widehat{p}_{3}\cosh ( x+2y)\right ] ) \nonumber \\ & & \times \left [ \theta _ { 3}\left ( -2\widehat{p}_{0}\widehat{t}\alpha / i4\pi \widehat{t}^{2}(\alpha + \beta ) \right ) -\frac{\exp -[\widehat{p}% _ { 0}^{2}\alpha ^{2}/(\alpha + \beta ) ] } { 2\widehat{t}\sqrt{\pi ( \alpha + \beta ) % } } \right]\ } \label{102}\end{aligned}\ ] ] eq . ( [ 102 ] ) gives the thermal part of the @xmath243 taking into account all landau levels . to compare the thermal and the lll contributions we plot @xmath377 vs. @xmath378 in fig . 3 . @xmath201from fig . 3 we see that in all the range of temperatures considered , the thermal contribution is consistently smaller in several orders than the lll contribution . this result , along with those obtained in figs . 1 and 2 , imply that @xmath379 , thereby validating the lll approximation in the strong and weakly - strong field cases ( @xmath380 , @xmath381 @xmath382 ) . in addition , we would like to underline the consistency of these numerical results with the arguments used in sec . [ strong - b - selfen ] ( see discussion after eq . ( [ 44 ] ) ) , in the sense that the finite temperature part should be much smaller than the vacuum contribution . two natural environments where the results of the present paper can be of interest are , on one side , stars possessing large magnetic fields ( neutron stars , magnetars , etc ) , and , on the other , the early universe , which presumably was permeated by large primordial magnetic fields that eventually became the seeds @xcite , @xcite of the galactic fields that are observed today at scales of @xmath383 kpc @xcite and larger . the effects of primordial magnetic fields on neutrino propagation are of special interest during the neutrino decoupling era , on which a large number of these particles escaped the electroweak plasma with the possibility to develop significant oscillations . the strength of the primordial magnetic field in the neutrino decoupling era can be estimated from the constraints derived from the successful nucleosynthesis prediction of primordial @xmath384 abundance @xcite@xmath385 as well as on the neutrino mass and oscillation limits @xcite . these constraints , together with the energy equipartition principle , lead to the relations @xmath386 with @xmath387 being the muon mass . then , it is reasonable to assume that between the qcd phase transition epoch and the end of nucleosynthesis a primordial magnetic field in the above range could have been present @xcite . unlike the stellar material , whose density can be of order one , we know that the early universe was almost charge symmetric , with a particle - antiparticle asymmetry of only @xmath388 . therefore , when considering possible consequences of our results for the early universe , we should restrict the discussion to the neutral case ( @xmath13 ) . for the neutral case , it is known that at weak field all the corrections in the neutrino energy density , whether they depend on the field and/or temperature , are second order in the fermi coupling constant @xcite , @xcite , @xcite , therefore negligible small . nevertheless , it can be seen from eqs . ( [ 63])-([64 ] ) that if sufficiently strong magnetic fields were present in the primeval plasma , they would yield corrections to the energy density that are linear in the fermi coupling constant . moreover , as shown in sec . [ numerical ] , the strong - field approximation that led to these corrections is reliable even for magnetic fields in a more realistic range ( i.e. as those satisfying condition ( [ 105 ] ) ) . notice that a field satisfying ( [ 105 ] ) would be effectively strong with respect to the electron - neutrino , but weak for the remaining neutrino flavors . if such a field existed during the decoupling era , it could significantly affect the @xmath389 and @xmath390 resonant oscillations @xcite , as the field would differently modify the energy of @xmath391 compared to those of @xmath392 , @xmath393 and @xmath394 the interesting new thing here is that despite these oscillations would take place in an essentially neutral medium , because of the strong field they will be as significant as those produced by the msw mechanism @xcite in a dense medium . a peculiarity of the strong - field effects on the neutrino energy density is to give rise to anisotropic resonant oscillations . that is , the oscillation probability depends on the direction of the neutrino propagation with respect to the magnetic field . another interesting question related to primordial magnetic fields is whether they influenced neutrino propagation prior to the electroweak phase transition , since some of the mechanisms of primordial magnetic field generation allow their existence at very early epochs @xcite , @xcite . before the electroweak transition a primordial magnetic field could only exist in the form of a u(1 ) hypermagnetic field @xcite . any non - abelian `` magnetic '' field would decay at high temperatures because it would acquire a non - perturbative infrared magnetic mass @xmath395 . the implications of primordial hypermagnetic fields in neutrino propagation before the electroweak phase transition have been studied in refs . @xcite , @xcite . in @xcite the strong - field effects led to a large anisotropy in the leptons energies . the anisotropy was due to the degeneracy in the energy , which in the leading order does not depend on the transverse momentum component . it is in contrast with the magnetic field case , which takes place after the electroweak phase transition , where , although an anisotropy in the neutrino energy is also present , no degeneracy occurs ( see eqs . [ 63])-([64 ] ) ) . this different behavior is a direct consequence , on one hand , of the nonzero neutrino hypercharge and hence of the minimally coupling of these particles to the hypermagnetic field ; and on the other , of their electrical neutrality , which implies that they can interact with a magnetic field through radiative corrections only . we underline that the anisotropy in the energy spectrum of neutrinos in background fields could provide an independent way to verify the existence of primordial magnetic fields , since a field - induced anisotropy would be reflected as an imprint in a yet undetected and elusive cosmic background of neutrinos produced during the decoupling era . for astrophysical applications we should turn our attention to the star interiors characterized by high densities of charged particles under strong magnetic fields . as shown in sec . [ index - ref ] , in the charged medium case the strong - field approximation gives rise to a modification of the kinetic part of the neutrino energy ( first term in eq.([68 ] ) ) , that is not present in leading order at weak field ( eq . ( [ 71 ] ) ) @xcite . due to this new term the largest field - dependent contribution to the neutrino kinetic energy comes from neutrinos propagating perpendicularly to the field . it is worth noticing that the anisotropic term associated with the effective magnetic moment in the dispersion relations ( [ 68 ] ) and ( [ 71 ] ) changes sign when neutrinos reverse their direction of motion . on the other hand , the kinetic energy term in ( [ 68 ] ) remains unchanged when the neutrinos moves in opposite direction , even though it depends on the direction of propagation of the neutrinos with respect to the magnetic field . as known @xcite , the anisotropy associated to the magnetic - moment contribution to the neutrino energy can be relevant for a possible explanation of the peculiar high pulsar velocities . it would be interesting to consider the combined effects of the two different anisotropies to understand whether they could affect the dynamics of proto - neutron stars . another possible ground of applications of our results is supernova neutrinos . core collapse supernovae are dominated by neutrinos and their transport properties . in addition , the observation of neutrinos from supernova , which is essential to confirm the basic picture of supernova explosion , can be affected by neutrino oscillations . as we have pointed out in this paper , neutrino transport and oscillations can be both modified by the presence of a strong magnetic field . magnetic fields as strong as @xmath396 to @xmath397 g could exist in the first seconds of neutrino emission inside the supernova core @xcite . thus , the electron - neutrino energy spectrum found in this paper , within the strong - field limit for the charged medium , should be considered for any study of neutrino oscillations under those conditions . in this paper we carried out a thorough study of the propagation of neutrinos in strongly magnetized neutral and charged media . we started from the most general structure of the neutrino self - energy in a magnetic field , expressing it as the sum of four independent covariant terms with coefficients that are functions of the physical variables of the theory and whose values depend on the approximation considered . general expressions of the four coefficients at one - loop approximation were given in eqs . ( [ 25]-[28 ] ) . the coefficients were then calculated in the strong - field limit using the lll aproximation for the electrons . the lll was assumed to be valid in the parameter range @xmath329 , @xmath381 @xmath382 . to justify it one should keep in mind that under these conditions most electrons would not have enough energy to overcome the gap between the landau levels . hence , they will be mainly confined to their lower levels and the leading contribution would come from the lll . this assumption was also corroborated for the above parameters range by numerical calculations summing in all landau levels . the dispersion relation of the neutrinos was written as a function of the four coefficients of the self - energy structures , allowing in this way to straightforwardly obtain the neutrino s energy in the strong - field limit for each physical case . in concordance with results previously obtained in charged media at weak fields @xcite,@xcite , in the strong - field case an energy term associated with the interaction between the magnetic field and the effective magnetic moment was also found at leading order in @xmath398 . this interaction energy disappears in the neutral medium , since in a charged - symmetric plasma the contribution to the effective magnetic moment coming from electrons and positrons cancels out . a main outcome of our investigation was to show that in strongly magnetized systems a term of different nature emerges in both charged and neutral media . the new term , which is linear in the magnetic field and of first order in @xmath398 , enters as a correction to the neutrino kinetic energy in the presence of a strong - magnetic field . this correction is present even in a strongly magnetized vacuum , since it is related to the vacuum part ( @xmath170 @xmath13 ) of the neutrino self - energy at @xmath399 . a characteristic of the field - dependent corrections to the neutrino energy is that they produce an anisotropic index of refraction , since neutrinos moving along different directions have different field - dependent dispersion relations . we should underline that while the magnetic moment interaction term produces a maximum field effect for neutrinos propagating along the field lines , the field correction to the kinetic energy does not contribute to those propagation modes , but on the contrary , the maximum kinetic - energy effect takes place for neutrinos propagating perpendicularly to the field direction . we stress that the anisotropy does not differentiate between neutrinos and antineutrinos . the charged medium results reported in the current work can be of interest for the astrophysics of neutrinos in stars with large magnetic fields . on the other hand , our finding for the neutral medium can have applications in cosmology , if the existence of high primordial magnetic fields is finally confirmed . contrary to some authors belief @xcite,@xcite that , regardless of the field intensity , the neutrino dispersion relation in the early universe is well approximated by the dispersion relation in the zero field medium , our results indicate that strong , and even weakly strong , magnetic fields can give rise to a contribution to the neutrino energy that is several orders larger than the pure thermal contribution . the field - dependent correction to the neutrino energy in a neutral medium with strong magnetic field can have an impact in neutrino flavor - oscillations in the primeval plasma @xcite and therefore affect primordial nucleosynthesis . hence , this new effect could be important to establish possible limits to the strength of the primordial magnetic field . ee is indebted to the members of the mathematics dept . of mit , and specially to dan freedman , for the warm hospitality . ejf and vi would like to thank the institute of space studies of catalonia ( spain ) , where part of this work was accomplished , for the warm hospitality during their visit . the work of ee has been supported by dgicyt ( spain ) , contract pr-2003 - 0352 and project bfm2003 - 00620 . the work of ejf and vi was supported by nsf - grant phy-0070986 . v. ginzburg , high energy gamma ray astrophysics ( north - holland , amsterdam , 1991 ) ; d. bhattacharya and g. srinivasan , in x - ray binaries , ed . w.h.g . lewin , j. van paradijs , and e.p.j . van den heuvel , cambridge univ . press , cambridge , 1995 , pag . 495 . r. c. duncan and c. thompson , astrophys . j.*392 * ( 1992 ) l9 ; b. paczynski , acta astron.*42 * ( 1992 ) 145 ; v. v. usov , nature**357 * * ( 1992 ) 472 ; c. thompson and r. c. duncan astrophys . j.*473 * ( 1996 ) 322 ; c. kouveliotou et al . , nature * 393 * ( 1998 ) 235 ; astrophys . j.*510 * ( 1999 ) l115 . y. sofue , m. fujimoto , and r. wielebinski , ann . astron . astrophys.*24 * ( 1986 ) 459 ; p. p. kronberg , rep . phys.*57 * ( 1994 ) 325 ; r. beck et . astron . astrophys.*34 * ( 1996 ) 153 ; r. beck et . al . , ann . astron . astrophys.*34 * ( 1996 ) 153 . t. vachaspati , phys . lett . b**265 * * ( 1991 ) 258 ; k. enqvist and p. olesen , phys . lett . b**319 * * ( 1993 ) 178 ; b. cheng and a. v. olinto , phys . d**50 * * ( 1994 ) 2421 ; g. baym , d. bodeker and l. mclerran , phys . rev . d**53 * * ( 1996 ) 662 ; a. brandenburg , k. enqvist , and p. olesen , phys . rev . d**54 * * , ( 1996 ) 1291 ; phys . lett . b**392 * * ( 1997 ) 395 ; g. sigl , a. v. olinto , and k jedamzik , phys . rev . d**55 * * ( 1997 ) 4582 ; m. joyce and m. e. shaposhnikov , phys . lett.*79 * ( 1997 ) 1193 ; k. enqvist , int . d**7 * * ( 1998 ) 331 ; m. giovannini and m. e. shaposhnikov , phys . d**57 * * ( 1998 ) 2186 ; o. trnkvist , phys.rev.d**58 * * ( 1998 ) 043501 ; m. hindmarsh and a. everett , phys . rev . d**58 * * ( 1998 ) 103505 ; d. grasso and a. riotto , phys . lett . b**418 * * ( 1998 ) 258 ; j. ahonen and k. enqvist , phys . rev . d**57 * * ( 1998 ) 664 ; p. cea and l. tedesco , phys . lett . b**450 * * ( 1999 ) 61 ; int . j. mod d**12 * * ( 2003 ) 663 ; k. dimopoulos , t. prokopec , o. tornkvist and a. c. davis , phys . rev . d**65 * * ( 2002 ) 063505 ; d. boyanovsky , h. j. de vega , and m. simionato , phys . rev . d**67 * * ( 2003 ) 123505 , astro - ph/0305131 ; z. berezhiani and a. d. dolgov , astro - ph/0305595 ; a. d. dolgov , astro - ph/0306443 ; m. giovannini astro - ph/0312614 . e. j. ferrer and v. de la incera , neutrinos under strong magnetic fields . proceedings of neutrinos , flavor physics and precision cosmology , fourth tropical workshop ( aip conference proceedings v689 , pag . 372 , editor j. f. nieves and r. r. volkas ) , cairns , queensland , australia , june 2003 . v. b. semikoz and j. w. f. valle , nucl . phys . b**425 * * ( 1994 ) 651 , p. elmfors , d. grasso and g. raffelt , nucl . phys . b**479 * * ( 1996 ) 3 ; a. erdas , c. w. kim and t. h. lee , phys . d**58 * * ( 1998 ) 085016 ; d. grasso , nucl . b ( proc . suppl . ) * 70 * ( 1999 ) 267 ; a. erdas , and c. isola , phys . lett . b**494 * * ( 2000 ) 262 . e. j. ferrer , in proceedings of the international conference quantization , gauge theory , and strings , 5 - 10 jun 2000 , moscow , russia ( scientif world publi . , singapore 2001 , v. ii , pag.301 , editors a. semikhatov , m. vasiliev and v. zaikin ) ; e. elizalde , e. j. ferrer and v. de la incera , neutrino dispersion in intense magnetic field . proceedings of particle physics and cosmology second tropical workshop ( aip conference proceedings v540 , pag . 372 , editor j. f. nieves ) , san juan , puerto rico , apr . 2000 . v. i. ritus , in issues in intense - field quantum electrodynamics ( v.l . ginzburg , ed . ) , proceedings of p.n . lebedev physical institute , vol . 111 , p. 5 , moscow , nauka , 1979 . * 168 * , nova science , commack , 1987 ) . c. n. leung , y. j. ng , and a. w. ackley , phys . d**54 * * ( 1996 ) 4181 ; d. -s . lee , c. n. leung , y. j. ng , phys . rev.*55 * ( 1997 ) 6504 ; e. j. ferrer and v. de la incera , phys . rev d**58 * * ( 1998 ) 065008 ; int . a * 14 * ( 1999 ) 3963 ; phys . b**481**(2000 ) 287 ; e. elizalde , e. j. ferrer and v. de la incera , phys . rev . d**68 * * ( 2003 ) 096004 . m. j. duff and m. ramon - medrano , phys . rev . d**12 * * ( 1975 ) 3357 ; g. k. savvidy , phys . lett b**71 * * ( 1977 ) 133 ; s. g. matinyan and g. k. savvidy , nucl . phys . b**134 * * ( 1978 ) 539 ; h. b. nielsen and p. olesen , nucl . phys . b**144 * * ( 1978 ) 376 ; v. v. skalozub , sov . j. nucl . phys * 43 * ( 1986 ) 665 . for a discussion on the relation between the index of refraction and the different velocity notions see : l. brillouin , wave propagation and group velocity , academic press , new york , 1960 ; r. l. smith , am . j. phys . * 38 * ( 1970 ) 978 ; a. kuzmich , et al , phys . rev . lett.*86 * ( 2001 ) 3925 . j. l. latorre , p. pascual , and r. tarrach , nucl . phys . b**437 * * ( 1995 ) 60 , g. m. shore , nucl . phys . b**460 * * ( 1996 ) 379 ; k. scharnhorst , ann . phys . * 7 * ( 1998 ) 700 ; w. dittrich and h. gies , phys . rev . d**58 * * ( 1998 ) 025004 ; a. d. dolgov and i. d. novikov , phys . lett . b**442 * * ( 1998 ) 82 . j. cannellos , e. j. ferrer and v. de la incera , phys . b**542 * * ( 2002 ) 123 ; j. cannellos , e. j. ferrer and v. de la incera , self energy of chiral leptons in a background hypermagnetic field . proceedings of mrst-2002 , perimeter institute , waterloo , canada , may 15 - 17 , 2002 ( aip conf . ( 2003 ) 117 ) .

the recent measurements of the wilkinson microwave anisotropy probe ( wmap ) satellite @xcite on the cosmic microwave background radiation ( cmbr ) have shown that the matter abundance in the universe is @xmath2 , where @xmath3 with @xmath4 being the energy density of the @xmath5-th species and @xmath6 the critical energy density of the universe and @xmath7 is the present value of the hubble parameter in units of @xmath8 . the baryon abundance is also found by these measurements to be @xmath9 . consequently , the cold dark matter ( cdm ) abundance in the universe is @xmath10 . the @xmath11 confidence level ( c.l . ) range of this quantity is then , roughly , @xmath12 . taking @xmath13 , which is its best - fit value from the hubble space telescope ( hst ) @xcite , and assuming that the total energy density of the universe is very close to its critical energy density ( i.e. @xmath14 ) , as implied by wmap , we conclude that about @xmath15 of the energy density of the present universe consists of cdm . the question then is , what the nature , origin , and composition of this important component of our universe is . particle physics provides us with a number of candidate particles out of which cdm can be made . these particles appear naturally in various particle physics frameworks for reasons completely independent from cdm considerations and are , certainly , not invented for the sole purpose of explaining the presence of cdm in the universe . the basic properties that such a candidate particle must satisfy are the following : ( @xmath5 ) it must be stable or very long - lived , which can be achieved by an appropriate symmetry , ( @xmath16 ) it should be electrically and color neutral , as implied by astrophysical constraints on exotic relics ( like anomalous nuclei ) , but can be interacting weakly , and ( @xmath17 ) it has to be non - relativistic , which is usually guaranteed by assuming that it is adequately massive , although even very light particles such as axions can be non - relativistic for different reasons . so , what we need as constituent of cdm is a weakly interacting massive particle . there are several possibilities , but we will concentrate here on the major particle physics candidates which are the axion , the lightest neutralino , the axino , and the gravitino ( for other candidates , see e.g. ref . note that the last three particles exist only in supersymmetric ( susy ) theories . in sec . [ sec : axion ] , we examine the possibility that the axions are constituents of cdm . [ sec : mssm ] is devoted to outlining the salient features of the minimal supersymmetric standard model ( mssm ) , which will be used as a basic frame for discussing susy cdm . in sec . [ sec : relic ] , we summarize the calculation of the relic abundance of the lightest neutralino , which is normally the lightest supersymmetric particle ( lsp ) , and investigate the circumstances under which it can account for the cdm in the universe . in secs . [ sec : axino ] and [ sec : gravitino ] , we discuss , respectively , axinos and gravitinos as constituents of cdm . in sec . [ sec : quasi ] , we present a susy grand unified theory ( gut ) model which solves the bottom - quark mass problem by naturally and modestly violating the exact unification of the third generation yukawa couplings . we study the parameter space of the model which is allowed by neutralino dark matter considerations as well as some other phenomenological constraints . finally , in sec . [ sec : concl ] , we summarize our conclusions . the most natural solution to the strong @xmath18 problem ( i.e. the apparent absence of @xmath18 violation in strong interactions implied by the experimental bound on the electric dipole moment of the neutron ) is the one provided by a peccei - quinn ( pq ) symmetry @xcite . this is a global @xmath19 symmetry , @xmath20 , which carries qcd anomalies and is spontaneously broken at a scale @xmath21 , the so - called axion decay constant or simply pq scale . astrophysical @xcite and cosmological constraints imply that @xmath22 . the upper bound originates @xcite from the requirement that the relic energy density of axions does not overclose the universe . it should be noted , however , that this upper bound can be considerably relaxed if the axions are diluted @xcite by entropy generation after their production at the qcd phase transition ( for more recent applications of this dilution mechanism , see e.g. ref . @xcite ) . the axion is a pseudo nambu - goldstone boson corresponding to the phase of the complex pq field , which breaks @xmath20 by its vacuum expectation value ( vev ) . after the end of inflation @xcite , this phase appears homogenized over the universe ( supposing that the pq field is non - zero ) with a value @xmath23 , which is known as the initial misalignment angle . naturalness suggests that @xmath23 is of order unity . this angle remains frozen until the qcd phase transition , where the qcd instantons come into play . they break explicitly the pq symmetry to a discrete subgroup @xcite since this symmetry carries qcd anomalies . so , a sinusoidal potential for the phase of the pq field is generated and this phase starts oscillating coherently about a minimum of the potential . the resulting state resembles pressureless matter consisting of static axions with mass @xmath24 , where @xmath25 is the qcd scale . for @xmath26 , the mass of the axion @xmath27 . note that axions , although very light , are good candidates for being constituents of the cdm in the universe since they are produced at rest . also , they are very weakly interacting since their interactions are suppressed by the axion decay constant @xmath21 . the relic abundance of axions can be calculated by using the formulae of ref . @xcite , where we take the qcd scale @xmath28 and ignore the uncertainties for simplicity . we find @xmath29 ( note that a primordial magnetic helicity , may @xcite influence this abundance ) . so , for natural values of @xmath30 and @xmath26 , axions can contribute significantly to cdm , which can even consist solely of axions . the main disadvantage of axionic dark matter is that it leads to isocurvature perturbations if the pq field emerges with non - zero ( homogeneous ) value at the end of inflation . indeed , during inflation , the angle @xmath23 acquires a superhorizon spectrum of perturbations as all the almost massless degrees of freedom . at the qcd phase transition , these perturbations turn into isocurvature perturbations in the axion energy density , which means that the partial curvature perturbation in axions is different than the one in photons . the recent results of wmap @xcite put stringent bounds @xcite on the possible isocurvature perturbation . so , a large axion contribution to cdm is disfavored in models where the inflationary scale is superheavy ( i.e. of the order of the susy gut scale ) and the pq field is non - zero at the end of inflation . we now wish to turn to the discussion of the main susy candidates for dark matter : the lightest neutralino @xmath31 , the axino @xmath32 and the gravitino @xmath33 . we will consider them mainly within the simplest susy framework , which is the mssm . it is , thus , important to first outline the basics of mssm . we consider the mssm embedded in some general susy gut model . we further assume that the gut gauge group breaking down to the standard model ( sm ) gauge group @xmath34 occurs in one step at a scale @xmath35 , where the gauge coupling constants of strong , weak , and electromagnetic interactions unify . ignoring the yukawa couplings of the first and second generation , the effective superpotential below @xmath36 is @xmath37 where @xmath38 and @xmath39 are the quark and lepton @xmath40 doublet left handed superfields of the third generation and @xmath41 , @xmath42 , and @xmath43 the corresponding @xmath40 singlets . also , @xmath44 , @xmath45 are the electroweak higgs superfields and @xmath46 the @xmath47 antisymmetric matrix with @xmath48 . the gravity - mediated soft susy - breaking terms in the scalar potential are given by @xmath49 @xmath50 , \label{vsoft}\ ] ] where the sum is taken over all the complex scalar fields @xmath51 and tildes denote superpartners . the soft gaugino mass terms in the lagrangian are @xmath52 where @xmath53 , @xmath54 and @xmath55 are the bino , winos and gluinos respectively . the lagrangian of mssm is invariant under a discrete @xmath56 matter parity symmetry under which all `` matter '' ( i.e. quark and lepton ) superfields change sign . combining this symmetry with the @xmath56 fermion number symmetry under which all fermions change sign , we obtain the discrete @xmath56 r - parity symmetry under which all sm particles are even , while all sparticles are odd . by virtue of r - parity conservation , the lsp is stable and , thus , can contribute to the cdm in the universe . it is important to note that matter parity is vital for mssm to avoid baryon- and lepton - number - violating renormalizable couplings in the superpotential , which would lead to highly undesirable phenomena such as very fast proton decay . so , the possibility of having the lsp as cdm candidate is not put in by hand , but arises naturally from the very structure of mssm . the susy - breaking parameters @xmath57 , and @xmath58 ( @xmath59 ) are all of the order of the soft susy - breaking scale @xmath60 , but are otherwise unrelated in the general case . however , if we assume that soft susy breaking is mediated by minimal supergravity ( msugra ) , i.e. supergravity with minimal khler potential , we obtain soft terms which are universal `` asymptotically '' ( i.e. at @xmath36 ) . in particular , we obtain a common scalar mass @xmath61 , a common trilinear scalar coupling @xmath62 , and a common gaugino mass @xmath63 . the mssm supplemented by universal boundary conditions is known as constrained mssm ( cmssm ) @xcite . it is true that msugra implies two more asymptotic relations : @xmath64 and @xmath65 , where @xmath66 and @xmath67 is the ( asymptotic ) gravitino mass . these extra conditions are usually not included in the cmssm . imposing them , we get the so - called very cmssm @xcite , which is a very restrictive version of mssm and will not be considered in these lectures . the cmssm can be further restricted by imposing asymptotic yukawa unification ( yu ) @xcite , i.e. the equality of all three yukawa coupling constants of the third family at @xmath36 : @xmath68 exact yu , which makes the cmssm considerably more predictive , can be obtained in guts based on a gauge group such as @xmath69 or @xmath70 under which all the particles of one family belong to a single representation with the additional requirement that the masses of the third family fermions arise primarily from their unique yukawa coupling to a single superfield representation which predominantly contains the electroweak higgs superfields . it should be noted that exact yu in the cmssm leads to unacceptable values for the bottom - quark mass @xmath71 and , thus , must be corrected in order to become consistent with observations . we will ignore this problem for the moment , but we will return to it in sec . [ sec : quasi ] . now , we assume that our effective theory below @xmath36 is the cmssm with yu . this theory depends on the following parameters ( @xmath72 ) : @xmath73 where @xmath74 with @xmath75 being the gut gauge coupling constant and @xmath76 is the ratio of the two electroweak vevs . the parameters @xmath77 and @xmath36 are evaluated consistently with the experimental values of the electromagnetic and strong fine - structure constants @xmath78 and @xmath79 , and the sine - squared of the weinberg angle @xmath80 at @xmath81 . to this end , we integrate @xcite numerically the renormalization group equations ( rges ) for the mssm at two loops in the gauge and yukawa coupling constants from @xmath36 down to a common but variable @xcite susy threshold @xmath82 ( @xmath83 are the stop - quark mass eigenstates ) . from @xmath84 to @xmath81 , the rges of the non - susy sm are used . the set of rges needed for our computation can be found in many references ( see e.g. ref . we take @xmath85 which , as it turns out , leads to gauge coupling unification at @xmath36 with an accuracy better than 0.1@xmath86 . so , we can assume exact unification once the appropriate susy particle thresholds are taken into account . the unified third generation yukawa coupling constant @xmath87 at @xmath36 and the value of @xmath88 at @xmath84 are estimated using the experimental inputs for the top - quark mass @xmath89 and the @xmath90-lepton mass @xmath91 . our integration procedure of the rges relies @xcite on iterative runs of these equations from @xmath36 to low energies and back for every set of values of the input parameters until agreement with the experimental data is achieved . the susy corrections to @xmath92 are taken from ref . @xcite and incorporated at @xmath84 . assuming radiative electroweak symmetry breaking , we can express the values of the parameters @xmath93 ( up to its sign ) and @xmath94 ( or , equivalently , the mass @xmath95 of the @xmath18-odd neutral higgs boson @xmath96 ) at @xmath84 in terms of the other input parameters by minimizing the tree - level renormalization group ( rg ) improved potential @xcite at @xmath84 . the resulting conditions are @xmath97 where @xmath98 , @xmath99 are the soft susy - breaking scalar higgs masses . we can improve the accuracy of these conditions by including the full one - loop radiative corrections to the potential from ref . @xcite at @xmath84 . we find that the corrections to @xmath93 and @xmath95 from the full one - loop effective potential are minimized @xcite by our choice of @xmath84 . so , a much better accuracy is achieved by using this variable susy threshold rather than a fixed one . furthermore , we include in our calculation the two - loop radiative corrections to the masses @xmath100 and @xmath101 of the @xmath18-even neutral higgs bosons @xmath7 and @xmath102 . these corrections are particularly important for the mass of the lightest @xmath18-even neutral higgs boson @xmath7 . finally , the susy corrections to @xmath71 are also included at @xmath84 using the relevant formulae of ref . @xcite . as already mentioned , the predicted value of the bottom - quark mass is not compatible with experiment . however , we will ignore this problem for the moment . the sign of @xmath93 is taken to be positive , since the @xmath103 case is excluded because it leads @xcite to a neutralino relic abundance which is well above unity , thereby overclosing the universe , for all @xmath95 s permitted by @xmath104 . we are left with @xmath105 and @xmath62 as free input parameters . the lsp is the lightest neutralino @xmath106 . the mass matrix for the four neutralinos is @xmath107 in the @xmath108 basis . here , @xmath109 , @xmath110 , and @xmath111 , @xmath112 are the mass parameters of @xmath53 , @xmath113 in eq . ( [ gaugino ] ) . in cmssm , the lightest neutralino turns out to be an almost pure bino @xmath53 . the lsps are stable due to the presence of the unbroken r - parity , but can annihilate in pairs since this symmetry is a discrete @xmath56 symmetry . this reduces their relic abundance in the universe . if there exist sparticles with masses close to the mass of the lsp , their coannihilation @xcite with the lsp leads to a further reduction of the lsp relic abundance . it should be noted that the number density of these sparticles is not boltzmann suppressed relative to the lsp number density . they eventually decay yielding an equal number of lsps and , thus , contributing to the relic abundance of the lsps . of particular importance is the next - to - lsp ( nlsp ) , which , in cmssm , turns out to be the lightest stau mass eigenstate @xmath114 . its mass is obtained by diagonalizing the stau mass - squared matrix @xmath115 in the gauge basis ( @xmath116 ) . here , @xmath117}}$ ] is the soft susy - breaking mass of the left [ right ] handed stau @xmath118}$ ] and @xmath119 the tau - lepton mass . the stau mass eigenstates are @xmath120 where @xmath121 is the @xmath122 mixing angle . the large values of @xmath0 and @xmath90 yukawa coupling constants implied by yu cause soft susy - breaking masses of the third generation squarks and sleptons to run ( at low energies ) to lower physical values than the corresponding masses of the first and second generation . furthermore , the large values of @xmath88 implied by yu lead to large off - diagonal mixings in the sbottom and stau mass - squared matrices . these effects reduce further the physical mass of the lightest stau , which is the nlsp . another effect of the large values of the @xmath0 and @xmath90 yukawa coupling constants is the reduction of the mass @xmath95 of the @xmath18-odd neutral higgs boson @xmath96 and , consequently , the other higgs boson masses to smaller values . we now turn to the calculation of the cosmological relic abundance of the lightest neutralino @xmath31 ( almost pure @xmath53 ) in the cmssm with yu according to the standard cosmological scenario ( for non - standard scenaria , see e.g. ref . @xcite ) . in general , all sparticles contribute to @xmath123 , since they eventually turn into lsps , and all the ( co)annihilation processes must be considered . the most important contributions , however , come from the lsp and the nlsp . so , in the case of the cmssm , we should concentrate on @xmath31 ( lsp ) and @xmath124 ( nlsp ) and consider the coannihilation of @xmath31 with @xmath114 and @xmath125 . the important role of the coannihilation of the lsp with sparticles carrying masses close to its mass in the calculation of the lsp relic abundance has been pointed out by many authors ( see e.g. refs . @xcite ) . here , we will use the method of ref . @xcite , which was also used in ref . note that our analysis can be readily applied to any mssm scheme where the lsp and nlsp are the bino and stau respectively . in particular , it applies to the cmssm without yu , where we have @xmath88 as an extra free input parameter . the relevant quantity , in our case , is the total number density @xmath126 since the @xmath114 s and @xmath125 s decay into @xmath31 s after freeze - out . at cosmic temperatures relevant for freeze - out , the scattering rates of these ( non - relativistic ) sparticles off particles in the thermal bath are much faster than their annihilation rates since the ( relativistic ) particles in the bath are considerably more abundant . consequently , the number densities @xmath127 ( @xmath128 , @xmath114 , @xmath125 ) are proportional to their equilibrium values @xmath129 to a good approximation , i.e. @xmath130 . the boltzmann equation ( see e.g. ref . @xcite ) is then written as @xmath131 where @xmath102 is the hubble parameter , @xmath132 is the `` relative velocity '' of the annihilating particles , @xmath133 denotes thermal averaging and @xmath134 is the effective cross section defined by @xmath135 with @xmath136 being the total cross section for particle @xmath5 to annihilate with particle @xmath137 averaged over initial spin states . in our case , @xmath134 takes the following form @xmath138 for @xmath139 , we use the non - relativistic approximation @xmath140 @xmath141 here @xmath142 , 1 , 1 ( @xmath128 , @xmath114 , @xmath125 ) is the number of degrees of freedom of the @xmath5-th particle with mass @xmath143 and @xmath144 with @xmath145 being the photon temperature . using boltzmann equation ( which is depicted in eq . ( [ boltzmann ] ) ) , we can calculate the relic abundance of the lsp at the present cosmic time . it has been found @xcite to be given by @xmath146 with @xmath147 here @xmath148 is the planck scale , @xmath149 is the effective number of massless degrees of freedom at freeze - out @xcite and @xmath150 with @xmath151 being the freeze - out photon temperature calculated by solving iteratively the equation @xcite @xmath152 the constant @xmath153 is chosen to be equal to @xmath154 @xcite . the freeze - out temperatures which we obtain here are of the order of @xmath155 and , thus , our non - relativistic approximation ( see eq . ( [ ri ] ) ) is _ a posteriori _ justified . away from s - channel poles and final - state thresholds , the quantities @xmath156 are well approximated by applying the non - relativistic taylor expansion up to second order in the relative velocity @xmath132 : @xmath157 actually , this corresponds @xcite to an expansion in s and p waves . the thermally averaged cross sections are then easily calculated @xmath158 here , we approximated the masses of the incoming particles by the neutralino mass , i.e. @xmath159 . the reduced mass of the incoming particles is then equal to @xmath160 . we also thermally averaged over the relative velocity rather than the separate velocities of the incoming particles , which would be more accurate . using eqs . ( [ sigmaeff1 ] ) , ( [ sigmaeff2 ] ) , ( [ sigmaeff3 ] ) , and ( [ average ] ) , one obtains @xmath161 where we sum over @xmath162 , @xmath163 , and @xmath164 with @xmath165 , @xmath166 , and @xmath167 , @xmath168 given by @xmath169 here @xmath170 , 4 , 2 for @xmath171 , @xmath163 , and @xmath172 respectively . .feynman diagrams [ cols="^,^,^ " , ] [ tab : charges ] a moderate violation of exact yu can be naturally accommodated in this model by adding a new higgs superfield @xmath173 with yukawa couplings @xmath174 . actually , ( * 15,2,2 * ) is the only representation of @xmath175 , besides ( * 1,2,2 * ) , which possesses such couplings to the matter superfields . in order to give superheavy masses to the color non - singlet components of @xmath173 , we need to include one more higgs superfield @xmath176 with the superpotential coupling @xmath177 , whose coefficient is of the order of @xmath36 . after the breaking of @xmath175 to @xmath34 , the two color singlet @xmath178 doublets @xmath179 , @xmath180 contained in @xmath173 can mix with the corresponding doublets @xmath181 , @xmath182 in @xmath7 . this is mainly due to the terms @xmath177 and @xmath183 . actually , since @xmath184 there are two independent couplings of the type @xmath183 ( both suppressed by the string scale @xmath185 , as they are non - renormalizable ) . one of these couplings is between the @xmath186 singlets in @xmath187 and @xmath188 and the other between the @xmath186 triplets in these combinations . so , we obtain two bilinear terms @xmath189 and @xmath190 with different coefficients , which are suppressed by @xmath191 . these terms together with the terms @xmath192 and @xmath193 from @xmath177 , which have equal coefficients , generate different mixings between @xmath181 , @xmath179 and @xmath182 , @xmath180 . consequently , the resulting electroweak doublets @xmath44 , @xmath45 contain @xmath194 violating components suppressed by @xmath191 and fail to form a @xmath186 doublet by an equally suppressed amount . so , yu is naturally and moderately violated . unfortunately , as it turns out , this violation is not adequately large for correcting the bottom - quark mass within the framework of the cmssm with @xmath1 . in order to allow for a more sizable violation of yu , we further extend the model by including the superfield @xmath195 with the coupling @xmath196 . to give superheavy masses to the color non - singlets in @xmath195 , we introduce one more superfield @xmath197 with the coupling @xmath198 , whose coefficient is of order @xmath36 . the superpotential terms @xmath198 and @xmath199 imply that , after the breaking of @xmath175 to @xmath34 , @xmath195 acquires a vev of order @xmath36 . the coupling @xmath196 then generates @xmath186 violating unsuppressed bilinear terms between the doublets in @xmath176 and @xmath7 . these terms can overshadow the corresponding ones from the non - renormalizable term @xmath183 . the resulting @xmath186 violating mixing of the doublets in @xmath7 and @xmath173 is then unsuppressed and we can obtain stronger violation of yu . to further analyze the mixing of the doublets in @xmath7 and @xmath173 , observe that the part of the superpotential corresponding to the symbolic couplings @xmath177 , @xmath196 is properly written as @xmath200 where @xmath201 is a mass parameter of order @xmath36 , @xmath202 is a dimensionless parameter of order unity , @xmath203 denotes trace taken with respect to the @xmath204 and @xmath178 indices , and the superscript t denotes the transpose of a matrix . after the breaking of @xmath175 to @xmath34 , @xmath195 acquires a vev @xmath205 . substituting it by this vev in the above couplings , we obtain @xmath206 where the ellipsis in eq . ( [ mass ] ) contains the colored components of @xmath176 , @xmath173 and @xmath207 . inserting eqs . ( [ mass ] ) and ( [ triplet ] ) into eq . ( [ expmix ] ) , we obtain @xmath208 so , we get two pairs of superheavy doublets with mass @xmath201 . they are predominantly given by @xmath209 the orthogonal combinations of @xmath181 , @xmath210 and @xmath182 , @xmath211 constitute the electroweak doublets @xmath212 the superheavy doublets in eq . ( [ superdoublets ] ) must have vanishing vevs , which readily implies that @xmath213 and @xmath214 . equation ( [ ew ] ) then gives @xmath215 , @xmath216 . from the third generation yukawa couplings @xmath217 , @xmath218 , we obtain @xmath219 where @xmath220 . from eqs . ( [ top ] ) and ( [ bottomtau ] ) , we see that yu is now replaced by the yquc @xmath221 for simplicity , we restricted ourselves here to real values of @xmath153 only which lie between zero and unity , although @xmath153 is , in general , an arbitrary complex quantity with @xmath222 . below the gut scale @xmath36 , the particle content of our model reduces to this of mssm ( modulo sm singlets ) . we assume universal soft susy breaking scalar masses @xmath61 , gaugino masses @xmath63 , and trilinear scalar couplings @xmath62 at @xmath36 . therefore , the resulting mssm is the so - called cmssm @xcite with @xmath1 supplemented by the yquc in eq . ( [ minimal ] ) . with these initial conditions , we run the mssm rges @xcite between @xmath36 and a common variable susy threshold @xmath84 ( see sec . [ sec : mssm ] ) determined in consistency with the susy spectrum of the model . at @xmath84 , we impose radiative electroweak symmetry breaking , evaluate the susy spectrum and incorporate the susy corrections @xcite to the @xmath0-quark and @xmath90-lepton masses . note that the corrections to the @xmath90-lepton mass ( almost 4@xmath86 ) lead @xcite to a small reduction of @xmath88 . from @xmath84 to @xmath81 , the running of gauge and yukawa coupling constants is continued using the sm rges . for presentation purposes , @xmath63 and @xmath61 can be replaced @xcite by the lsp mass @xmath223 and the relative mass splitting between this particle and the lightest stau @xmath224 ( recall that @xmath124 is the nlsp in this case ) . for simplicity , we restrict this presentation to the @xmath225 case ( for @xmath226 see refs . so , our input parameters are @xmath223 , @xmath227 , @xmath153 , and @xmath88 . for any given @xmath228 in the range in eq . ( [ mbrg ] ) and with fixed @xmath89 and @xmath229 , we can determine the parameters @xmath153 and @xmath88 at @xmath84 so that the yquc in eq . ( [ minimal ] ) is satisfied . we are , thus , left with @xmath223 and @xmath227 as free parameters . restrictions on the parameters of our model can be derived by imposing a number of cosmological and phenomenological requirements ( for similar recent analyses , see refs . these constraints result from @xmath230 _ cdm considerations_. as discussed in sec . [ sec : mssm ] , in the context of the cmssm , the lsp can be the lightest neutralino which is an almost pure bino . it naturally arises @xcite as a cdm candidate . we require its relic abundance , @xmath231 , not to exceed the @xmath11 c.l . upper bound on the cdm abundance derived @xcite by wmap : @xmath232 we calculate @xmath231 using micromegas @xcite , which is certainly one of the most complete publicly available codes . among other things , it includes all possible coannihilation processes @xcite and one - loop qcd corrections @xcite to the higgs decay widths and couplings to fermions . @xmath230 _ branching ratio of @xmath104_. taking into account the experimental results of ref . @xcite on this ratio , @xmath233 , and combining @xcite appropriately the experimental and theoretical errors involved , we obtain the @xmath11 c.l . range @xmath234 although there exist more recent experimental data @xcite on the branching ratio of @xmath104 , we do not use them here . the reason is that these data do not separate the theoretical errors from the experimental ones and , thus , the derivation of the @xmath11 c.l . range is quite ambiguous . in any case , the @xmath11 c.l . limits obtained in ref . @xcite on the basis of these latest measurements are not terribly different from the ones quoted in eq . ( [ bsgb ] ) . in view of this and the fact that , in our case , the restrictions from @xmath233 are overshadowed by other constraints ( see sec . [ sec : parameters ] ) , we limit ourselves to the older data . we compute @xmath233 by using an updated version of the relevant calculation contained in the micromegas package @xcite . in this code , the sm contribution is calculated following ref . the charged higgs ( @xmath235 ) contribution is evaluated by including the next - to - leading order ( nlo ) qcd corrections @xcite and @xmath88 enhanced contributions @xcite . the dominant susy contribution includes resummed nlo susy qcd corrections @xcite , which hold for large @xmath88 . @xmath230 _ muon anomalous magnetic moment_. the deviation , @xmath236 , of the measured value of @xmath237 from its predicted value in the sm , @xmath238 , can be attributed to susy contributions , which are calculated by using the micromegas routine @xcite . the calculation of @xmath238 is not yet stabilized mainly because of the instability of the hadronic vacuum polarization contribution . according to recent calculations ( see e.g. refs . @xcite ) , there is still a considerable discrepancy between the findings based on the @xmath239 annihilation data and the ones based on the @xmath90-decay data . taking into account the results of ref . @xcite and the experimental measurement of @xmath237 reported in ref . @xcite , we get the following @xmath11 c.l . ranges : @xmath240 following the common practice @xcite , we adopt the restrictions to parameters induced by eq . ( [ g2e ] ) , since eq . ( [ g2 t ] ) is considered as quite oracular , due to poor @xmath90-decay data . it is true that there exist more recent experimental data @xcite on @xmath237 than the ones we considered and more updated estimates of @xmath236 than the one in ref . @xcite ( see e.g. ref . @xcite ) . however , only the @xmath11 c.l . upper limit on @xmath236 enters into our analysis here and its new values are not very different from the one in eq . ( [ g2e ] ) . @xmath230 _ collider bounds_. here , as it turns out , the only relevant collider bound is the @xmath11 c.l . lep lower bound @xcite on the mass of the lightest @xmath18-even neutral higgs boson @xmath7 : @xmath241 the susy corrections to the lightest @xmath18-even higgs boson mass @xmath100 are calculated at two loops by using the feynhiggsfast program @xcite included in the micromegas code @xcite . we will now try to delineate the parameter space of our model with @xmath1 which is consistent with the constraints in sec . [ sec : pheno ] . the restrictions on the @xmath242 plane , for @xmath225 and the central values of @xmath243 and @xmath244 , are indicated in fig . [ figa ] by solid lines , while the upper bound on @xmath223 from eq . ( [ cdmb ] ) , for @xmath245~{\rm gev}$ ] , is depicted by a dashed [ dotted ] line . we observe the following : * the lower bounds on @xmath223 are not so sensitive to the variations of @xmath228 . * the lower bound on @xmath223 from eq . ( [ mhb ] ) overshadows all the other lower bounds on this mass . * the upper bound on @xmath223 from eq . ( [ cdmb ] ) is very sensitive to the variations of @xmath228 . in particular , one notices the extreme sensitivity of the almost vertical part of the corresponding line , where the lsp annihilation via an @xmath96-boson exchange in the s - channel is @xcite by far the dominant process , since @xmath95 , which is smaller than @xmath246 , is always very close to it as seen from fig . this sensitivity can be understood from fig . [ figc ] , where @xmath95 is depicted versus @xmath223 for various @xmath228 s . we see that , as @xmath228 decreases , @xmath95 increases and approaches @xmath246 . the @xmath96-pole annihilation is then enhanced and @xmath231 is drastically reduced causing an increase of the upper bound on @xmath223 . * for low @xmath227 s , bino - stau coannihilations @xcite take over leading to a very pronounced reduction of the lsp relic abundance @xmath231 , thereby enhancing the upper limit on @xmath223 . so , we obtain the almost horizontal tail of the allowed region in fig . [ figa ] . for @xmath1 , @xmath225 , @xmath243 and @xmath244 , we find the following allowed ranges of parameters : @xmath247 the splitting between the bottom ( or tau ) and top yukawa coupling constants @xmath248 ranges between 0.25 and 0.29 . it is interesting to note that our susy gut model gives rise @xcite naturally to a modified version of shifted hybrid inflation @xcite . hybrid inflation @xcite , which is certainly one of the most promising inflationary scenarios , uses two real scalars : one which provides the vacuum energy density for driving inflation and a second which is the slowly varying field during inflation . this scheme , which is naturally incorporated @xcite in susy guts ( for an updated review , see ref . @xcite ) , in its standard realization has the following property @xcite : if the gut gauge symmetry breaking predicts topological defects such as magnetic monopoles @xcite , cosmic strings @xcite , or domain walls @xcite , these defects are copiously produced at the end of inflation . in the case of monopoles or walls , this leads to a cosmological catastrophe @xcite . the breaking of the @xmath175 symmetry predicts the existence of doubly charged monopoles @xcite . so , any ps susy gut model incorporating the standard realization of susy hybrid inflation would be unacceptable . one way to remedy this is to invoke @xcite thermal inflation @xcite to dilute the primordial monopoles well after their production . alternatively , we can construct variants of the standard susy hybrid inflationary scenario such as smooth @xcite or shifted @xcite hybrid inflation which do not suffer from the monopole overproduction problem . in the latter scenario , we generate @xcite a shifted inflationary trajectory so that @xmath175 is already broken during inflation . this could be achieved @xcite in our susy gut model even before the introduction of the extra higgs superfields , but only by utilizing non - renormalizable terms . the inclusion of @xmath249 and @xmath250 does not change this situation . the inclusion of @xmath195 and @xmath197 , however , very naturally gives rise @xcite to a shifted path , but now with renormalizable interactions alone . we showed that particle physics provides us with a number of candidate particles out of which the cdm of the universe can be made . these particles are not invented solely for explaining the cdm , but they are naturally there in various particle physics models . we discussed in some detail the major candidates which are the axion , the lightest neutralino , the axino , and the gravitino . the last three particles exist only in susy theories and can be stable provided that they are the lsp . the axion is a pseudo nambu - goldstone boson associated with the spontaneous breaking of a pq symmetry . this is a global anomalous @xmath19 symmetry invoked to solve the strong @xmath18 problem . it is , actually , the most natural solution to this problem which is available at present . the axions are extremely light particles and are generated at the qcd phase transition carrying zero momentum . we argued that these particles can easily provide the cdm in the universe . however , if the pq field emerges with non - zero value at the end of inflation , they lead to isocurvature perturbations , which , for superheavy inflationary scales , are too strong to be compatible with the recent results of the wmap satellite on the cmbr anisotropies . the most popular cdm candidate is , certainly , the lightest neutralino which is present in all susy models and can be the lsp for a wide range of parameters . we considered it within the simplest susy framework which is the mssm whose salient properties were summarized . we used exclusively the constrained version of mssm which is known as cmssm and is based on universal boundary conditions . in this case , the lightest neutralino is an almost pure bino , whereas the nlsp is the lightest stau . we sketched the calculation of the neutralino relic abundance in the universe paying particular attention not only to the neutralino pair annihilations , but to the neutralino - stau coannihilations too . it is very important for the accuracy of the calculation to treat poles and final - state thresholds properly and include the one - loop qcd corrections to the higgs boson decay widths and the fermion masses . we find that two effects help us reduce the neutralino relic abundance and satisfy the wmap constraint on cdm : the resonantly enhanced neutralino pair annihilation via an @xmath96-pole exchange in the s - channel , which appears in the large @xmath88 regime , and the strong neutralino - stau coannihilation , which is achieved when these particles are almost degenerate in mass . the axino , which is the susy partner of the axion , can also be the lsp in many cases since its mass is a strongly model - dependent parameter in the cmssm . it is produced thermally by 2-body scattering or decay processes in the thermal bath , or non - thermally by the decay of sparticles which are already frozen out of thermal equilibrium . for small axino masses , tp is more important yielding a very narrow favored region in the parameter space . for larger axino masses , however , ntp is more efficient and the favored region in the parameter space becomes considerably wider . one finds that , in the case of the cmssm , almost any point on the @xmath251 plane can be allowed by axino cdm considerations . the required reheat temperatures though are quite small ( @xmath252 ) . the mass of the gravitino is a practically free parameter in the cmssm . so , the gravitino can easily be the lsp and , in principle , contribute to the cdm of the universe . it is produced thermally by 2-body scattering processes in the thermal bath as well as non - thermally by the decay of the nlsp , which can be either the neutralino or the stau . in contrast to the axino case , however , the nlsp can now have quite a long lifetime . the electromagnetic showers resulting from the nlsp decay can destroy the successful predictions of bbn . so , we obtain strong constraints which allow only very limited regions of the parameter space of the cmssm . as it turns out , ntp in these regions is not efficient enough to account for cdm . we can , however , make these regions cosmologically favored by raising @xmath253 to enhance tp of gravitinos . we studied the cmssm with @xmath1 and @xmath225 applying a yquc which originates from a ps susy gut model . this condition yields an adequate deviation from yu which allows an acceptable @xmath228 . we , also , imposed the constraints from the cdm in the universe , @xmath104 , @xmath254 and @xmath100 . we found that there exists a wide and natural range of cmssm parameters which is consistent with all the above constraints . the parameter @xmath88 ranges between about 58 and 59 and the asymptotic splitting between the bottom ( or tau ) and the top yukawa coupling constants varies in the range @xmath255 for central values of @xmath228 and @xmath256 . the predicted lsp mass can be as low as about @xmath257 . moreover , the model resolves the @xmath93 problem of mssm , predicts stable proton , generates the baryon asymmetry of the universe via primordial leptogenesis , and gives rise to a new version of shifted hybrid inflation which is based solely on renormalizable interactions . we thank l. roszkowski , p. sikivie , and f.d . steffen for useful suggestions . this work was supported by european union under the contract mrtn - ct-2004 - 503369 as well as the greek ministry of education and religion and the epeak program pythagoras . # 1#2#3int . mod . phys . * # 1 * , # 3 ( # 2 ) # 1#2#3phys . b * # 1 * , # 3 ( # 2 ) # 1#2#3z . c * # 1 * , # 3 ( # 2 ) # 1#2#3phys . # 1 * , # 3 ( # 2 ) # 1#2#3rev . # 1 * , # 3 ( # 2 ) # 1#2#3phys . rep . * # 1 * , # 3 ( # 2 ) # 1#2#3phys . rev . d * # 1 * , # 3 ( # 2 ) # 1#2#3nucl . phys . * b#1 * , # 3 ( # 2 ) # 1#2#3nucl . b ( proc . sup . ) * # 1 * , # 3 ( # 2 ) # 1#2#3mod . # 1 * , # 3 ( # 2 ) # 1#2#3mod . lett . a * # 1 * , # 3 ( # 2 ) # 1#2#3annu . nucl . part . sci . * # 1 * , # 3 ( # 2 ) # 1#2#3sov . # 1 * , # 3 ( # 2 ) # 1#2#3jetp lett . * # 1 * , # 3 ( # 2 ) # 1#2#3acta phys . # 1 * , # 3 ( # 2 ) # 1#2#3riv . nuovo cim . * # 1 * , # 3 ( # 2 ) # 1#2#3ann . phys . * # 1 * , # 3 ( # 2 ) # 1#2#3prog . # 1 * , # 3 ( # 2 ) # 1#2#3astrophys . j. lett . * # 1 * , # 3 ( # 2 ) # 1#2#3nature * # 1 * , # 3 ( # 2 ) # 1#2#3astrophys . j. * # 1 * , # 3 ( # 2 ) # 1#2#3astron . j. * # 1 * , # 3 ( # 2 ) # 1#2#3mnras * # 1 * , # 3 ( # 2 ) # 1#2#3gen . # 1 * , # 3 ( # 2 ) # 1#2#3science * # 1 * , # 3 ( # 2 ) # 1#2#3bull . astron . soc . * # 1 * , # 3 ( # 2 ) # 1#2#3_ibid . _ * # 1 * , # 3 ( # 2 ) # 1#2#3comput . . commun . * # 1 * , # 3 ( # 2 ) # 1#2#3astropart . phys . * # 1 * , # 3 ( # 2 ) # 1#2#3eur . j. c * # 1 * , # 3 ( # 2 ) # 1#2#3nucl . meth . a * # 1 * , # 3 ( # 2 ) # 1#2#3j . high energy phys . * # 1 * , # 3 ( # 2 ) # 1#2#3j . # 1 * , # 3 ( # 2 ) # 1#2#3astrophys . * # 1 * , # 3 ( # 2 ) # 1#2#3lect . notes phys . * # 1 * , # 3 ( # 2 ) # 1#2#3j . phys . a * # 1 * , # 3 ( # 2 ) # 1#2#3prog . part . # 1 * , # 3 ( # 2 ) # 1#2#3jetp ( sov . phys . ) * # 1 * , # 2 ( # 3 ) c.l . bennett et al : ; d.n . spergel et al : w.l . freedman et al : d.v . ahluwalia - khalilova , d. grumiller : ; ; m. cirelli , n. fornengo , a. strumia : hep - ph/0512090 r. peccei , h. quinn : ; s. weinberg : ; f. wilczek : d.s.p . dearborn , d.n . schramm , g. steigman : j. preskill , m.b . wise , f. wilczek : ; l.f . abbott , p. sikivie : m. dine , w. fischler : p.j . steinhardt , m.s . turner : ; k. yamamoto : g. lazarides , c. panagiotakopoulos , q. shafi : ; g. lazarides , r.k . schaefer , d. seckel , q. shafi : k. choi , e.j . chun , j.e . kim : ; k. dimopoulos , g. lazarides : a.h . for a recent review see g. lazarides : p. sikivie : ; g. lazarides , q. shafi : ; h. georgi , m.b . wise : m.s . turner : l. campanelli , m. giannotti : astro - ph/0512458 h.v . peiris et al : c. gordon , a. lewis : ; p. crotty , j. garca - bellido , j. lesgourgues , a. riazuelo : ; c. gordon , k.a . malik : ; m. bucher , j. dunkley , p.g . ferreira , k. moodley , c. skordis : ; k. moodley , m. bucher , j. dunkley , p.g . ferreira , c. skordis : ; m. beltrn , j. garca - bellido , j. lesgourgues , a. riazuelo : k. dimopoulos , g. lazarides , d. lyth , r. ruiz de austri : ; g. lazarides , r. ruiz de austri , r. trotta : ; g. lazarides : g.l . kane , c. kolda , l. roszkowski , j.d . wells : j.r . ellis , k.a . olive , y. santoso , v.c . spanos : b. ananthanarayan , g. lazarides , q. shafi : ; . for a more recent update see u. sarid : hep - ph/9610341 ( in : _ snowmass 1996 , new directions for high - energy physics _ ) gmez , g. lazarides , c. pallis : ; m. drees , m.m . nojiri : v. barger , m. s. berger , p. ohmann : ; m. drees , m.m . nojiri : ; m. olechowski , s. pokorski : d. pierce , j. bagger , k. matchev , r. zhang : g. gamberini , g. ridolfi , f. zwirner : h. baer , c. chen , m. drees , f. paige , x. tata : f.m . borzumati , m. olechowski , s. pokorski : h. baer , m. brhlik , d. castao , x. tata : k. griest , d. seckel : c. pallis : ; ; hep - ph/0510234 m. drees , m.m . nojiri : ; m. drees : hep - ph/9703260 j. ellis , t. falk , k.a . olive : ; j. ellis , t. falk , g. ganis , k.a . olive , m. schmitt : j. ellis , t. falk , k.a . olive , m. srednicki : , ( e ) e.w . kolb , m.s . turner : _ the early universe _ ( addison - wesley , redwood city ca 1990 ) k. griest , m. kamionkowski , m.s . turner : p. gondolo , g. gelmini : t. falk , k.a . olive , m. srednicki : ; j. edsj , p. gondolo : t. nihei , l. roszkowski , r. ruiz de austri : m.e . gmez , g. lazarides , c. pallis : t. nihei , l. roszkowski , r. ruiz de austri : v.d . barger , c. kao : j.r . ellis , t. falk , g. ganis , k.a . olive , m. srednicki : phys . b * 510 * , 236 ( 2001 ) ; l. roszkowski , r. ruiz de austri , t. nihei : ; a.b . lahanas , d.v . nanopoulos , v.c . spanos : hep - ph/0211286 ( in : _ oulu 2002 , beyond the desert _ ) a. djouadi , j. kalinowski , m. spira : g. blanger , f. boudjema , a. pukhov , a. semenov : p. gondolo , j. edsj , l. bergstrm , p. ullio , e.a . baltz : astro - ph/0012234 ( in : _ york 2000 , the identification of dark matter _ ) l. covi , j.e . kim , l. roszkowski : ; l. covi , h .- b . kim , j.e . kim , l. roszkowski : ; a. brandenburg , f.d . steffen : l. roszkowski : hep - ph/0102325 ( in : _ york 2000 , the identification of dark matter _ ) ; j.e . kim : astro - ph/0205146 ( in : _ cape town 2002 , dark matter in astro- and particle physics _ ) k. rajagopal , m.s . turner , f. wilczek : ; e.j . chun , j.e . kim , h.p . nilles : l. covi , l. roszkowski , r. ruiz de austri , m. small : ; a. brandenburg , l. covi , k. hamaguchi , l. roszkowski , f.d . steffen : l. covi , l. roszkowski , m. small : j. ellis , j.e . kim , d. nanopoulos : ; t. moroi , h. murayama , m. yamaguchi : m. bolz , w. buchmller , m. plmacher : e. holtmann , m. kawasaki , k. kohri , t. moroi : phys . d * 60 * , 023506 ( 1999 ) ; t. gherghetta , g.f . giudice , a. riotto : ; t. asaka , k. hamaguchi , k. suzuki : ; j.l . feng , a. rajaraman , f. takayama : ; ; j.l . feng , s. su , f. takayama : ; l. roszkowski , r. ruiz de austri , k .- y . choi : j. high energy phys . * 08 * , 080 ( 2005 ) ; d.g . cerdeno , k .- y . choi , k. jedamzik , l. roszkowski , r. ruiz de austri : hep - ph/0509275 m. bolz , a. brandenburg , w. buchmller : f.d . steffen : hep - ph/0507003 g. lazarides , c. pallis : hep - ph/0404266 ( in : _ vrnjacka banja 2003 , mathematical , theoretical and phenomenological challenges beyond the standard model _ ) g. lazarides , c. pallis : hep - ph/0406081 g. lazarides , c. panagiotakopoulos : ; s. khalil , g. lazarides , c. pallis : l. hall , r. rattazzi , u. sarid : ; m. carena , m. olechowski , s. pokorski , c.e.m . wagner : m. carena , d. garcia , u. nierste , c.e.m . wagner : s.f . king , m. oliveira : s. abel et al . ( sugra working group collaboration ) : hep - ph/0003154 h. baer , j. ferrandis , k. melnikov , x. tata : t. blaek , r. dermek , s. raby : ; d. auto et al : u. chattopadhyay , a. corsetti , p. nath : ; c. pallis : g. lazarides , q. shafi , c. wetterich : ; g. lazarides , q. shafi : m.e . gmez , g. lazarides , c. pallis : ; c. pallis , m.e . gmez : hep - ph/0303098 r. barate et al . ( aleph collaboration ) : ; k. abe et al . ( belle collaboration ) : ; s. chen et al . ( cleo collaboration ) : r. jeannerot , s. khalil , g. lazarides : r. jeannerot , s. khalil , g. lazarides , q. shafi : g. lazarides : hep - ph/0011130 ( in : _ cascais 2000 , recent developments in particle physics and cosmology _ ) ; r. jeannerot , s. khalil , g. lazarides : hep - ph/0106035 ( in : _ cairo 2001 , high energy physics _ ) i. antoniadis , g.k . leontaris : g. lazarides , q. shafi : g.r . dvali , g. lazarides , q. shafi : g. lazarides , c. panagiotakopoulos , q. shafi : n. ganoulis , g. lazarides , q. shafi : ; g. lazarides , q. shafi : g. lazarides , q. shafi : m. fukugita , t. yanagida : g. lazarides : ; g. lazarides , n.d . vlachos : ; t. dent , g. lazarides , r. ruiz de austri : ; m.e . gmez , c. pallis : hep - ph/0303094 ( in : _ hamburg 2002 , supersymmetry and unification of fundamental interactions _ ) j. ellis , k.a . olive , y. santoso , v.c . spanos : ; a.b . lahanas , d.v . nanopoulos : ; h. baer , c. balzs : ; u. chattopadhyay , a. corsetti , p. nath : h. goldberg : ; j.r . ellis , j.s . hagelin , d.v . nanopoulos , k.a . olive , m. srednicki : b. aubert et al . ( babar collaboration ) : hep - ex/0207074 ; hep - ex/0207076 j.r . ellis , s. heinemeyer , k.a . olive , g. weiglein : a.l . kagan , m. neubert : ; p. gambino , m. misiak : m. ciuchini , g. degrassi , p. gambino , g. giudice : ; g. degrassi , p. gambino , g.f . giudice : s. martin , j. wells : m. davier , hep - ex/0312065 ( in : _ pisa 2003 , sighad 03 _ ) a. hcker : hep - ph/0410081 ( in : _ beijing 2004 , ichep 2004 _ ) ; a. vainshtein : g.w . bennett et al . ( muon @xmath258 - 2 collaboration ) : phys . rev . lett . * 89 * , 101804 ( 2002 ) , ( e ) g.w . bennett et al . ( muon @xmath258 - 2 collaboration ) : phys . rev . lett . * 92 * , 161802 ( 2004 ) ; a. aloisio et al . ( kloe collaboration ) : aleph , delphi , l3 and opal collaborations , the lep higgs working group for higgs boson searches : hep - ex/0107029 ( in : _ budapest 2001 , high energy physics _ ) ; lhwg - note/2002 - 01 , + http://lephiggs.web.cern.ch/lephiggs/papers/july2002_sm/index.html s. heinemeyer , w. hollik , g. weiglein : hep - ph/0002213 a.b . lahanas , d.v . nanopoulos , v.c . spanos : a.d . linde : e.j . copeland , a.r . liddle , d.h . lyth , e.d . stewart , d. wands : ; g.r . dvali , q. shafi , r.k . schaefer : ; g. lazarides , r.k . schaefer , q. shafi : v.n . senoguz , q. shafi : ; g. lazarides , c. panagiotakopoulos : g. t hooft : ; a. polyakov : jetp lett . * 20 * , 194 ( 1974 ) t.w.b . kibble , g. lazarides , q. shafi : ya.b . zeldovich , i.yu . kobzarev , l.b . okun : t.w.b . kibble : g. lazarides , m. magg , q. shafi : g. lazarides , q. shafi : g. lazarides , c. panagiotakopoulos , q. shafi : ; d.h . lyth , e.d .

the use of ultraviolet emission to shed light on the evolutionary history of galaxies is not straightforward . the presence of dust in galaxies represents one of the major obstacles complicating a direct quantification of the star formation activity in local and high redshift galaxies . absorption by dust grains reddens the spectra at short wavelengths completely modifying the spectral energy distribution of galaxies . since the uv radiation is preferentially emitted by young stars that are generally more affected by attenuation from surrounding dust clouds than older stellar populations , rest - frame uv observations can lead to incomplete and/or biased reconstructions of the star formation activity and star formation history of galaxies affected by dust absorption , unless proper corrections are applied . + radiative transfer models suggest that the total - ir ( tir ) to uv luminosity ratio method ( i.e. @xcite ) is the most reliable estimator of the dust attenuation in star - forming galaxies because it is almost completely independent of the extinction mechanisms ( i.e. dust / star geometry , extinction law , see @xcite ) . this method is based on the assumption that a fraction of photons emitted by newly formed young stars are absorbed by the dust . the dust heats up and subsequently re - emits the energy in the mid- and far - infrared . the amount of uv attenuation can thus be quantified by means of an energy balance . the reliability of this method has also made the @xmath1 ratio ideal to calibrate empirical methods to correct for dust attenuation when far - infrared observations are not available , like the well known @xmath2 relation ( @xcite , where @xmath3 is the slope of the uv continuum spectrum in the range 1300@xmath42600 ) . + although almost independent from dust - geometry and extinction law , the @xmath1 ratio method unfortunately depends on the age of the underlying stellar populations . in systems with low specific star formation rate ( ssfr ) the dust heating by old stars becomes important , and only a fraction of the dust emission is related to the ultraviolet absorption @xcite . + in the past this _ age effect _ has been generally considered negligible since uv observations were available only for active star - forming systems for which the @xmath1 vs. @xmath0 can be assumed to be independent from the star formation history ( sfh , @xcite ) . however after the launch of the galaxy evolution explorer ( galex , @xcite ) this may not be the case any more . galex is delivering to the community an unprecedented amount of uv data on local and high redshift galaxies covering the whole range of morphologies and star formation histories : from starbursts , to passive spirals and elliptical galaxies ( e.g. @xcite ) . can we extend the use of the _ standard method _ ( calibrated on active star forming galaxies and starbursts ) to correct for dust attenuation to the thousands of galaxies detected by galex ? in recent years various studies have been undertaken to address these issues ( i.e. @xcite ) . however more attention has been always given to the @xmath1 vs. @xmath3 relation than to the @xmath1 vs. @xmath0 relation . @xcite suggested that quiescent galaxies deviate from the @xmath2 relation of starburst galaxies , because they tend to have redder ultraviolet spectra at fixed @xmath1 ratio . they interpreted the different behavior of starbursts and normal galaxies as due to a difference in the star formation histories , proposing that the offset from the starburst @xmath2 relation can be predicted using the birthrate parameter @xmath5 ( @xcite ; i.e. the ratio of the current to the mean past star formation activity ) . in this work we use a different approach , investigating the dependence of the @xmath1 vs. @xmath0 relation on the galaxy star formation history , to quantify the influence of the age - independent correction on the interpretation of uv observations . we will adopt a very simple dust geometry model and extinction law to show that the use of standard ( i.e. age independent ) methods can strongly bias our interpretation of uv properties of local and high redshift galaxies . the main goal of this paper is to provide the community with some new empirical relations based on observable quantities suitable for deriving dust attenuation corrections taking into account the _ age effect_. more detailed geometries and dust models have been developed in the last decade ( e.g. @xcite ) . however , since our results do not depend on the geometry or extinction law adopted ( as shown in appendix a ) , we decided to adopt a very simple approach in order to make our recipes useful for the widest possible range of astrophysical applications . in section 2 we present our model and discuss the variation of the @xmath6 vs. @xmath7 relation with the age of the underlying stellar population . in section 3 we apply our model to a sample of nearby field and cluster galaxies and in section 4 we use this sample to test new empirical recipes to determine @xmath7 . additional applications of these techniques are discussed in section 5 . in order to investigate the correlation between the uv dust attenuation @xmath0 and the @xmath1 ratio for different stellar populations we used a library of spectral energy distributions ( seds ) obtained using the bruzual & charlot population synthesis models @xcite . we adopted a @xcite initial mass function ( imf ) and a star formation history ( sfh ) a la sandage in the formalism used by @xcite : @xmath8 where @xmath9 is the star formation rate per unit mass , @xmath10 is the age of the galaxy ( we assumed @xmath10=13 gyr at the present epoch ) and @xmath11 is the time at which the star formation rate reaches the highest value over the whole galaxy history : short @xmath11 correspond to galaxies dominated by old stellar populations while long @xmath11 correspond to young ( i.e. star forming ) galaxies . we investigated a range of @xmath11 between 0.1 and 25 gyr , with steps of 0.2 gyr for 0.1@xmath1210 gyr and 0.5 gyr for @xmath1310 gyr and metallicities in the range 0.02@xmath142.5 z@xmath15 in five steps : 0.02 , 0.2 , 0.4 , 1 , and 2.5 z@xmath15 . the library obtained is able to reproduce all the seds typically observed in local galaxies @xcite and it is only a function of @xmath11 will be only used to indicate the shape of the sed and it must not be adopted to quantify the real age of the stellar populations . ] . similar libraries can be obtained assuming different sfh ( e.g. exponential sfh , see appendix a ) . + each synthetic sed in our library has been reddened using different values of @xmath7 in the range 0@xmath16 4 mag , the typical range observed in normal star - forming galaxies ( e.g. @xcite ) . higher uv attenuations are normally associated with strong starbursts and highly obscured objects which are outside the goal of the present work . for each @xmath7 , @xmath17 ( the attenuation at each @xmath18 ) has been derived assuming a large magellanic cloud ( lmc ) extinction curve @xcite : @xmath19 and a simple sandwich model for dust geometry @xcite , where a thin layer of dust of thickness @xmath20 is embedded in a thick layer of stars : @xmath21\left(1+{\rm e}^{-\tau_{dust}(\lambda ) \cdot { \rm sec}(i)}\right ) + & & { } \nonumber \\ { } + \left[\frac{\zeta(\lambda)}{\tau_{dust}(\lambda ) \cdot { \rm sec}(i)}\right ] \cdot \left(1-{\rm e}^{-\tau_{dust}(\lambda ) \cdot { \rm sec}(i)}\right ) \bigg ) ~~[{\rm mag}]\end{aligned}\ ] ] where @xmath22 is the galaxy inclination , @xmath23 is the optical depth and the dust to stars scale height ratio @xmath24 depends on @xmath18 ( in units of ) as @xcite : @xmath25 in the case of the fuv band ( @xmath26 ) @xmath27 . in this case @xmath28(fuv ) can be derived by inverting eq . [ alambda ] : @xmath29 \cdot \big(0.0259 + 1.2002 \times a({\rm fuv } ) + 1.5543 \times & & { } \nonumber \\ { } \times a({\rm fuv})^2 - 0.7409 \times a({\rm fuv})^3 + 0.2246 \times a({\rm fuv})^4 \big ) & & \end{aligned}\ ] ] using the lmc extinction law @xmath30 , we then derive : @xmath31 and compute the complete set of @xmath17 using eq.[alambda ] . in order to simplify the calculations in the following we assume @xmath22=0 . we adopted the lmc extinction law as a compromise between a milky way extinction law , with a strong bump at 2175 , and the small magellanic cloud , with almost no bump at @xmath322000 . recent analysis have also pointed out that a lmc extinction law can reproduce galex observations better than a mw or smc extinction law @xcite the reddened sed is then subtracted from the original _ dust - free _ sed , providing ( once integrated over all wavelengths ) the amount of energy absorbed by dust which , assuming an energetic balance , corresponds to the total infrared radiation ( tir ) emitted by the galaxy . finally , by convolving the reddened sed with the galex - fuv filter we estimated the fuv flux and the @xmath6 ratio for each model . the relations between @xmath7 and @xmath6 so obtained for different values of @xmath11 and stellar metallicity are shown in fig.[tauafuv ] . + the mean age of the stellar populations plays a crucial role in the @xmath6 vs @xmath7 relation , as already pointed out by several theoretical ( e.g. @xcite ) and observational ( e.g. @xcite ) studies : for the same @xmath6 ratio active star forming galaxies ( @xmath337 gyr ) are more attenuated than objects with low ssfr ( @xmath34 4 - 5 gyr ) or quiescent galaxies ( @xmath35 2 gyr ) . on comparison , we find a very weak dependence ( @xmath360.1 mag ) of the @xmath6 vs. @xmath7 on stellar metallicity ( see fig.[tauafuv ] ) . in table [ convfuv ] we provide the best polynomial fit ( averaged over the whole metallicity range considered ) to the @xmath7 and @xmath6 for different values of @xmath11 . the origin of the _ age effect _ is clearly visible in fig.[enbudg ] where we compare the amount of the energy absorbed by dust at long ( @xmath37 ) and short ( @xmath38 ) wavelengths for different values of @xmath11 . for @xmath39 gyr the high - energy photons ( mainly uv ) contribute to less than the 50% of the total energy absorbed by dust and the far infrared emission is mainly due to the re - emission of the stellar radiation emitted by intermediate age stars in the optical . only for @xmath40 gyr the uv radiation dominates the dust heating , contributing more than the 75% of the whole energy absorbed and re - emitted in the far infrared . this also implies that the far - infrared radiation is not always a good star formation indicator . + we remark that our results do not strongly depend on the parameters adopted in our model ( e.g. dust geometry , attenuation law ) . the variation in the @xmath6 vs. @xmath7 relations when different attenuation laws are considered is @xmath410.2 mag , as discussed in appendix a. this is confirmed by the good agreement between our model and the one of @xcite ( black dotted line in fig.[tauafuv ] ) in the case of strong star forming systems and starburst galaxies ( @xmath338 gyr ) . our results are also independent of the shape of the adopted sfh and galaxy age . in fact the @xmath6 vs. @xmath7 relation mainly depends on the specific star - formation and , for the same specific star formation activity , it is not strongly affected by the real shape of the past sfh ( see appendix a ) . moreover , in fig.[tauafuv ] we show the @xmath6 vs. @xmath7 relation proposed by @xcite ( empty pentagons ) for a birthrate parameter @xmath420.06 ( the lowest birthrate parameter considered in their model , corresponding to @xmath344.1 gyr in our formalism ) . the level of agreement ( @xmath320.2 mag ) between @xcite and our model is quite comforting , considering that the @xcite model is based on different sfh ( i.e. exponential+bursty model ) and geometry ( i.e. the time - dependent scenario proposed by @xcite ) . + as already discussed by @xcite , the flux ratio method would not be applicable in the case of an embedded starburst in a galaxy with a second older , less embedded stellar population . at uv and ir wavelengths , the starburst would dominate , but at optical and near - ir wavelengths the older population would dominate . however , at least for spiral galaxies in the local universe this seems not to be the case @xcite and smooth sfh like an a la sandage and an exponential well reproduce the observations ( e.g. @xcite ) . we therefore conclude that our results are model - independent within 0.2 mag ( see appendix a ) . in order to quantify the impact of the _ age effect _ on the estimate of the uv attenuation , we computed @xmath7 for a sample of spiral galaxies and compared our estimate with the one obtained using the standard relation calibrated on active star - forming galaxies ( e.g. @xmath338 gyr , @xmath434 mag , @xmath440.5 ) . the sample here adopted is an extension of the optically selected sample described in @xcite , composed by late - type galaxies ( later than s0a ) including giant and dwarf systems extracted from the virgo cluster catalogue ( vcc , @xcite ) and from the cgcg catalogue @xcite . the data include galaxies in the virgo , abell1367 and abell262 clusters and part of the coma - a1367 supercluster ( including cluster and field galaxies ) observed as part of the all - sky imaging survey ( ais ) and of the nearby galaxy survey ( ngs ) carried out by galex in two uv bands : fuv ( @xmath45 ) and nuv ( @xmath46 ) . we include in our analysis only late - type galaxies detected in both nuv and fuv galex bands and in both 60 @xmath47 and 100 @xmath47 iras bands : 191 galaxies in total ( @xmath3270% in high density environments ) . uv and far - infrared data have been combined with multifrequency data available . these are optical and near - ir h imaging @xcite , most of which are available from the goldmine galaxy database @xcite ( http://goldmine.mib.infn.it ) . data from uv to near - ir have been corrected for galactic extinction according to @xcite . + we assume a distance of 17 mpc for the members of virgo cluster a , 22 mpc for virgo cluster b , and 32 mpc for objects in the m and w clouds @xcite . members of the cancer , a1367 , and coma clusters are assumed to lie at distances of 65.2 , 91.3 , and 96 mpc , respectively . isolated galaxies in the coma supercluster are assumed at their redshift distance , adopting @xmath48 = 75 @xmath49 . the fuv attenuation of each galaxy in our sample has been computed using the following sed fitting procedure , implemented into the galaxy observed simulated sed interactive program ( gossip , @xcite ) . in order to take into account the _ age effect _ we reddened each synthetic sed and fit them to the observed fuv - to - near - infrared sed obtained using all the available photometric magnitudes . + in details , for each galaxy we compute the observed @xmath6 ratio using iras and galex observations . the tir flux emitted in the range 1 - 1000 @xmath47 , is obtained following @xcite : @xmath50 where @xmath51 is the far - infrared flux , defined as the flux between 42 and 122 @xmath47 @xcite : @xmath52\ ] ] and @xmath53 and @xmath54 are the iras flux densities measured at 60 and 100 @xmath47 ( in jansky ) . using the relations presented in the previous section , we then convert the observed @xmath6 ratio into @xmath7 for each value of @xmath11 and @xmath55 considered in our model and determine @xmath17 as described in sec.[model ] . each @xcite sed is then reddened with the @xmath17 obtained following this procedure and fitted to the observed ( i.e. not corrected for internal extinction ) sed , using a @xmath56 fitting technique similar to the one described in @xcite and assuming a conservative photometric error of 0.15 mag for each band . for each galaxy , the value of @xmath56 determines the weight of a given model ( as @xmath57 ) and a probability distribution function ( pdf ) for @xmath11 can be build by combining weights for each model . we then normalize the final pdfs and use the peak value ( i.e. the most probable one ) as the best estimate of @xmath11 and the range of @xmath11 containing 68.2% of the pdf s area ( equivalent to the use of constant @xmath56 contours ) as estimate of the 1 @xmath58 error . the sed fitting provides not only the best value of @xmath11 but also the right estimate of @xmath7 and its uncertainty . the combined use of the observed @xmath6 ratio for each galaxy and of the relations presented in the previous section makes this method ideal to estimate the uv attenuation for a wide range of morphological types and star formation histories . the difference between the fuv attenuation obtained with the sed fitting technique ( @xmath59 ) and the one obtained by blindly applying a constant conversion calibrated on star forming galaxies and starbursts ( @xmath338 gyr , @xmath60 consistent with the one presented by @xcite ) is show in fig.[diffafuv_tau ] as a function of @xmath11 ( for clarity only the errors on a(fuv ) are shown ) . as already shown in fig.[tauafuv ] , for low values of @xmath11 the standard conversion overestimate the uv dust attenuation : in particular for @xmath32 30% ( 59/191 galaxies ) of our sample the use of a constant @xmath6 vs. @xmath7 relation leads to an overestimate of more than 0.5 mag of the uv attenuation and for @xmath3216% this systematic error exceeds 1 mag . in order to investigate the properties of the most deviating objects in fig.[diffafuv_tau ] we divide our sample according to their hi - deficiency and optical linear diameter @xmath61 : hi - def = @xmath62 @xcite ] : a value of @xmath63 = 0.4 has been used to separate healthy ( blue circles ) from hi - deficient star forming spirals ( red triangles ) . hi - deficient spirals are among the most deviating objects in fig.[diffafuv_tau ] , consistent with the fact that gas deficient objects have normally a ssfr significantly lower than that expected from their luminosity @xcite . however also some healthy spirals show a significant offset from the _ age independent _ @xmath6 vs. a(fuv ) relation . these are massive early type spirals with low ssfr , similar to m31 in the local group ( i.e. @xmath647 mag , @xcite ) . + the @xmath6 ratio can not thus be considered as a good proxy of the uv attenuation for sample of galaxies spanning a wide range of ssfr , in particular in clusters , and relations like the @xmath2 relation @xcite can not be blindly used to determine @xmath7 . this is shown in fig . [ afuv_beta ] where we compare the @xmath7 vs. @xmath3 relation obtained from our sed fitting technique ( filled symbols ) with the one obtained when @xmath7 is computed using the @xmath7 vs. @xmath6 relation for strong star forming galaxies ( empty symbols ) . the @xmath3 parameter has been computed from the fuv - nuv colour following @xcite : @xmath3=2.201@xmath65 - 1.804 . galaxies with @xmath66 ( @xmath670.75 ) are mainly hi - deficient galaxies and have a fuv dust attenuation @xmath322 mag lower than the one obtained using the standard @xmath68 relation : a difference significantly larger than the typical error on the estimate of a(fuv ) ( @xmath410.5 mag ) by our fitting technique . the systematic errors introduced in the data is similar when the @xmath6 ratio is not available and @xmath7 is determined using empirical methods based on the @xmath69 colour like the one proposed by @xcite , and once again calibrated on strong star forming galaxies ( solid and dashed line in fig.[afuv_beta ] ) : for @xmath70 - 0.2 , a uv attenuation in the range 2.99@xmath163.32 mag is predicted , whereas our method gives an average value of @xmath71 1 mag . this is mainly due to the fact that in low star forming systems the uv spectral slope is strongly contaminated by the old stellar populations , whose contribution increases the value of @xmath3 even if the fuv attenuation is low . this result is consistent with the recent analysis of a sample of @xmath321000 galaxies selected from sdss presented by @xcite , who found that a large fraction of galaxies having red @xmath69 colours is also characterized by large 4000 break ( @xmath72 ) , suggesting that part of the dust heating comes from old stellar populations and not from extremely obscured star forming regions . we can therefore conclude that the @xmath2 relation can be blindly used to estimate @xmath7 only for @xmath73 ( @xmath740.75 ) . it clearly appears that a correct estimate of the uv attenuation requires information about the shape of the galaxy sed . a sed fitting technique like the one here described represents therefore the best method available to quantify @xmath7 and properly correct uv observations . unfortunately sed fitting is only possible when large multiwavelength data sets are available , which is not always the case . for this reason we investigated the possibility to derive different techniques which can be used to estimate @xmath7 when sed fitting is not possible . ideally , these techniques will have a small rms and not display the systematic shift relative to the galaxy age observed in fig.[diffafuv_tau ] . the crucial step is to find a good proxy for @xmath11 ( and therefore the shape of the sed ) . this is particularly important for @xmath75 gyr where the @xmath6 vs. @xmath7 relation strongly depends on the age of the stellar populations . the first natural choice is to look for a colour with the largest possible dynamical range , sensitive to small changes in the shape of the sed @xcite . therefore in fig.[fuvh_tau ] we plot the distribution of @xmath11 for our sample as a function of the observed ( i.e. not corrected for internal extinction ) @xmath76 colour . for @xmath137 gyr the two variables are not correlated and galaxies show approximately the same @xmath76 colour independently from the value of @xmath11 , reflecting the large error on the estimate of @xmath11 . this is as expected since in young stellar populations ( @xmath137 gyr ) the variations in the observed @xmath76 are mainly due to dust attenuation and not to age . however for @xmath777 gyr ( i.e. the range in which we are interested ) the @xmath76 is a very good proxy for @xmath11 ( pearson correlation coefficient @xmath78 ) . a simple lest - square fitting gives us : @xmath79 which can be used to estimate @xmath11 from the @xmath76 colour . the dispersion on this relation varies quite remarkably with the colour . we therefore determine the typical dispersion within 0.5 mag wide fuv - h bins by combining the @xmath11 s pdf for each galaxy in the bin and estimating the 1@xmath58 error as described in the previous section . the dispersion in the relation ( indicated by the shaded area in fig.[fuvh_tau ] ) is @xmath80 for 6@xmath819 mag , increasing significantly ( up to @xmath821 ) for redder or bluer colours . similar relations are found when we consider different far - ultraviolet - optical colours and are presented in table [ taucol ] . we therefore propose two different ways to determine @xmath7 depending on whether or not far infrared observations are available . in appendix b we also provide similar recipes in order to determine @xmath83 in the case fuv observations are not available . + i ) * far - infrared observations are available . * determine the observed @xmath6 ratio and use one of the relations presented in table [ taucol ] to determine @xmath11 . finally use the value of @xmath11 so obtained to choose the conversion between the @xmath6 ratio and @xmath7 from the relations provided in table [ convfuv ] . as discussed above we suggest using the colour covering the widest possible dynamical range , i.e. first choice @xmath76 , last choice @xmath84 . the error on the estimate of the fuv dust attenuation ( @xmath85 ) obtained with this method depends on the observational errors on the fuv - optical / near - infrared colour and on the intrinsic dispersion of the colour-@xmath11 relation adopted . in order to estimate @xmath85 , for each galaxy in our sample we generated 1000 random galaxies having @xmath76 colour following a gaussian distribution centered on the observed colour with @xmath58=0.2 mag ( more accurate estimate of the @xmath76 colour would correspond to a lower uncertainty on @xmath7 ) . for each @xmath76 colour the @xmath11 s pdf are used to random generate the correspondent value of @xmath11 vs @xmath11 relation in the estimate of @xmath11 . ] and then the final value of @xmath7 by applying the relations in table [ taucol ] . in fig.[test ] ( upper panel ) we compare this method with the sed fitting technique . the shaded area indicates the 1 @xmath58 error on the estimate of @xmath7 . as expected , the error on the estimate of @xmath7 varies significantly with @xmath11 : from @xmath32 + 0.1/-0.25 mag for @xmath336 gyr ( @xmath865 mag ) to a maximum of + 0.4/-0.8 mag for @xmath87 gyr ( 7@xmath818.5 mag ) . even if the uncertainty for low @xmath11 is quite large ( due to the large variation of the @xmath6 vs. @xmath7 relation with @xmath11 ) , it is clear that our empirical method is able to remove the systematic overestimate of @xmath7 when the age independent @xmath6 vs. @xmath7 conversion is used ( filled symbols in fig.[test ] ) . + ii ) * far - infrared observations are not available . * in this case the first step is to find a way to estimate the @xmath6 ratio from the available observations . recently various methods have been proposed based on the use of the @xmath69 colour ( or @xmath3 , @xcite ) , h - band luminosity @xcite , gas metallicity @xcite , effective surface brightness @xcite , @xmath72 or optical and ultraviolet colours @xcite . once the @xmath6 ratio has been determined it is possible to proceed as described above : i.e. estimate @xmath11 from an ultraviolet - optical / near infrared colour ( see table [ taucol ] ) and than convert @xmath6 in @xmath7 using the proper relation in table [ convfuv ] . to test this second method we estimate the @xmath6 ratio from the @xmath69 colour @xcite : @xmath88 and then determine @xmath11 from the @xmath76 colour . the typical uncertainty in this method has been computed following the same procedure described in the previous point . also in this case our recipe is able to remove the systematic overestimate on @xmath7 . this is clearly evident in fig.[test ] ( lower panel ) where our method is compared with the age independent method based on the @xmath69 colour proposed by @xcite . unfortunately the uncertainty on the determination of @xmath7 considerably increases to @xmath32 + 0.5/-0.7 mag for @xmath336 gyr ( @xmath865 mag ) reaching a maximum of @xmath32 + 1/-1.2 mag for @xmath87 gyr ( 7@xmath818.5 mag ) when the total infrared luminosity is estimated from empirical relations involving colours or luminosities . as discussed by @xcite this is mainly due to the intrinsic scatter in the empirical relation used to determine the @xmath6 ratio . + in the very unlikely scenario in which there is no sufficient data to estimate the @xmath6 ratio and/or @xmath11 as described above , the only possibility is to use the morphological type to obtain a rough estimate of @xmath7 . in table [ afuv_type ] we give the average value and standard deviation of @xmath7 obtained for our sample in bins of morphological type . for our sample this method provides an estimate of @xmath7 with an average error of @xmath320.7 mag . however we recommend to apply this technique only when accurate an morphological classification is available ( i.e. the local universe ) , otherwise the error on @xmath7 will be considerably larger than the one obtained for our sample . the simple recipes presented here have the widest scientific application possible , requiring only an uv and uv - optical colour . in order to show their real power , in the following we will discuss two different applications of these methods to the study of the uv properties of nearby galaxies and compare them with the age independent techniques usually adopted . the _ age effect _ on the @xmath6 vs. @xmath7 relation is particularly important in panchromatic studies of resolved galaxies @xcite since in different regions the dust can be heated up by different stellar populations @xcite . + we test our method on the nearby spiral galaxy ngc 4569 ( m90 ) , the brightest spiral galaxy in the virgo cluster recently studied by @xcite . ngc 4569 can be considered as the prototype of hi deficient galaxy having only about one - tenth of the atomic gas of a comparable field galaxy of similar type and size . this galaxy has a truncated h@xmath89 and hi radial profile ( at a radius of @xmath32 5 kpc , @xcite ) and shows significant colour gradients with star formation activity only in the central 5 kpc . therefore , given its significant age gradients , the contribution of uv photons to the dust heating probably varies with the distance from the galaxy center , making this object ideal to compare with our recipes with age independent techniques . as described in @xcite we determined the @xmath6 ratio profile combining galex - fuv image and spitzer 24 , 70 , and 160 @xmath90 radial profiles . then , we use the observed @xmath76 colour profile to estimate @xmath11 at each radius , determining which @xmath6 vs. @xmath7 relation to use at each radius . the @xmath7 profile so obtained is shown in fig.[n4569 ] ( left panel , empty circles ) and compared to the one estimated using the age independent @xmath6 vs. @xmath7 conversion @xcite . as expected , our method gives a fuv attenuation @xmath321.5 mag lower than the standard conversion implying a factor @xmath324 difference in the sfr surface density profile obtained from the fuv profile ( fig.[n4569 ] , right panel ) . the results obtained with our technique are supported by the very good agreement with the sfr density profile independently obtained from the h@xmath89 line ( corrected for extinction using balmer decrement ; fig . [ n4569 ] right panel , empty triangles ) . in fact only a recent ( @xmath91 yr ) starburst @xmath32100 times stronger than the normal star formation rate in ngc 4569 could reconcile the difference between the sfr obtained from the h@xmath89 and from the fuv corrected with the standard methods @xcite : an extremely unlikely scenario as discussed by @xcite and @xcite . + this exercise shows the crucial importance of the age dependent @xmath6 vs @xmath7 relations in the panchromatic study of resolved galaxies . the need of empirical methods to correct for dust attenuation is particular important for the study of uv properties of large samples of galaxies lacking far - infrared rest - frame data . of particular importance for our understanding of galaxy evolution is the correct interpretation of the uv - optical colour magnitude ( cm ) relation , as recently discussed by @xcite , @xcite and @xcite . it is very difficult to use a single correction technique able to deal simultaneously with starbursts , low ssfr objects and elliptical galaxies . of particular importance is the correct estimate of @xmath7 for transition objects : galaxies with low ssfr for which the standard corrections calibrated on active star forming systems are likely to be not valid . in order to quantify the influence of the _ age effect _ on our interpretation of the uv - optical colour magnitude relation we compare our method with the age independent recipes using a sample of galaxies extracted from galex observations of the coma cluster region ( gi - cycle 1 , cortese et al . , in preparation ) . this sample includes all galaxies detected in both fuv and nuv bands as well as with sdss - dr6 @xmath92 photometry and with spectroscopic redshift data available ( 833 galaxies ) . since for this large sample far infrared observations are not available , we use the second method described in sec.[recipe ] , determining the @xmath6 ratio from the @xmath69 colour and @xmath11 from the @xmath93 colour . the @xmath94 colour-@xmath95 magnitude relation obtained is compared with the observed one ( i.e. not corrected for internal extinction ) in fig.[cm ] ( left panel ) . whereas galaxies in the blue sequence ( @xmath96 4 mag ) have an average fuv dust attenuation @xmath711.5 - 2 mag , galaxies in the red sequence ( @xmath976 mag ) show ( as expected ) a very low amount of attenuation ( @xmath71 0.5 mag ) and occupy almost the same parameter space as uncorrected data . the separation between the red and blue sequences is therefore more evident after the correction for dust attenuation . we remark that our technique should not be blindly applied to elliptical galaxies , since in these objects uv emission comes from old stellar populations and not young stars ( e.g. @xcite ) . however if an accurate morphological classification is not available the use of the recipes presented here does not introduce a strong systematic bias in the data , as shown in fig.[cm ] ( left panel ) . this result indicates that our procedures are reliable for old as well as young stellar populations , contrary to previous empirical methods , calibrate and valid only for active star - forming galaxies ( e.g. @xcite ) . this is clearly evident in fig.[cm ] ( right panel ) where the cm relation obtained with our recipe is compared with the one determined by using the age - independent method to convert the @xmath69 colour into @xmath7 ( e.g. @xcite ) . the difference between the two cm relations is quite impressive : while no significative difference is observed in blue - sequence galaxies , the red sequence shifts @xmath322 mag towards bluer colours and the gap between the blue and red sequence ( i.e. the so - called `` green valley '' , @xcite ) is considerably reduced . a similar result has been recently shown by @xcite when comparing the cm relation determined using the balmer decrement as indicator of dust attenuation with the one estimated using the recipes proposed by @xcite . this result shows how strong the bias can be when blindly using the recipes to estimate @xmath7 without taking into account the _ age effect _ , leading to an incorrect interpretation of the data and reconstruction of galaxy s evolution history . we have investigated the dependence of the @xmath6 vs. @xmath7 relation on the average age of galaxy stellar populations . our simple method has shown that even for spiral galaxies the use of a standard ( i.e. not age dependent ) conversion of the @xmath6 into @xmath7 leads to a systematic overestimate ( i.e. @xmath981 mag ) of the dust attenuation in galaxies with low specific star - formation , mainly anemic cluster spirals . this systematic bias strongly affects our interpretation of uv observations and can produce a significant overestimate ( up to a factor 10 ) of the star formation rate , particularly in hi deficient galaxies . therefore we have developed different methods for determining the uv dust attenuation taking into account the age dependence of the @xmath6 vs. @xmath7 relation . these recipes require only an uv colour and an uv - optical / near infrared colour providing an estimate of the uv attenuation with an average uncertainty varying from @xmath32 + 0.1/-0.25 mag for @xmath336 gyr ( @xmath865 mag ) to a maximum of + 0.4/-0.8 mag for @xmath87 gyr ( 7@xmath818.5 mag ) , when far - infrared observations are available , and from @xmath32 + 0.5/-0.7 mag for @xmath336 gyr ( @xmath865 mag ) to a maximum of @xmath32 + 1/-1.2 mag for @xmath87 gyr ( 7@xmath818.5 mag ) when far - infrared data are not available . the small amount of multiwavelength data necessary for their application makes these methods extremely useful for the widest possible range of application eventually providing us with a less biased view of the uv properties of galaxies in the local universe and at higher redshift . we want to thank the anonymous referee , whose comments and suggestions were extremely useful for improving the present manuscript . we thank jonathan davies for useful discussions and comments on this manuscript . lc is supported by the uk particle physics and astronomy research council . galex ( galaxy evolution explorer ) is a nasa small explorer , launched in april 2003 . we gratefully acknowledge nasa s support for construction , operation , and science analysis for the galex mission , developed in cooperation with the centre national detudes spatiales of france and the korean ministry of science and technology . this research has made use of the nasa / ipac extragalactic database , which is operated by the jet propulsion laboratory , california institute of technology , under contract to nasa and of the goldmine database are the results obtained in this work and the recipes proposed to estimate the uv dust attenuation model - dependent ? in order to answer to this question , in this section we investigate the dependence of our results on the three free parameters entering our model : the shape of the extinction law , the dust geometry and the star formation history . in fig.[modgeo ] ( left ) we show the @xmath6 vs @xmath7 relation obtained for different values @xmath11 , a sandwich geometry and three different extinction laws : lmc ( solid line ) , milky way ( dotted line ) and small magellanic cloud ( dashed line ) . as already pointed out by several authors ( e.g. @xcite ) , the @xmath6 vs @xmath7 relation is almost independent of the extinction law with a typical variation limited to @xmath410.2 mag , considerably lower than the effect of @xmath11 on the @xmath6 vs @xmath7 relation . similar results are obtained if we investigate the effect of different dust geometries . in fig.[modgeo ] ( right ) we compare the results obtained for the sandwich model ( solid line ) with a simple slab geometry ( @xcite , dashed line ) . we also added the results for a @xcite attenuation law ( dotted line ) . even in this case , we observed no dependence of the @xmath6 vs. @xmath7 relation on the dust geometry within @xmath410.2 mag . similar results are obtained if we use the homogeneous and clumpy dust models proposed by @xcite ( not shown ) . finally , we fitted the 191 galaxies in our sample with various sed libraries in order to test the dependence of the results presented in sec.[data ] and sec.[recipe ] from sfh and age . the sed library used in sec.[model ] has been obtained by fixing the shape of the sfh ( assumed to be a la sandage ) , the age of the galaxy ( assumed equal to 13 gyr ) and the stellar initial mass function ( imf , assumed to be a salpeter imf ) . we produced different set of sed libraries varying every time one of these three free parameters and we fitted them to the data . in particular we considered an exponential sfh , a fixed ( @xmath10=13 gyr ) or free ( in the range 0@xmath9915 gyr ) age and a @xcite imf . in fig.[modsfh ] is presented the difference in the estimate of @xmath7 between the library adopted in sec.[model ] and some of the various combination adopted . for each combination the standard deviation in the estimate of @xmath7 is lower than 0.1 - 0.2 mag . this result was somehow expected since the @xmath6 vs. @xmath7 relation is only affected by the actual shape of the sed and not by the shape of the past ( older than 1 - 2 gyr ) sfh . we remark that our results are not applicable only in the case of an embedded starburst in a galaxy with a second older , less embedded stellar population @xcite . at uv and ir wavelengths , the starburst would dominate , but at optical and near - ir wavelengths the older population would dominate . however , at least for spiral galaxies in the local universe this seems not to be the case @xcite and smooth sfh like a la sandage and exponential well reproduce the observations ( e.g. @xcite ) we can therefore conclude that the results obtained from our model and the recipes presented in this work are model - independent at least within @xmath320.2 mag . this uncertainty is still lower or equal to the uncertainty in the estimate of @xmath7 from our recipes . in this appendix we discuss the possibility to extend our recipes to the nuv band in the case that galex - fuv observations are available . the galex - nuv filter ( @xmath46 ) lies in a region where the sed of galaxies with @xmath100 4 gyr is strongly dependent on stellar metallicity for @xmath101 ( e.g. see fig . 6 in @xcite ) . this strongly affects the @xmath102 vs @xmath83 relation which , for low values of @xmath11 , can vary of even 0.4 mag from @xmath103 to @xmath104 , apparently complicating the use of our model directly on the nuv band . however for none of the galaxies with @xmath100 4 gyr in our sample the best fitting model has metallicity lower than @xmath103 , consistent with the fact that in the local universe evolved galaxies tend to be metal rich ( e.g. @xcite ) . therefore for @xmath100 4 gyr we compute the average @xmath102 vs @xmath83 relation by combining only the relations obtained in the metallicity range @xmath105 . this makes our recipes not valid for @xmath100 4 gyr and @xmath106 . the average @xmath102 vs. @xmath83 relations for different values of @xmath11 so obtained are presented in table [ convnuv ] . in table [ taunuv ] are given the relations to use to estimate @xmath11 from the nuv - optical / near - infrared colours . their typical dispersion is consistent with the one observed in the fuv - optical / near - infrared colours vs. @xmath11 relations . as discussed in sec.[recipe ] the suggested recipes to determine @xmath83 result : + i ) * far - infrared observations are available . * determine the observed @xmath102 ratio and use one of the relations presented in table [ taunuv ] to determine @xmath11 . finally use the value of @xmath11 obtained with this procedure to choose the conversion between the @xmath6 ratio and @xmath7 from the relations provided in table [ convnuv ] . the typical error on the estimate of @xmath83 ( computed through a montecarlo simulation similar to the one described in sec . 4 ) varies between @xmath32 + 0.1/-0.2 mag for @xmath338 gyr , to + 0.3/-0.7 mag for @xmath87 gyr . + ii ) * far - infrared observations are not available . * in this case the first step is to find a way to determine the @xmath102 from the observations available . once the @xmath102 ratio has been determined it is sufficient to proceed as described in the previous point : i.e. estimate @xmath11 from an ultraviolet - optical / near infrared colour and than convert @xmath102 in @xmath83 using the proper relation in table [ convnuv ] . the typical error on the estimate of @xmath83 ( computed through a montecarlo simulation similar to the one described in sec . 4 ) varies between @xmath32 + 0.4/-0.6 mag for @xmath338 gyr , and to @xmath32 + 0.9/-1 mag for @xmath87 gyr . + in table [ anuv_type ] are presented the average value of @xmath83 obtained for the different morphological types in our sample . the morphological type can be used to estimate @xmath83 only if the to previous methods can not be applied and an accurate morphological classification is available . the relationship between the @xmath6 ratio and the fuv attenuation @xmath7 obtained from our model for different values of @xmath11 . solid and dashed lines show the relations for stellar metallicity @xmath55=2.5 z@xmath15 and @xmath55=0.02 z@xmath15 respectively . the dotted black line indicates the age independent relation given by buat et al . empty pentagons show the relation proposed by kong et al . ( 2004 ) for birthrate parameter @xmath420.06.,width=321 ] the importance of the uv emission in the dust heating . the ratio of the energy absorbed by dust at @xmath107 and at @xmath108 is shown as a function of @xmath11 ( i.e. the age of the universe at which the galaxy sfr peaks).,width=321 ] the difference between the fuv attenuation for active star forming galaxies ( @xmath33 8 gyr ; consistent with buat et al . ( 2005 ) ) and the one obtained from our sed fitting technique as a function of @xmath11 . circles indicate healthy spirals ( hi - def@xmath360.4 ) and triangles hi - deficient star forming objects ( hi - def@xmath980.4 ) . the distributions of the difference for healthy ( solid histogram ) and deficient ( dashed histogram ) spirals are shown in the upper - right panel . the unshaded area indicates the range of @xmath11 on which the standard @xmath6 vs. @xmath7 relation for star - forming galaxies is usually calibrated.,width=321 ] the @xmath7 vs. @xmath3 relation for our optically selected sample . a(fuv ) is computed from our sed fitting technique ( filled symbols ) and using the standard recipe for active star forming galaxies ( empty symbols ) respectively . the solid and dashed lines indicate the relations between @xmath7 and @xmath69 proposed by salim et al . symbols are as in fig.[diffafuv_tau].,width=302 ] relation between the observed ( i.e. not corrected for internal extinction ) @xmath76 colour and @xmath11 for our sample . the dotted line indicates the best linear fit for galaxies having @xmath777 gyr . the shaded area indicate the typical dispersion of the data computed in fuv - h bins 0.5 mag wide . symbols are as in fig.[diffafuv_tau].,width=302 ] residuals in the estimate of @xmath7 obtained with the different methods described in sec . the @xmath6 ratio is estimated using iras observations ( upper panel ) and the @xmath69 colour ( lower panel ) respectively . the shaded areas show the typical uncertainties in the new recipes presented in this paper , estimated by generating 1000 random galaxies having @xmath76 colour following a gaussian distribution centered on the observed colour with @xmath58=0.2 mag ( see sec . 4 for details ) . for comparison , the residuals obtained when an age independent method is used are indicated by the filled symbols . symbols are as in fig.[diffafuv_tau].,width=302 ] difference in the estimate of @xmath7 with the sed technique described in sec.[data ] ( @xmath10913 gyr , a la sandage sfh and salpeter imf ) and various combination of sfh ( sandage or exponential ) , age ( variable or fixed to 13 gyr ) and initial mass function ( imf , salpeter or chabrier ) . the dashed lines indicates the 1@xmath58 difference in the estimate of @xmath7 . , width=320 ]

the observed properties of galaxies can be attributed to a wide variety of complex factors . in order to understand the environmental causes that might trigger intense star - formation in the local universe , such as that observed at high redshifts for lyman break galaxies , we explore two regimes : large - scale , to study the connection between halo mass and high star formation rates ( sfrs ) ; and small scale , calculating merger rates . at large scales , a galaxy s sfr may be connected to how the galaxy accretes gas from its large - scale environment , which depends on the halo mass . smaller scale environmental effects , from nearby galaxies , may induce star formation from interactions and eventual mergers . the study of galaxy evolution has been connected to that of larger structure formation their growth related to overdensities arising from primordial density fluctuations . studies of galaxy bias , the distribution of galaxies relative to underlying dark matter distribution , attempt to describe the fate and origin of present - day galaxies . the link between halo mass and star formation has been discussed and explored in several papers @xcite . a simple description involves gas flowing to centers of dark matter potentials , reaching sufficient density to radiate and cool , losing pressure support and settling into a disk , and then forming stars . the system is eventually accreted onto a larger structure . such accretion by clusters can shock - heat the gas into high temperatures , preventing new stars from forming . therefore , we expect that active star formation can occur for halo masses whose virial temperatures are less than 10@xmath8k , where shock - heating inhibits new star formation . redshift evolution of the link between halo mass and star formation , determined by large scale environmental effects , might explain observations such as the butcher - oemler @xcite effect . in this effect , clusters at high redshifts host more blue , star - forming galaxies than local clusters , which are predominantly inhabited by `` red and dead '' galaxies . similarly , using the two - point angular correlation function to measure clustering strength , @xcite find that uv - selected , star - forming galaxies have migrated from cluster regions , with high correlation lengths ( r@xmath9 ) , at high redshift to more isolated regions at low redshifts . from the view of hierarchical structure formation , higher redshift clusters have lower halo masses compared to low - z clusters , thus high - z clusters have virial temperatures low enough to allow star formation . adding to this framework , @xcite discuss the importance of hot vs. cold accretion in the simplified picture above , gas does not get heated to the virial temperature necessarily . massive galaxies draw gas via quasi - spherical , `` hot mode '' accretion , in which the gas does shock - heat to the virial temperature . however , less massive galaxies acquire gas from flow along filaments in `` cold mode '' accretion where the gas radiates away its gravitational energy and never reaches the virial temperature ( see @xcite and references therein ) . these simulations show that the large - scale galaxy environment influences gas accretion processes . therefore , a galaxy s halo mass contributes significantly to how it acquires gas , the fuel that affects its star formation efficiency . similar shock physics may affect the relation between small scale merger effects and star formation . @xcite predict that the merger of a halo with a larger structure , such as a cluster , suppresses star formation as gas is shock - heated to the virial temperature of the more massive halo and then cools onto the central object of the most massive halo . thus , these results suggest that the mass of the larger companion is a consideration for whether mergers might enhance star formation . according to the work of @xcite , mergers with clusters should not result in enhanced star formation . nevertheless , other studies show that at these small scales galaxy - galaxy interactions and close neighbors can stimulate star formation activity . in hierarchical structure formation , the merger process is an important ingredient for galaxy evolution . over thirty years ago , @xcite suggested that tidal forces during the merger process can funnel gas into the central regions of galaxies , triggering strong star formation activity . further theoretical studies have investigated the connection between mergers and high star formation rates ( see @xcite , @xcite , and @xcite ) . @xcite have shown that nearly half of the highest star - forming galaxies have close companions ( galaxies within a projected radius of 100 kpc ) , by correlating 10@xmath10 galaxies from the sloan digital sky survey ( sdss ) . according to this research , tidal interactions are the primary cause for enhanced star formation for these most active star - forming galaxies . several studies using various star formation indicators from h@xmath11 , co , far - infrared luminosities for interacting galaxies have shown that enhanced star formation results from galaxy interactions ( e.g. , @xcite , @xcite , and references therein ) . using the galaxy evolution explorer ( _ galex _ ) , @xcite have uncovered a population of local , intensely star - forming galaxies . these galaxies are selected based on their fuv luminosities ( l@xmath12 @xmath13 2@xmath1410@xmath15 ) . the most compact of these uv - luminous galaxies ( uvlgs ) , with i@xmath16 l@xmath2 kpc@xmath17 , share many similar properties with z@xmath73 lbgs : specific sfrs , metallicities , morphologies , kinematics , and attenuations @xcite . we refer to these systems as lyman break analogs ( lbas ) , to discriminate this special subset of uvlgs from lower surface brightness uvlgs . given that lbgs are found at higher redshifts ( z@xmath182 - 3 ) where major and minor mergers are generally more prevalent than for z@xmath180.1 lbas , lbas and lbgs may be expected to differ in their recent merger histories , environments , and location within large scale structure . in this paper , we study the environments , as quantified by correlation functions , of low to intermediate redshift ( 0.4@xmath0z@xmath01.2 ) uvlgs in the cosmos field . we extend our analysis to lbas to compare with z@xmath73 lbgs , although this analysis is limited because of the rarity of z@xmath01 lbas . we measure two - point auto - correlation functions ( acfs ) and cross - correlation functions ( ccfs ) of uvlgs with other galaxy samples to probe the mass of the dark halos within which uvlgs reside . we connect these observations with those of lbgs at high redshift ( z@xmath73 ) , which are known to strongly cluster @xcite . recent hubble space telescope ( _ hst _ ) images of z@xmath70.2 lbas show disturbed morphologies , signs of mergers and interactions @xcite . we relate small - scale correlations to pair fractions in order to estimate merger rates . studying the locations of current uvlgs , as examples of extreme star - forming galaxies , we explore the relationship between high rates of star - formation , host halo mass , and merger rate as a function of redshift . in the following section , we discuss our sample selection . in the three subsequent sections , we outline the theory ( [ sec : theory ] ) and techniques for measuring the angular correlation functions([sec : meas ] ) , and derivations of physical quantities ( [ sec : quant ] ) used to analyze our data . we discuss our results for the large and small scale environments of uvlgs and lbas in [ sec : results ] , and discuss these in the context of other environmental studies in [ sec : disc ] . a summary of our analysis can be found in the final section : [ sec : conc ] . throughout this paper we adopt the cosmology given by ( h@xmath9 , @xmath19 , @xmath20 ) = ( 70 km s@xmath3 mpc@xmath3 , 0.3 , 0.7 ) . unless explicitly stated , our number densities are in physical , proper units . correlation lengths are quoted in comoving units of h@xmath3mpc . with its large contiguous field and deep observations , the cosmos survey provides excellent data for environmental studies . the cosmos field covers 2 deg@xmath21 , centered on @xmath22 , @xmath23 . this region of sky has rich multiwavelength data , including : optical observations with the superb subaru ground - based telescope , atop mauna kea , hawaii , ultraviolet coverage from _ galex _ , radio observations from vla , and xray data from xmm . the inner 1.7 deg@xmath21 has been imaged by _ hst _ , using the advanced camera survey ( acs ) in the f814w filter ( see @xcite for details ) . deep photometric data from subaru provides photometric redshifts with @xmath24 , depending on i - band magnitude and redshift @xcite . from the i - band selected ( i@xmath025.5 ) cosmos photometric redshift catalog ( for details about this catalog see @xcite ) , we compare uvlg and lba samples , which draws from the combined cosmos+_galex _ data ( i@xmath025.5 , and nuv@xmath023 , see @xcite for more details ) , with three separate galaxy samples ( selected from the cosmos sample , with i@xmath025.5 ) for 0.4@xmath0z@xmath01.2 : ( 1 ) a volume limited sample in m@xmath25-``all '' , ( 2 ) blue luminous galaxies- `` blg '' , and ( 3 ) luminous red galaxies- `` lrg '' . we briefly note here while the `` all '' sample is volume - limited in m@xmath25 , it is not volume - limited for all colors . this is discussed in more detail further below . figure 1 shows the various cuts applied to the data to create these separate samples : all is comprised of galaxies with -24@xmath26 - 18 ; blg galaxies have @xmath27 - 20.0 and @xmath28 1.8 color ; and lrg set is selected by a luminosity cut , @xmath29 - 20.5 and color cut , @xmath30 @xmath13 2.0 . both colors and luminosities are given in rest - frame , k - corrected ab magnitudes . k - corrections were done using the _ kcorrect_4.1.4 _ idl routine @xcite . the spectral energy distribution ( sed ) fitting was done using the cosmos optical bands as well as the _ galex _ fuv and nuv bands , where available since the comparison samples are not required to be uv - detected . we modified this routine to use the closest band to the rest - frame filters for determining kcorrections for example , at z@xmath180.5 , the observed nuv magnitude at 2300 best approximates the rest - frame fuv magnitude at 1500(i.e , @xmath31 ) . the uvlgs are selected using their k - corrected , rest - frame fuv properties from the cosmos+_galex _ catalog : fuv @xmath32 - 19.54 , corresponding to the luminosity criterion of l@xmath33 @xcite . to estimate the fuv surface brightness , we use the value of r@xmath34 output from sextractor @xcite for the acs i - band data ( see @xcite for further detail ) , which corresponds to the half - light radius . we use this acs i - band half - light radius as a proxy for the uv size , acknowledging that the i - band at these redshifts corresponds to optical wavelengths . our sample of lbas were determined by the criteria : i@xmath35 l@xmath36 kpc@xmath17 . while the i - band size may overestimate the uv size by a factor of 2 ( @xcite , thereby , underestimating the fuv surface brightness by a factor of 4 ) , we find that @xmath37% of uvlgs are lbas , consistent with what is found at z@xmath7 0.2 @xcite . we summarize the sample sizes for each redshift bin in table 1 , and show the redshift and spatial distributions of these samples in figures 2 and 3 , respectively . the `` l '' shape of the color - magnitude diagrams in figure 1 indicates that the red and blue galaxies do not populate the magnitude space in the same way . the majority of red galaxies tend to be more luminous than blue galaxies ; the faint end for the red galaxies starts decreasing around m@xmath38 - 19 . while the volume - limited cut shown on the upper left panel of figure 1 selects a complete sample in m@xmath39 , the color - magnitude diagram at m@xmath40 - 19 shows a diminishing sample of red galaxies . the luminosity function ( lf ) for red galaxies shows that the faint end does have a decreasing slope ( @xcite , but our lf comparison of the red galaxies in the all sample with the theoretical lf does indicate sample incompleteness at higher redshifts . this incompleteness does not affect either the blg or lrg samples which are selected to be sufficiently bright to avoid magnitude limits . however , the all magnitude cut of m@xmath38 - 18 suffers from some incompleteness in the red population . this observation is worth noting for interpreting and comparing the all sample to other samples and studies , but does not affect the overall conclusions of this paper . the predicted n(z ) , shown as a black dotted line in figure [ fig : zdist ] , is derived using the luminosity functions observed by @xcite for the all , blg and lrg samples and by @xcite for the uvlg sample ( this is discussed further in [ sec : quant ] ) . the redshift distributions in figure 2 roughly follow the smooth predicted distribution marked by the black dotted line . the uvlg , as well as the lba , redshift distributions may be too noisy , because of poor number statistics , to interpret their shape . the overdensities apparent in the spatial maps ( figure [ fig : skydist ] ) correspond to known physical structures observed in the cosmos field ( see @xcite , @xcite , @xcite for more discussion of these structures ) . in the standard framework for defining the two - point correlation function , galaxies are distributed in space such that the probability , @xmath41 , of finding a galaxy in some volume , @xmath42 is , @xmath43 , where _ n _ is the mean number density of galaxies . extrapolating to the situation of finding two galaxies at some comoving separation , r@xmath44 , one within volume , @xmath45 and another within volume , @xmath46 , is then @xmath47 \label{eqn : xi}\ ] ] where @xmath48 is defined as the two - point correlation function . an approximation for the correlation function , @xmath49 , is a power - law with a single slope at all scales : @xmath50 where z refers to redshifft , @xmath51 and @xmath52 specify the slope and comoving clustering strength , respectively . given a uniform , poissonian distribution , @xmath53 , while correlated data ( @xmath54 ) means that there is an increased probability of finding a second galaxy within distance r of the first . various studies ( i.e. , @xcite for sdss galaxies and @xcite for high - z lbgs ) have found departures from a power - law with a single slope . rather , they find that the correlation function is better approximated by the sum of two contributions : the `` one - halo term '' , dominating at small scales for galaxy pairs residing within the same halo , and the `` two - halo term '' , dominating at large scales for galaxy pairs within separate haloes . this so - called hod modeling is beyond the scope of this paper . however , to account for the fact that the galaxy correlation function might not follow a power law with a single slope , we assume that @xmath55 can be fit by two power laws with different slope for small scales and for large scales . for data based on photometric redshifts , the three dimensional correlation function can not be measured directly , but it can be inferred from the projected two - dimensional angular correlation function . by analogy of the steps outlined above , the three - dimensional , real space @xmath56 is collapsed into a two - dimensional function : w(@xmath57 ) , which measures the excess probability that two galaxies are found within some angular separation , @xmath57 . @xmath58\delta\omega_1\delta\omega_2 \label{eqn : wtheta}\ ] ] where _ n _ is the surface density and @xmath59 and @xmath60 are the solid angle elements describing our field . like @xmath61 in eqn . [ eqn : xiform ] , w(@xmath57 ) is also expected to have a similar power law form : @xmath62 we convert the three - dimensional correlation function into the angular one , using the formalism introduced by @xcite and @xcite . therefore , we relate the comoving correlation length , @xmath63 , to @xmath64 and @xmath51 to @xmath65 by the following : @xmath66 where @xmath67 , @xmath68 is the comoving distance , @xmath69 and @xmath70 are the number densities , and f(r ) is the curvature term from the robertson - walker metric , @xmath71 } \ ] ] using the @xcite form for the evolution of @xmath52 : @xmath72 where r@xmath73 is the comoving correlation length at z@xmath180 , and @xmath74 gives the evolution of clustering ( i.e. , @xmath75 specifies a stable state , @xmath76 describes no evolution , and @xmath77 corresponds to the linear growth of fluctuations ) . throughout our analysis , we assume no evolution , or @xmath76 , which allows us to simplify eqn . [ eqn : r0 ] . over the relatively narrow redshift ranges for which we are integrating over , we can justify this assumption @xcite find that r@xmath9 evolves by @xmath78 20% between z@xmath180.2 and z@xmath181.2 for late type galaxies and @xmath78 50% for early type galaxies . auto - correlations compare distributions of a sample with itself , while cross - correlations compare distributions of one set of objects with a separate sample . naturally , in the auto - correlation cases , @xmath79 in eqn . [ eqn : r0 ] . we infer the real - space correlation length , r@xmath9 and power - law slope , @xmath51 , by fitting the angular correlation function . along with slope and amplitude parameters , we additionally calculate the integral constraint ( ic ) correction . in principle , our observed field may be superimposed on a void or overdensity , such that the observed number density in our field will not match the average number density in the universe . in theory , the integral in eqn . [ eqn : wtheta ] should average to 0 over all scales , i.e. , since there are n galaxies in the universe , excess at small scales must be balanced by deficits at larger scales . we correct the observed angular correlation function by the ic factor defined in @xcite . the angular correlation function , w(@xmath57 ) , measures the excess of clustering compared to a uniform , random distribution . to simulate the random distribution appropriately , we first create masks from the cosmos+_galex _ fields . these masks mark the locations where bright , saturated stars and field edges prevent galaxies in those locations from appearing in the catalog . the black regions in fig . [ fig : skydist ] display masked locations . populating the allowed , un - masked regions with random galaxies , we create the random set . the histogram of separations in the random set and data sets are referred to as rr(@xmath57 ) and d@xmath80d@xmath81(@xmath57 ) , respectively , where the subscripts 1 and 2 denote two separate data samples ; d@xmath80d@xmath82dd in the auto - correlation case . similarly , the separations between galaxies in the random and data sets are represented by d@xmath80r and d@xmath81r , or simply dr for autocorrelations . it has been shown that the landy - szalay @xcite estimator is one of the most robust ways of measuring w(@xmath57 ) , reducing the edge effects of finite fields . here we give the expression for the angular correlation function that we use to measure auto - correlations ( eqn . [ eqn : auto ] ) and cross - correlations ( eqn . [ eqn : cross ] ) : @xmath83 where r and n refer to the number of galaxies in the random and data samples ( the subscripts 1 and 2 for n@xmath80 and n@xmath81 distinguish the separate samples , in the cross - correlation case ) . our angular correlation function errors have been derived from boot - strapping , as described in @xcite . cosmic variance adds another important source for error . however , unlike the bootstrap errors which are scale dependent , the errors from cosmic variance shift the entire correlation function . using @xcite to calculate cosmic variance errors for our samples , we show these errors in our correlation function figures as shaded regions ( for example , see figs . [ fig : fit]-[fig : acor_comp ] ) . as described in @xmath84[sec : data ] , our samples span the redshift range 0.4@xmath0z@xmath01.2 . while we may reduce statistical noise by choosing the largest sample possible selecting galaxies in this entire redshift span , correlating data across such a wide redshift range has disadvantages . by correlating data in two - dimensions , which are nt physically correlated in three - dimensional , de - projected space , we risk weakening the actual correlations . therefore , we select two bin sizes : dz@xmath180.2 ( resulting in 4 overlapping bins : 0.4@xmath10.6 , 0.6@xmath10.8 , , 1.0@xmath11.2 ) and dz@xmath180.4 ( 2 bins : 0.4@xmath10.8 , and 0.8@xmath11.2 ) . comparing the angular correlation functions for different redshift bins requires converting the angular correlation function into physical de - projected units , by the following reasoning : @xmath85^{1/2 } ) d}\pi \\ & = & { \rm \frac{\frac{dn(r_p)}{da}-\left < \frac{dn}{da}\right > _ 0}{\left < \frac{dn}{dv}\right > _ 0}}\end{aligned}\ ] ] where r@xmath86 is the transverse distance and @xmath87 is the line - of - sight separation ; @xmath88 , @xmath89 , and @xmath90 are , respectively , dn / da , dn / d@xmath91 and comoving number density for a uniform , non - clustered distribution . finally , multiplying the w(@xmath57 ) equation ( eqn . [ eqn : wthetadef ] ) by d@xmath91/da and averaging these expressions over the redshift bins , we derive @xmath92 where d@xmath93 is the angular diameter distance . as such , we can compare how the correlation function of the narrower bin sizes compare to the wider redshift ranges . [ fig : acor_all ] shows how the wide bins differ from the narrow ones for samples with 0.8@xmath0z@xmath01.2 ; we plot the ratio of the correlation functions for the wide to narrow bins , displaying both w(@xmath57 ) , top , and w@xmath86(r@xmath86 ) , bottom . since the discrepancy was not severe , we decided that the wider redshift bins would provide better statistics . given the low abundance of uvlgs , their auto - correlation only proved possible for the full 0.4@xmath0z@xmath01.2 sample ; and for the lbas , their auto - correlation proved too noisy to be useful , and we were only able to attain reasonable results by cross - correlating with the largest sample : the all sample with dz@xmath180.4 bins . for each sample , we calculate separate fits to the angular correlation function in two regimes : large scales ( @xmath94kpc ) and small scales ( @xmath95kpc ) . we create a grid of a@xmath96 , spaced logarithmically , and delta , linearly spaced , values , with the best fit values corresponding to the minimum @xmath97 . the errors on this fit are derived as the values corresponding to min@xmath98 , and we make certain that the grid values extend significantly past these error values to avoid restricting the errors within the minimum or maximum value . the large scale is fit , additionally to the slope and amplitude , with the ic correction , discussed in the previous section . we show examples of our fitting to the separate samples in fig . [ fig : fit ] . the number density is required for calculating both limber s equation and the pair fraction . for our analysis , we use the observed number density . however , as a check we calculate the predicted physical number densities for the all , blg , and lrg samples . we use @xmath11 , m@xmath99 ( given in vega magnitudes , and converted to ab magnitudes using @xcite ) , and @xmath100 given for the combo-17 data in @xcite to determine the luminosity function , interpolating the lf between the specified redshift - dependent values to derive a smooth n(z ) across our entire redshift range . we use a gaussian kernel with @xmath101 , simulating the photo - z errors , to convolve the derived n(z) the results are shown as black dash - dotted lines in fig . [ fig : zdist ] , compared to the observed numbers . with the exception of some peaks and valleys , the observed blg and lrg samples follow the predicted values well . the agreement between the predicted and observed all distribution is poorer at higher redshifts the observed number density appears to be underestimated . it is possible that the peaks in the observed n(z ) in the lrg sample correspond to real structure . for the uvlg and lba samples , we predict the number density using the calibrated luminosity function from @xcite . in table [ tab : lffit ] , we list the values used to fit the luminosity function to derive n(z ) for all of our samples . for use in limber s equation , we smooth the observed redshift distribution , performing a boxcar average with a kernel size : @xmath102z@xmath180.08 . the discrepancy between results derived using the observed versus predicted number densities is negligible . quantifying the smaller scales , we relate the fits from the inner regions ( shown in red in fig . [ fig : fit ] ) to pair fractions and merger rate , similar to analysis done by @xcite for massive , luminous galaxies . the pair fraction is derived from integrating eqn . [ eqn : xi ] from 0 to some maximum radius , within which galaxies are assumed to eventually merge the pair fraction quantifies the number of galaxies that are found within some critical , merging radius of another galaxy . however , since we are interested in the physical separation between the galaxies rather than the comoving distance , we replace the comoving correlation length , @xmath103 , with the physical correlation length , @xmath104 . therefore , we calculate pair fractions by simplying the integral of eqn . [ eqn : xi ] with the following assumption : @xmath105 , into this expression @xmath106 here , @xmath107 is the number density of the second sample in the cross - correlation . following @xcite , we also use r@xmath108=30 kpc to define close , physical pairs . our errors for the pair fraction include cosmic variance uncertainty , estimated using the relations from @xcite . while referred to as a probability , the pair fraction is the mean number of galaxies found within r@xmath108 ( see @xcite ) and may exceed unity . the pair fraction has quantitative power in comparing small scale clustering between different samples . the merger rate is derived by dividing the pair fraction by the timescale for this merging event to occur ( assuming that the close pair merge into one merger remnant , such that the pair fraction is equal to twice the merger fraction ) . though the calculation is uncomplicated , the steps for deriving the merger timescale are quite complex and fraught with uncertainties . @xcite have compared pair fractions derived from their mock catalogs with merger rates from the millenium simulations to determine merger timescales . from these results , we infer that the merger timescale is @xmath109 myr h@xmath3 , using the most general cross - correlation case of uvlgs with all and estimating @xmath0 m@xmath110 m@xmath36 @xcite in the range z@xmath180.4@xmath11.2 . using this value , we determine merger rates for our data , shown in table [ tab : pf ] . according to @xcite , merger timescales have been underestimated by @xmath111 in previous studies for example , @xcite estimate the merger timescale to be 400 myr . to avoid the large uncertainties in the merger timescale , we compare our pair fraction results to other studies , rather than comparing merger rates . given the fits shown in fig . [ fig : fituvlg ] , we derive the correlation strength ( r@xmath9 ) for uvlgs . we determine that uvlgs have r@xmath4 h@xmath3mpc ( see table [ tab : lsauto ] ) , corresponding to m@xmath112 ( see @xcite ) . figure [ fig : r0_uvlg ] shows how our results compare with other published data , at various redshifts . the shaded region in this figure corresponds to the shaded arrow in figure 11 of @xcite this marks how r@xmath9 evolves with redshift for a halo of mass 10@xmath113m@xmath2 ( bottom of shaded region ) to 10@xmath114m@xmath2 ( top of shaded region ) . we discuss the evolution of uv - selected galaxies , in relation to other samples , in the following section . studying the autocorrelations of the separate samples with each other , we find that typically ( at all z ) lrgs cluster most strongly with themselves at small scales , compared to auto - correlations of blgs or all samples ( top panel of fig . [ fig : acor_comp ] ) . comparing auto - correlations of the samples with cross - correlations with uvlgs in the bottom panels of fig . [ fig : acor_comp ] , we find that at small scales ( r@xmath115kpc ) , uvlgs cluster less strongly with lrgs compared to lrgs with themselves ( _ i.e. _ @xcite ) ; however , it is difficult to differentiate the uvlg cross correlations with the all and blg samples compared to their auto - correlations . we compare the pair fractions of the autocorrelation of the other three samples with their cross - correlation with uvlgs in fig . [ fig : comp]100% of the galaxies are found in pairs . rather the pair fraction represents the mean number of secondary galaxies likely found near the primary all galaxy . see discussion in [ sec : quant ] . ] . at the smallest scales the all sample pairs more with uvlgs than with themselves , and given the large errors , the blg sample galaxies appear to pair with uvlgs with equal likelihood to pairing with themselves . yet , uvlgs are not likely companions of lrgs . finally , the lba - all ( purple ) pair fraction in table [ tab : pf ] has excessively large errors , preventing a meaningful comparison of the lba - all sample with these other cases . as we have discussed , small scale clustering translates to merger rates suggesting that uvlgs appear to merge often with all sample galaxies ( and possibly , with other blue , star - forming galaxies ) , and do not merge often with lrgs . however , as there are some small scale pairs of uvlg - lrgs , these systems do interact on some occasions . given that blue , star - forming galaxies typically do not reside in clusters , and uvlgs merge with these galaxies as often as they do with themselves , we expect that uvlgs will inhabit lower mass halos rather than massive clusters further constraining our results . comparing with other results , we show our pair fraction results for the all , blg , and lrg samples with other studies in fig . [ fig : pfpub ] . @xcite observe a sample of m@xmath116 galaxies ( orange squares ) to measure merger fractions , assumed to be half of the pair fractions since two pair galaxies interact in a single merger ; @xcite study pair fractions for galaxies with @xmath117m@xmath118 ( green triangles ) , and the magenta star marks the result for m@xmath119 , found by @xcite . since all of these samples are magnitude limited and color - selection independent , the all sample is most comparable ( m@xmath120 ) . it should be noted that many studies ( e.g. @xcite , @xcite , @xcite ) find that pair fractions are extremely sensitive to the limiting magnitude . in order to compare our pair fraction results with other work , we adopt the observation from @xcite that the correlation parameters are less sensitive to limiting depth than the number density . therefore , we modify the derived pair fraction for our all sample , using eqn . [ eqn : pf ] with the small - scale correlation function parameters in table [ tab : acor ] , but changing the number density to represent m@xmath121 selection . the modified pair fraction is shown as open circles in fig . [ fig : pfpub ] and agree well with published results . the picture that has emerged from this analysis depicts uvlgs as galaxies that inhabit halos with m@xmath122 . uvlgs are more likely to interact and eventually merge with other galaxies ( all ) than the all sample does with itself , and they are as likely to interact and have companions as other blue , star - forming galaxies . these results are consistent with other studies showing that high sfrs may be closely related to increased merger rates . uvlgs are unlikely to interact with lrgs , which are expected to reside at the centers of massive clusters , providing another constraint regarding uvlg halo masses . in this section , we first place uvlgs amongst other categories of star - forming galaxies ; then , we discuss our findings in light of other work related to uv - selected galaxies , star - formation and environments . in figure [ fig : r0_uvlg ] , we display the large scale environmental measure for several other classes of star - forming galaxies . the bm , bx , lbgs @xcite are selected by color and limiting magnitude to be both luminous and blue , exhibiting the lyman break in their spectra . as discussed in the introduction , the lbas have been found to have similar colors and specific star formation rates ( sfr / m@xmath123 ) , suggesting similar star formation histories and stellar populations . in @xcite , we found evidence for recent and continuous star formation activity , similar to the results found for z@xmath124 lbgs by @xcite and @xcite . we argue that lbas are most comparable to lbgs . the fuv@xmath125 sample by @xcite is a less extreme sample ( extending to fainter magnitudes ) of uv - selected galaxies compared to uvlgs . the other star - forming samples ( blgs , @xcite , @xcite , @xcite , @xcite ) are selected based on color and magnitude , and span a larger range of sfrs and stellar masses . as demonstrated in @xcite for the local universe , the correlation strength appears to decrease with increasing uv luminosity . yet , blue galaxies appear to have similar or lower r@xmath9 than the _ galex _ sample , including , the uvlg sample . @xcite finds that around a critical stellar mass of @xmath126 m@xmath36 , galaxies separate into two classes : massive , red , with low sfrs versus less massive , blue , with high sfrs . the typical uvlg stellar mass is @xmath127m@xmath36 ( lbas have @xmath128m@xmath129)@xcite . while extremely star - forming and blue , these galaxies have masses placing them on the boundary between these two classes suggesting , that they may be in the process of transitioning . this in - between stage is not well - understood as yet , but mergers and environmental processes may drive the transition in this regime . the more massive , infrared counterpart to uvlgs are luminous infrared galaxies ( lirgs ) and the more extreme ultra - luminous infrared galaxies ( ulirgs ) dusty star - forming galaxies with significant emission coming from central 1kpc . they experience significant merger events in the local universe @xcite . as yet , little research has been completed on studying the environments of local ulirgs . according to @xcite , mostly z@xmath130 ulirgs occupy the field environment , with few cases having cluster densities of abell richness class 0 or 1 . studying lirg pair fractions at z@xmath131 , @xcite find that @xmath132 of the ir luminosity density arises from close pairs , while morphologically classified mergers contribute to @xmath133 thereby , @xmath134 of the ir luminosity density can be attributed to early interactions ( as evidenced by close pairs ) and mergers ( morphologically determined ) . while uvlgs are not dusty , it seems that uvlgs and lirgs / ulirgs share environmental conditions , possibly contributing to their high sfrs . @xcite investigate the relationship between star formation , stellar mass and halo mass . assuming that a tight relationship between stellar mass and halo mass exists , based on observational `` abundance matching '' , they theoretically derive how mass assembly and star formation history evolve with time . based on averaged results ( they caution that exceptions to their results are completely plausible ) , @xcite determine that the relationship between star formation rate ( sfr ) and halo mass is gaussian - shaped , whose mean and normalization parameters evolve slightly with redshift . however the broadness of the sfr peak , spanning 11.9@xmath13513.1 for z@xmath70 , suggests that a distinct characteristic halo mass leading to higher sfrs does not exist . rather galaxy properties likely shift gradually across the halo mass scale . as discussed earlier , @xcite suggest two regimes of gas fueling and star formation efficiency , depending on halo mass : hot - mode accretion for massive galaxies and halos masses @xmath136 m@xmath36 which relate to lower sfrs , and cold - mode accretion in less massive galaxies , where gas is drawn along filamentary structure and able to feed star formation more efficiently . both of the results from @xcite and @xcite imply that the range of allowed halo masses for high sfrs could be broad with the possibility that there is some mixing of both hot- and cold- accretion modes for halos with masses , m@xmath137 . while the study by @xcite has been very thorough for z@xmath02 , they do allow that the situation might differ at higher redshifts . as illustrated by @xcite , at high redshifts observations show a strong correlation between rest - frame uv - luminosity and correlation length , or halo mass @xcite . yet , their results show that the nearly linear relation between uv - luminosity , or sfr , and halo mass seems to level off , or possibly even reverse at low redshifts @xcite . although we have studied a rare sample of galaxies , their close relation to well - studied high redshift lbgs allows us to make a direct comparison about the halo mass condition for lbg - like properties . @xcite find that lbg halo masses are @xmath138 . the uvlgs have halo masses that are consistent with these values . while the errors are admittedly too large to be conclusive , our result hints that the halo masses in extreme star - forming galaxies , such as lbgs and uvlgs , are consistent with masses @xmath139 . at large scales , we measured halos masses of uvlgs , in an attempt to connect star formation properties with large scale processes ; at small scales another important environmental quantity is measured the likelihood for mergers . it has been shown by @xcite that halos do not necessarily merge in the same way that galaxies merge . at small scales , the more relevant quantity is the halo occupation distribution ( hod ) , or the way in which galaxies occupy their halos . this information constrains galaxy formation models at the @xmath32 100 kpc scales . studying their small scale clustering , we find that uvlgs and lbas tend to have high pair fractions and merger rates with the all sample . images of the lbas offer similar clues , displaying companions or ongoing interactions ( such as tidal features and plumes ) in the high resolution images @xcite . consistent with the results found by @xcite , the most extreme star - forming galaxies are preferentially found in close pairs . studying lbgs at high redshifts ( z@xmath140 ) , @xcite also find evidence for small - scale clustering . they interpret that lbgs likely have fainter companions occupying the same halos , and find that the hod model fits their data well . using merger histories to connect z@xmath124 lbgs to present - day massive galaxies , @xcite argue that m@xmath141 galaxies undergo @xmath7 4 major mergers , with most of the interactions having occured before z@xmath181.5 . furthermore , they show that z@xmath180.5 galaxy with stellar mass @xmath142 likely originated as a 10@xmath143 galaxy , whose stellar mass has grown both by mergers and increased star formation activity . while the number of mergers plateaus for @xmath144 galaxies by z@xmath72 , less massive galaxies seem to continue merging until z@xmath180.5 . uvlgs have stellar masses @xmath145m@xmath36 , and lbas have lower masses ( @xmath146 m@xmath36 ) . if uvlgs follow the same merger histories and evolution as more general galaxies studied by @xcite , the present uvlg and lba masses suggest that their z@xmath124 predecessors had stellar masses between @xmath147 m@xmath36 and are probably still actively merging at z@xmath148 . could mergers be the main triggers for the increased star formation in these low dust , high star - forming galaxies ? our results , and more generally , the conclusion reached by @xcite find that interactions and mergers do lead to enhanced star formation . as discussed in the introduction , this concept is not new having theoretical roots from @xcite , mergers efficiently funnel gas into central regions of galaxies , providing fodder for star formation . this theory is supported by several observational studies ( see reviews by @xcite and @xcite ) . we can also reverse the query : do all mergers and interactions create uvlgs or lbgs ? mergers and interactions do enhance star formation , but these processes are also seen in lirgs and ulirgs which are dustier systems , apparently different from uvlgs and lbas . the overlap between these ir star - forming populations and uv - star - forming galaxies has not been studied in detail . we speculate that both of these systems may be related , as different phases of the merger process with various unknown factors such as orientation , timescales , and kinematics adding complexity to the evolution and final merger state ; or , while some fraction of mergers result in uvlgs , some other fraction result in dustier lirgs or a separate class of galaxies , altogether . @xcite has found that a larger fraction of lirgs appear in the merger phase than in the pre - merger , close pair phase . so it is plausible that uvlgs relate to the earliest interaction phase , exhibiting high sfrs , low dust and distorted features . the highest restframe - uv surface brightnesses , observed in lbas , corresponds to galaxies found to have a higher incidence of pairs . unfortunately , further investigation about their environment is beyond the scope of this statistical study because of their rarity . extrapolating from our data on local lbg analogs , it seems that close pairs are one relevant ingredient to their condition . however , it is unlikely that every merger will result in a low dust , high surface brightness , high sfr galaxy other ( unknown as yet ) constraints must also be placed in order to produce this particular breed of galaxy . using the cosmos data to compare angular cross - correlation functions of uvlgs with three other samples : all , blgs , and lrgs , we determine both large and small scale environments of uvlgs . in our analysis , we assume the following : 1 . ) the correlation function does not evolve with redshift . the full correlation function includes a term from evolution ( see eqns . [ eqn : xiform ] and [ eqn : r0z ] ) . we assume that @xmath76 , which describes non - evolution in the correlation function . this assumption is justified by the study by @xcite finding that r@xmath9 evolves by @xmath78 20% over z@xmath180.2@xmath11.2 range . the correlation function can be characterized by a power law at large scales , and a separate power law at small scales ( see @xcite ) . 3.)the size of the redshift bins do not affect correlation results . bins that are too wide might make real - space correlations appear weaker by projecting across an excessively wide redshift range . however , we assume that dz@xmath180.4 is not too wide , and check this assumption by comparing to more narrow dz@xmath180.2 bins finding that there is good agreement between the wide and narrow cases . we find that uvlgs inhabit halos with masses exceeding @xmath149m@xmath36 . this result is further constrained by the additional observation that uvlgs and lrgs ( which typically trace the most massive halos ) do not often form pairs . rather , uvlgs form close pairs with higher probability compared to the all sample forming pairs with themselves ; and uvlgs pair with blgs with the same probability as blg - blg pairs . therefore we conclude that mergers are relevant environmental effects responsible for triggering high levels of sfr , observed in uvlgs and the high redshift lbgs . we are not able to definitively measure the halo mass of this sample given our large errors . however , we explore the possibility that halo mass may play some role in the uvlg or lbg condition since both examples of uv - selected , high sfr galaxies have halo masses consistent with @xmath150m@xmath36 . ( galaxy evolution explorer ) is a nasa small explorer , launched in april 2003 . we gratefully acknowledge nasa s support for construction , operation , and science analysis for the galex mission , developed in cooperation with the centre national detudes spatiales of france and the korean ministry of science and technology . the hst cosmos program was supported through nasa grant hst - go-09822 . more information on the cosmos survey is available at http://www.astro.caltech.edu/cosmos . we thank michael blanton for access to the idl kcorrect ( version 4.1.4 ) analysis package . this work has greatly benefitted from the careful comments and suggestions made by the anonymous referee . a.r.b . gratefully recognizes ian mcgreer and andrei mesinger for their contributions to this analysis , and david hogg for insightful discussions . ccc all & 0.4 @xmath1 0.8 & 60872 + all & 0.8 @xmath1 1.2 & 107473 + all & 0.4 @xmath1 1.2 & 168345 + blg & 0.4 @xmath1 0.8 & 7678 + blg & 0.8 @xmath1 1.2 & 25100 + blg & 0.4 @xmath1 1.2 & 32778 + lrg & 0.4 @xmath1 0.8 & 11175 + lrg & 0.8 @xmath1 1.2 & 15861 + lrg & 0.4 @xmath1 1.2 & 27036 + uvlg & 0.4 @xmath1 0.8 & 752 + uvlg & 0.8 @xmath1 1.2 & 2737 + uvlg & 0.4 @xmath1 1.2 & 3489 + lba & 0.4 @xmath1 0.8 & 50 + lba & 0.8 @xmath1 1.2 & 232 + lba & 0.4 @xmath1 1.2 & 282 + cccccc all & 0.3 & -1.3 & -21.10 & 32.26 & @xcite + all & 0.5 & -1.3 & -21.30 & 33.32 & @xcite + all & 0.7 & -1.3 & -21.62 & 32.16 & @xcite + all & 0.9 & -1.3 & -21.35 & 41.26 & @xcite + all & 1.1 & -1.3 & -21.48 & 34.72 & @xcite + blg & 0.3 & -1.3 & -20.84 & 24.26 & @xcite + blg & 0.5 & -1.3 & -21.20 & 23.20 & @xcite + blg & 0.7 & -1.3 & -21.40 & 27.27 & @xcite + blg & 0.9 & -1.3 & -21.20 & 37.32 & @xcite + blg & 1.1 & -1.3 & -21.35 & 29.19 & @xcite + lrg & 0.3 & -0.5 & -20.73 & 21.91 & @xcite + lrg & 0.5 & -0.5 & -20.87 & 19.97 & @xcite + lrg & 0.7 & -0.5 & -21.20 & 17.75 & @xcite + lrg & 0.9 & -0.5 & -21.28 & 11.89 & @xcite + lrg & 1.1 & -0.5 & -21.68 & 5.32 & @xcite + uvlg & 0.3 & -1.5 & -19.53 & 6.00 & @xcite + uvlg & 0.6 & -1.5 & -19.63 & 10.30 & @xcite + uvlg & 1.0 & -1.5 & -19.89 & 8.00 & @xcite + [ tab : lffit ] ccccc all & 0.4 @xmath1 0.8 & 5.0 @xmath151 & 0.64 @xmath152 & 3.3 @xmath153 + all & 0.8 @xmath1 1.2 & 22.0 @xmath154 & 0.32 @xmath155 & 3.8 @xmath153 + blg & 0.4 @xmath1 0.8 & 16.6 @xmath156 & 0.44 @xmath157 & 3.9 @xmath158 + blg & 0.8 @xmath1 1.2 & 50.0 @xmath159 & 0.22 @xmath160 & 4.4 @xmath161 + lrg & 0.4 @xmath1 0.8 & 2.7 @xmath162 & 0.83 @xmath163 & 2.9 @xmath164 + lrg & 0.8 @xmath1 1.2 & 22.3 @xmath165 & 0.43 @xmath166 & 4.8 @xmath167 + uvlg & 0.4 @xmath1 1.2 & 32.8 @xmath168 & 0.32 @xmath169 & 4.8 @xmath170 + [ tab : lsauto ] cccccccc uvlg x all & 0.4 @xmath1 0.8 & 0.06 @xmath171 & 1.5 @xmath172 & 1.3 @xmath173 & 0.084 @xmath174 & 0.47 @xmath175 & 0.109 @xmath176 + uvlg x all & 0.8 @xmath1 1.2 & 0.03 @xmath177 & 1.5 @xmath178 & 1.3 @xmath179 & 0.159 @xmath180 & 1.18 @xmath181 & 0.276 @xmath182 + uvlg x blg & 0.4 @xmath1 0.8 & 0.07 @xmath183 & 1.5 @xmath184 & 1.4 @xmath185 & 0.011 @xmath186 & 0.07 @xmath187 & 0.015 @xmath188 + uvlg x blg & 0.8 @xmath1 1.2 & 0.24 @xmath189 & 1.2 @xmath172 & 1.9 @xmath190 & 0.039 @xmath191 & 0.14 @xmath192 & 0.032 @xmath193 + uvlg x lrg & 0.4 @xmath1 0.8 & 0.21 @xmath194 & 1.2 @xmath195 & 1.6 @xmath196 & 0.015 @xmath186 & 0.03 @xmath197 & 0.007 @xmath198 + uvlg x lrg & 0.8 @xmath1 1.2 & 0.16 @xmath199 & 1.2 @xmath200 & 1.8 @xmath201 & 0.023 @xmath202 & 0.11 @xmath203 & 0.025 @xmath204 + lba x all & 0.4 @xmath1 0.8 & 0.22 @xmath205 & 1.3 @xmath206 & 1.9 @xmath207 & 0.084 @xmath208 & 0.65 @xmath209 & 0.153 @xmath210 + lba x all & 0.8 @xmath1 1.2 & 0.01 @xmath211 & 1.6 @xmath179 & 1.2 @xmath212 & 0.159 @xmath213 & 2.61 @xmath214 & 0.608 @xmath215 + [ tab : pf ] cccccccc all & 0.4 @xmath1 0.8 & 0.56 @xmath216 & 1.05 @xmath217 & 2.0 @xmath200 & 0.084 @xmath218 & 0.14 @xmath219 & 0.033 @xmath220 + all & 0.8 @xmath1 1.2 & 0.30 @xmath221 & 1.12 @xmath222 & 1.9 @xmath172 & 0.159 @xmath180 & 0.26 @xmath221 & 0.060 @xmath223 + blg & 0.4 @xmath1 0.8 & 0.44 @xmath224 & 1.14 @xmath225 & 1.9 @xmath226 & 0.011 @xmath227 & 0.03 @xmath228 & 0.007 @xmath229 + blg & 0.8 @xmath1 1.2 & 0.34 @xmath230 & 1.11 @xmath231 & 1.9 @xmath232 & 0.039 @xmath202 & 0.07 @xmath219 & 0.017 @xmath233 + lrg & 0.4 @xmath1 0.8 & 0.38 @xmath234 & 1.23 @xmath235 & 2.0 @xmath236 & 0.015 @xmath227 & 0.06 @xmath237 & 0.015 @xmath238 + lrg & 0.8 @xmath1 1.2 & 0.18 @xmath239 & 1.41 @xmath240 & 2.1 @xmath241 & 0.023 @xmath242 & 0.24 @xmath243 & 0.056 @xmath244 + [ tab : acor ] 25.5 ) cosmos catalog : all ( upper panels ) , blg ( middle left ) , and lrg ( middle right ) and the color - magnitude distributions of uvlgs ( lower left ) and lbas ( lower right ) , selected from the cosmos+_galex _ ( i@xmath025.5 and nuv@xmath023 ) combined catalog . the upper left panel illustrates the luminosity - redshift distribution for the full data , with the black box defining a volume - limited sample ( all ) . the plot on the upper right shows how this sample of galaxies , all , occupies the color - magnitude diagram . the lower plots show color magnitude diagrams for the other samples . we define blgs to have @xmath2451.8 and @xmath246 - 20.0 ( shown on middle left ) and lrgs to have @xmath2472.0 and @xmath248 - 20.5 ( middle right ) . the lowest left panel shows the uvlg distribution in purple with contour levels representing the number of galaxies in color - magnitide bins corresponding to ( 3 , 6 , 12 , 24 ) galaxies ; similarly , the lowest right panel shows the same for lbas with contour levels corresponding to ( 1.5 , 3 , 6 ) galaxies . , title="fig:",width=249 ] 25.5 ) cosmos catalog : all ( upper panels ) , blg ( middle left ) , and lrg ( middle right ) and the color - magnitude distributions of uvlgs ( lower left ) and lbas ( lower right ) , selected from the cosmos+_galex _ ( i@xmath025.5 and nuv@xmath023 ) combined catalog . the upper left panel illustrates the luminosity - redshift distribution for the full data , with the black box defining a volume - limited sample ( all ) . the plot on the upper right shows how this sample of galaxies , all , occupies the color - magnitude diagram . the lower plots show color magnitude diagrams for the other samples . we define blgs to have @xmath2451.8 and @xmath246 - 20.0 ( shown on middle left ) and lrgs to have @xmath2472.0 and @xmath248 - 20.5 ( middle right ) . the lowest left panel shows the uvlg distribution in purple with contour levels representing the number of galaxies in color - magnitide bins corresponding to ( 3 , 6 , 12 , 24 ) galaxies ; similarly , the lowest right panel shows the same for lbas with contour levels corresponding to ( 1.5 , 3 , 6 ) galaxies . , title="fig:",width=249 ] 25.5 ) cosmos catalog : all ( upper panels ) , blg ( middle left ) , and lrg ( middle right ) and the color - magnitude distributions of uvlgs ( lower left ) and lbas ( lower right ) , selected from the cosmos+_galex _ ( i@xmath025.5 and nuv@xmath023 ) combined catalog . the upper left panel illustrates the luminosity - redshift distribution for the full data , with the black box defining a volume - limited sample ( all ) . the plot on the upper right shows how this sample of galaxies , all , occupies the color - magnitude diagram . the lower plots show color magnitude diagrams for the other samples . we define blgs to have @xmath2451.8 and @xmath246 - 20.0 ( shown on middle left ) and lrgs to have @xmath2472.0 and @xmath248 - 20.5 ( middle right ) . the lowest left panel shows the uvlg distribution in purple with contour levels representing the number of galaxies in color - magnitide bins corresponding to ( 3 , 6 , 12 , 24 ) galaxies ; similarly , the lowest right panel shows the same for lbas with contour levels corresponding to ( 1.5 , 3 , 6 ) galaxies . , title="fig:",width=249 ] 25.5 ) cosmos catalog : all ( upper panels ) , blg ( middle left ) , and lrg ( middle right ) and the color - magnitude distributions of uvlgs ( lower left ) and lbas ( lower right ) , selected from the cosmos+_galex _ ( i@xmath025.5 and nuv@xmath023 ) combined catalog . the upper left panel illustrates the luminosity - redshift distribution for the full data , with the black box defining a volume - limited sample ( all ) . the plot on the upper right shows how this sample of galaxies , all , occupies the color - magnitude diagram . the lower plots show color magnitude diagrams for the other samples . we define blgs to have @xmath2451.8 and @xmath246 - 20.0 ( shown on middle left ) and lrgs to have @xmath2472.0 and @xmath248 - 20.5 ( middle right ) . the lowest left panel shows the uvlg distribution in purple with contour levels representing the number of galaxies in color - magnitide bins corresponding to ( 3 , 6 , 12 , 24 ) galaxies ; similarly , the lowest right panel shows the same for lbas with contour levels corresponding to ( 1.5 , 3 , 6 ) galaxies . , title="fig:",width=249 ] 25.5 ) cosmos catalog : all ( upper panels ) , blg ( middle left ) , and lrg ( middle right ) and the color - magnitude distributions of uvlgs ( lower left ) and lbas ( lower right ) , selected from the cosmos+_galex _ ( i@xmath025.5 and nuv@xmath023 ) combined catalog . the upper left panel illustrates the luminosity - redshift distribution for the full data , with the black box defining a volume - limited sample ( all ) . the plot on the upper right shows how this sample of galaxies , all , occupies the color - magnitude diagram . the lower plots show color magnitude diagrams for the other samples . we define blgs to have @xmath2451.8 and @xmath246 - 20.0 ( shown on middle left ) and lrgs to have @xmath2472.0 and @xmath248 - 20.5 ( middle right ) . the lowest left panel shows the uvlg distribution in purple with contour levels representing the number of galaxies in color - magnitide bins corresponding to ( 3 , 6 , 12 , 24 ) galaxies ; similarly , the lowest right panel shows the same for lbas with contour levels corresponding to ( 1.5 , 3 , 6 ) galaxies . , title="fig:",width=249 ] 25.5 ) cosmos catalog : all ( upper panels ) , blg ( middle left ) , and lrg ( middle right ) and the color - magnitude distributions of uvlgs ( lower left ) and lbas ( lower right ) , selected from the cosmos+_galex _ ( i@xmath025.5 and nuv@xmath023 ) combined catalog . the upper left panel illustrates the luminosity - redshift distribution for the full data , with the black box defining a volume - limited sample ( all ) . the plot on the upper right shows how this sample of galaxies , all , occupies the color - magnitude diagram . the lower plots show color magnitude diagrams for the other samples . we define blgs to have @xmath2451.8 and @xmath246 - 20.0 ( shown on middle left ) and lrgs to have @xmath2472.0 and @xmath248 - 20.5 ( middle right ) . the lowest left panel shows the uvlg distribution in purple with contour levels representing the number of galaxies in color - magnitide bins corresponding to ( 3 , 6 , 12 , 24 ) galaxies ; similarly , the lowest right panel shows the same for lbas with contour levels corresponding to ( 1.5 , 3 , 6 ) galaxies . , title="fig:",width=249 ] z@xmath00.8 ( left ) , and 0.8@xmath0z@xmath01.2(right ) for samples from top to bottom : all ( top ) , blg , lrg , uvlg , and lba ( bottom ) . the distribution is smoothed to a scale of 4.8 . the spatial distributions are displayed over the mask ( black denotes regions where no reliable observations exist) the allowed regions of the masks , appearing in these plots as dark violet regions underlying the colored points , are used for populating the random catalogs . colors in this figure progress from deep violet to red with increasing density . , title="fig:",height=624 ] + 0.2 ) to wide ( dz@xmath180.4 ) redshift bins , by displaying the ratios of w(@xmath57 ) , at top , and w@xmath249(r@xmath249 ) , at bottom . for simplicity , we have chosen to display only one representative redshift range ( 0.4@xmath10.8 ) for the autocorrelation of the all ( left ) , blg ( middle ) and lrg ( right ) samples . the ratios of the 0.4@xmath0z@xmath00.6 data to the wider 0.5@xmath0z@xmath00.9 are shown as black points , and the ratios of 0.6@xmath10.8 to 0.5@xmath10.9 appear as orange points . while the narrow bin data appears a factor of 1.5@xmath12 higher than that of the wide bins in the w(@xmath57 ) plots ( top ) , we show that converting this into w@xmath249(r@xmath249 ) units shows little discrepancy between the separate narrow redshift ranges and the wider redshift ( the dashed red line marks where the narrow and wide bins give equivalent results . in most cases , the correlation function derived using wide z bins corresponds to the average correlation function of the narrower bins . therefore , selecting a dz@xmath180.4 redshift bin will not systematically affect the correlation strength.,width=624 ] 1.2 redshift data . the shaded regions within the dashed lines provide errors caused by cosmic variance . the topmost panel compares auto - correlation of the all sample ( black ) with blg sample ( open navy circle ) and lrg sample ( open red circle ) . the second panel displays the all auto - correlation ( black ) with the ` all x uvlgs ' cross - correlation ( purple ) . the third panel , similarly , shows the blg auto - correlation ( blue ) in comparison with the ` blg x uvlgs ' cross - correlation ( purple ) . finally , in the bottom panel the auto - correlation of ` lrgs ' is shown in red , with the cross - correlation of ` lrgs ' with uvlgs shown in purple . while on small scales , the auto - correlations of the all and blg samples are indistinguishable from their cross correlations with the uvlgs , the lrgs appear to cluster with themselves more strongly than with uvlgs . , height=624 ] z@xmath01.2 uvlgs . we compare with correlation length data for star - forming galaxies from other studies . we also include high redshift results for lbgs , bx , and bm ( see @xcite for their description ) with low - redshift _ galex _ results to compare uvlgs with other uv - selected samples ( all shown in magenta ) . the dark green shaded region shows the evolution of a 10@xmath25010@xmath114 m@xmath36 halo with redshift ( identical to @xcite).,height=384 ] m@xmath118 ; @xcite ( magenta star ) studies a sample with m@xmath119 ; @xcite ( orange square ) finds merger fraction , assumed to be half the pair fraction , for m@xmath116 . these magnitude limited samples , with no color selection , are most comparable to our all sample . from eqn . [ eqn : pf ] we see that the pair fraction depends on number density , which depends strongly on limiting magnitude . in order to compare our pair fraction with other studies , we extrapolate the pair fraction of the all sample for m@xmath251 , changing only the number density , and assuming that the correlation lengths is the same as our all sample ( see discussion in @xcite ) . these points are shown as open circles , and appear to agree well with the other studies.,height=432 ]

observables in relativistic heavy ion experiments are often reported as functions of measured total inelastic cross section fraction ( centrality ) and are related to the initial collision geometry using monte carlo glauber ( mcg ) or optical glauber models @xcite . an example is charged particle multiplicity @xmath4 versus number of participating nucleons @xmath0 . it is therefore imperative that the level of accuracy of the glauber models be understood and that relevant experimental information be used to estimate collision centrality . for instance , previous studies @xcite concluded that the uncertainties in relating peripheral collision data to the corresponding initial collision geometry were very large , thus contributing to the omission of those data from publications @xcite . in another example , studies of high-@xmath5 suppression of charged particle production for peripheral to mid - central collisions are presently limited by large uncertainties in the estimated number of binary collisions which is used as a scaling variable @xcite . in this work we re - examine the accuracy of monte carlo glauber model descriptions of minimum - bias multiplicity frequency distribution data from rhic using updated density and multipliciity production model parameters and a sensitive power - law inspired representation introduced by trainor and prindle ( tp ) @xcite . after demonstrating the accuracy of our monte carlo glauber model in describing rhic data we then use the power - law analysis method of ref . @xcite to estimate the uncertainties in the mapping relationships between initial stage collision geometry quantities and centrality for heavy ion collision systems relevant to the rhic program . we find that previous uncertainties @xcite in centrality measures are too pessimistic . the centrality analysis method of trainor and prindle @xcite exploits the approximate power - law dependence of the multiplicity frequency distribution data @xmath6 ( @xmath7 is the number of triggered events ) for minimum - bias collisions and uses proton - proton ( p - p ) multiplicity production data to constrain the peripheral collision end - point of the @xmath6 distribution . their analysis demonstrates that uncertainties in centrality determination and mcg geometry measures can be significantly reduced , particularly in the peripheral region . in this paper mcg results and errors for centrality bin average quantities @xmath8 ( mean impact parameter ) , @xmath9 , @xmath10 , and @xmath11 @xcite are presented for minimum - bias au - au collisions at @xmath3 = 20 , 62 , 130 and 200 gev and for cu - cu at 62 and 200 gev . in this study centrality is based on charged particle production at midrapidity . other centrality definitions appear in the literature but will not be considered here . however , the mcg model and analysis methods presented here and in ref . @xcite can be readily extended to any practical centrality determination method . the mcg model and updated input parameters are discussed in sec . [ sec : model ] . the accuracy of the model for describing @xmath6 data from rhic is demonstrated in sec . [ sec : datafit ] . mcg predictions and errors are presented and discussed in sec . [ sec : results ] . comparison of the mcg results and analytic parametrizations from ref . @xcite are presented in sec . [ sec : analytic ] . in sec . [ sec : discussion ] we demonstrate via simulations how energy deposition data from transverse particle production , which are generally available at the trigger level from the rhic experiments @xcite , can be used to further reduce the impact of background contamination and to mitigate the effects of event loss due to trigger and collision vertex finding inefficiencies . a summary and conclusions are presented in sec . [ sec : summary ] . computational details are contained in three appendixes at the end . the monte carlo glauber collision model used here is based on a standard set of assumptions @xcite which are appropriate for ultra - relativistic heavy ion collisions . these assumptions include the characterization of the collision in terms of a classical impact parameter @xmath12 , straight - line propagation of each incident nucleon through the oncoming nucleus , and a fixed transverse interaction range determined by the nucleon - nucleon ( n - n ) total inelastic cross section ( @xmath13 ) in free space . the impact parameter was selected at random and the positions of the nucleons relative to the geometrical centers of the colliding nuclei were randomly distributed according to a spherical density @xmath14 . center - of - mass constraints were not imposed on the nucleon positions . nucleon pairs in the colliding nuclei were assumed to interact hadronically if their relative impact parameter was @xmath15 . charged hadron multiplicity was assigned using the phenomenological two - component model of kharzeev and nardi @xcite for `` soft '' plus `` hard '' particle production processes where the mean number of charged hadrons in the acceptance ( @xmath16 ) per unit pseudorapidity ( @xmath17 ) was computed according to & = & ( 1-x)n_pp + x n_pp n_bin . [ eq1 ] parameters @xmath18 and @xmath19 depend on collision energy . event - wise multiplicities were obtained by sampling the gaussian distribution @xcite ( n_ch,|n_ch ) & = & ( - ) [ eq2 ] where @xmath20 is a multiplicity fluctuation width parameter . alternate distributions ( _ e.g. _ poisson , negative binomial @xcite ) produce quantitative effects on the multiplicity frequency distribution for very peripheral collisions but do not affect the present estimates of systematic errors . @xmath1 and @xmath4 were both required to be @xmath21 in order for the simulated collision to be used in the analysis . centrality bins were defined using the multiplicity frequency distribution @xmath6 . centrality bin average quantities , @xmath8 , @xmath9 , @xmath10 and @xmath22 , were calculated using the events within each bin . the acceptance for this study was @xmath23 and full @xmath24 in azimuth . monte carlo glauber simulations require the distribution of the centers of the nucleons in the nuclear ground state , @xmath25 , the point - matter density . for @xmath26cu and @xmath27au these were estimated using the measured charge densities @xcite and the hartree - fock calculations of negele using the density matrix expansion ( dme ) framework @xcite for the neutron - proton density differences . the charge densities for @xmath26cu and @xmath27au were represented by a woods - saxon distribution , ( r ) & = & _ 0 \ { 1 + } ^-1 , [ eq3 ] where the radius and diffuseness parameters are listed in table [ tablei ] . @cllll density & & + parameter & empirical@xmath28 & dme@xmath29 & empirical@xmath28 & dme@xmath29 + @xmath30 & 4.214@xmath310.026 & 4.232 & 6.38@xmath310.06 & 6.443 + @xmath32 & 0.586@xmath310.018 & & 0.535@xmath310.027 & + @xmath33 & 3.925@xmath310.022 & 3.899 & 5.33@xmath310.05 & 5.423 + @xmath34 & 4.195@xmath310.085 & 4.213 & 6.43@xmath310.10 & 6.495 + @xmath35 & 0.581@xmath310.031 & & 0.568@xmath310.047 & + @xmath36 & 3.901 & 3.875 & 5.41 & 5.502 + @xmath28 charge density results based on electron scattering analysis @xcite ; estimates of point matter densities as discussed in the text . @xmath29 density matrix expansion predictions @xcite . the point matter densities assumed for the present analysis are spherically symmetric with a woods - saxon radial distribution where the half - density and rms radii were estimated by adding the hartree - fock dme point matter charge distribution differences to the measured charge density radii where , c_pt , m & = & c_chrg + _ dme , + r^2_pt , m ^1/2 & = & r^2_chrg ^1/2 + _ dme , [ eq4 ] where @xmath30 and @xmath33 denote the measured radii from electron scattering while the quantities in the square brackets are the dme predictions . the diffusivity parameter @xmath35 was obtained from the quantities in eq . ( [ eq4 ] ) and the approximate relation , @xmath37 $ ] @xcite . the nominal radius and diffusivity parameters assumed here for au and cu are listed in table [ tablei ] . the uncertainties in @xmath34 and @xmath35 were obtained by summing the independent errors in the proton @xcite and neutron point matter densities in quadrature . the latter is constrained by theoretical and experimental information about the neutron - proton density difference . theoretical nuclear structure model predictions for the neutron@xmath38proton rms radii differences for isotopes in the cu and au mass range agree to within about @xmath39 fm @xcite . in general the theoretical predictions agree with analyses of medium energy proton - nucleus elastic scattering data @xcite which are typically uncertain by about @xmath40 fm for isotopes in the cu and au mass range . the uncertainty in the neutron density rms radii relative to the proton density was therefore assumed to be @xmath41 fm and the corresponding uncertainty in the matter rms radii was @xmath42 fm , where @xmath43 are the number of neutrons and nucleons in the isotope . theoretical contributions to the errors in @xmath34 and @xmath35 were conservatively estimated by requiring each to independently account for the @xmath31(n / a)0.09 fm error . the latter theoretical errors were combined in quadrature with the corresponding errors for @xmath30 and @xmath32 from analysis of electron scattering data to obtain the final errors listed in table [ tablei ] . the n - n total inelastic cross sections used here were based on total cross section measurements for p - p collisions ( @xmath311 mb uncertainty ) and elastic total cross sections for p - p and p-@xmath44 ( @xmath310.5 mb error ) @xcite . proton - neutron total cross section data are not available in the energy range studied here . the results for energies @xmath45 = 20 , 62 , 130 and 200 gev are respectively 33 , 35.3 , 38.7 and 41.7 mb , each being uncertain by @xmath31 1.1 mb . it is possible that the effective interaction cross section between colliding nucleons inside a nucleus differs from that in free space ( density dependence ) or that the strength and range of the effective n - n interaction changes with each successive collision as in the `` used '' nucleon scenario @xcite or the `` strict '' participant scaling model @xcite ( limiting case of the used - nucleon model in which nucleons interact only once ) . the study of density dependent effects is well beyond the scope of the present analysis . parameters @xmath46 at @xmath45 = 62 , 130 and 200 gev are 2.01 , 2.25 and 2.43 ( @xmath47 error for each ) , respectively , using the energy dependent parametrization of the ua5 @xcite and cdf data given in ref . @xcite . for the @xmath45 = 20 gev data , which is outside the energy range parametrized in @xcite , the average of isr @xcite and fnal @xcite measurements summarized in ref . @xcite was assumed where @xmath48 . binary scattering parameter @xmath19 in eq . ( [ eq1 ] ) was estimated by fitting @xmath49 data from star @xcite , phenix @xcite , and phobos @xcite versus @xmath2 with eq . ( [ eq1 ] ) rewritten as , & = & n_pp[1+x(-1 ) ] , [ eq5 ] assuming the above values for @xmath50 . the resulting values of @xmath19 are 0.07 , 0.09 , 0.09 and 0.13 ( @xmath310.03 errors for each ) for the 20 , 62 , 130 and 200 gev data , respectively . analysis of the 19.6 gev au - au data by the phobos experiment @xcite assumed @xmath51 @xcite and obtained @xmath52 , leading to claims in the literature that hard scattering contributions to multiplicity do not change with collision energies from 20 to 200 gev @xcite . a more recent compilation @xcite of @xmath50 measurements in this energy range indicates a larger value of @xmath48 . fitting eq . ( [ eq5 ] ) to the combined phenix @xcite and phobos @xcite 19.6 gev data with @xmath53 resulted in a smaller @xmath54 . phobos 62 gev au - au data @xcite , when plotted versus @xmath2 , do not linearly extrapolate to @xmath50 in contrast to what is expected from fig . 4 of ref . @xcite and fig . 1 of ref . @xcite from the phobos experiment . however , 62 gev au - au results from the star experiment linearly extrapolate to @xmath50 at @xmath55 , resulting in the fitted value @xmath56 used here . @xmath19 parameters from the three rhic experiments generally agree for the 130 and 200 gev data ; the phobos values of @xmath56 @xcite and 0.13 @xcite at 130 and 200 gev , respectively , were confirmed and used here . it is possible that parameter @xmath19 is affected by other processes in addition to hard and semi - hard partonic scattering in the initial collision stage where the latter mechanisms would be expected to follow a @xmath57 dependence . the variance of the multiplicity distribution @xmath58 in eq . ( [ eq2 ] ) is given by @xmath59 where values of parameter @xmath60 or @xmath61 represent multiplicity fluctuation suppression or excess , respectively , relative to pure statistical fluctuations ( @xmath62 ) . in general a non - vanishing integral @xcite of two - particle correlations @xcite over the acceptance requires @xmath63 . in principle , correlation measures of the type reported in ref . @xcite could be used to determine parameter @xmath20 using the relationship between fluctuations and correlations developed in refs . @xcite and discussed in appendix a. on the other hand , the kharzeev and nardi two - component multiplicity model with distribution @xmath58 constitutes a phenomenology for describing event - wise multiplicity frequency distribution data . the phenomenological approach based on fits to multiplicity distribution data , shown in the next section , was used to estimate parameter @xmath20 in the present analysis . the uncertainty in the width of @xmath58 was estimated by fitting the data . the uncertainty in the analytic form of the multiplicity fluctuation distribution was accounted for by comparing monte carlo glauber results assuming the gaussian @xmath58 in eq . ( [ eq2 ] ) ( with @xmath62 ) with results assuming a negative binomial distribution ( nbd ) @xcite in place of eq . ( [ eq2 ] ) as explained in appendix a. the 130 gev au - au minimum - bias negatively charged hadron multiplicity frequency distribution data from star @xcite for 60k events were fit by adjusting parameters @xmath50 ( for negative hadrons ) and @xmath20 where @xmath19 was fixed to 0.09 as discussed above . the acceptance was defined by transverse momentum @xmath64 gev/@xmath65 , @xmath66 , and @xmath67 . the fit ( solid histogram ) is shown in the left panel of fig . [ fig1 ] in comparison with data ( solid dots ) , where the optimum values of @xmath50 and @xmath20 are @xmath68 ( consistent with @xmath69 for charged particle yields ) and @xmath70 , respectively . similar fits ( not shown ) to the charged particle minimum - bias multiplicity distribution from the same data set resulted in @xmath71 . parameter @xmath20 was therefore set to @xmath72 for all four energies . in the tp analysis @xcite it was shown that minimum - bias multiplicity frequency distributions for relativistic heavy ion collision experiments and monte carlo glauber models approximately follow a power - law distribution . this is illustrated by plotting the data and mcg fit from the left panel of fig . [ fig1 ] on log - log axes in the middle panel . except near the end - points the data can be described to within 10% with a slope ( exponent ) of approximately @xmath73 . the collision event yield is approximately proportional to @xmath74 . therefore distribution @xmath75 versus @xmath76 is approximately constant , where & = & = 4n_h^-^3/4 as shown in the right - hand panel of fig . [ fig1 ] for data ( solid dots ) and mcg fit ( solid line ) . the power - law plotting format in the right - hand panel enables a more sensitive comparison between the model fit and data than in the usual semi - log format ( left panel ) . the mcg model fit is consistent with the data except for the first two data points at low @xmath76 . except near the end - points the @xmath75 data are constant on @xmath76 to within an overall variation of 20% in comparison to the @xmath77 data which span nearly two orders of magnitude within the same multiplicity range . the results in fig . [ fig1 ] demonstrate the efficacy of the monte carlo glauber model with two - component multiplicity production for accurately describing the measured multiplicity frequency distributions at rhic . based on this outcome we conclude that the present model is reasonable to use in estimating centrality bin average quantities and their systematic errors for rhic data . ensembles of one - million , minimum - bias ( _ i.e. _ random impact parameter ) monte carlo collisions were generated for each of the six systems studied here . centrality bin averaged quantities @xmath78 , @xmath8 , @xmath9 , @xmath10 and @xmath2 are listed in tables [ tableii]-[tablevii ] using the nominal parameter values discussed above . statistical errors are typically 0.1 - 0.2% and always @xmath79% of the nominal bin averages and in all instances are much less than the systematic errors discussed below . results in table [ tableii ] for @xmath9 for au - au collisions at 200 gev are systematically larger than published monte carlo glauber predictions in ref . @xcite by approximately 5% for peripheral centrality bins . this systematic increase is primarily caused by multiplicity fluctuations in the present model and basing centrality on @xmath4 here rather than on @xmath0 as was done in ref . @xcite . average charged particle multiplicity per participant pair for @xmath66 at mid - pseudorapidity as a function of centrality ( @xmath2 ) is shown in fig . [ fig2 ] for monte carlo simulated au - au collisions at @xmath3 = 200 gev ( solid dots ) . the p - p limit , @xmath80 , is indicated by the solid square symbol . the data display a linear dependence on @xmath81 as expected from eq . ( [ eq5 ] ) except for the most - peripheral 90 - 100% centrality bin where multiplicity fluctuations significantly reduce the average @xmath4 . the slope agrees with @xmath82 ; the linear extrapolation to the p - p limit is evident . ( color online ) monte carlo glauber results for @xmath83 for one - million au - au collisions at @xmath3 = 200 gev using the nominal parameters discussed in the text ( solid dots ) . linear fit ( dashed line ) to the au - au results ( excluding the 90 - 100% centrality bin ) accurately extrapolates to the p - p multiplicity from ua5 @xcite . monte carlo data for au - au include statistical errors only which are smaller than the symbols.,width=288 ] the sources of uncertainty which cause systematic error in the collision geometry quantities can be organized into three categories . the first includes fitting uncertainties that occur when the corrected , multiplicity frequency distribution data are described by adjusting the parameters of the multiplicity production model in eqs . ( [ eq1]-[eq2 ] ) . the second includes the uncertainties in the matter density , nucleon - nucleon total inelastic cross section , and functional representation of the multiplicity fluctuation distribution @xmath58 . the third includes the uncertainty in the corrected multiplicity frequency distribution data @xmath6 . the latter arise from trigger inefficiency , collision vertex finding inefficiency , background contamination , and particle tajectory finding inefficiency . systematic errors in centrality bin average quantities due to fitting uncertainty were estimated by comparing the nominal results to that obtained with parameters @xmath19 and @xmath20 in eqs . ( [ eq1]-[eq2 ] ) individually varied within their respective fitting errors ( secs . [ subsec : twocomponent ] and [ sec : datafit ] ) . uncertainty in parameter @xmath50 directly affects @xmath78 but does not affect @xmath8 , @xmath9 , @xmath10 or @xmath2 . error estimates due to uncertainties in the matter density radius and diffuseness parameters and the n - n inelastic cross section require a refitting of the @xmath6 distribution . this was accomplished via a phenomenological adjustment of the multiplicity production model described in appendix b. fit recovery via @xmath84 minimization for mcg simulations with ample statistics ( of order @xmath85 collisions ) is computationally intensive . the method in appendix b is fast and accurate . errors due to the uncertainty in the mathematical representation of @xmath58 were estimated by comparing the nominal mcg results using the gaussian distribution in eq . ( [ eq2 ] ) with results assuming a negative binomial distribution @xcite as explained in appendix a. uncertainties in the @xmath6 data arising from trigger inefficiency , collision vertex finding inefficiency , and background contamination mainly affect the low multiplicity region . minimum - bias trigger efficiencies at rhic are @xmath86% @xcite , @xmath87% @xcite , 96% @xcite and @xmath88% @xcite . trigger inefficiency causes the lower @xmath4 half - max point of the @xmath89 distribution to shift to larger @xmath90 . the position of this point for the @xmath89 distribution when corrected for trigger inefficiency ( assuming the power - law behavior ) has a relative uncertainty of about @xmath3110% . 2 to @xmath313% . the fractional uncertainty in the lower half - max position is @xmath31(0.02 to 0.03)@xmath91% where @xmath92 and @xmath93 are the lower and upper half - max points of the @xmath89 distribution . ] however , knowledge that the lower half - max point is constrained by p - p scattering allows the relative uncertainties of the lower half - max point on @xmath90 to be reduced to 1/4 of the uncertainties in @xmath50 , or to @xmath310.8% , @xmath310.9% , @xmath311% , and @xmath312% for the 200 , 130 , 62 and 20 gev data , respectively . primary collision vertex reconstruction efficiency is approximately 100% for events with @xmath4 of order a few tens and greater but falls precipitously for @xmath94 @xcite . phobos @xcite and star @xcite both report uncertainties in their overall vertex finding efficiencies of about @xmath95% for minimum - bias collisions . vertex finding inefficiency increases the slope and half - max position of the low multiplicity edge of the @xmath89 distribution . an uncertainty of @xmath95% in the overall efficiency results in about @xmath3110% uncertainty in the low @xmath90 half - max position . however , the p - p data constrain this uncertainty . in this analysis we assumed 100% vertex finding efficiency for collisions producing @xmath96 @xcite , or @xmath97 , and allowed the slope of the lower edge of the @xmath89 distribution to vary such that the half - max position varied by @xmath310.8% , @xmath310.9% , @xmath311% , and @xmath312% for the 200 , 130 , 62 and 20 gev data , respectively , as in the preceding paragraph . the principle sources of background contamination are from ultra - peripheral two - photon interactions @xcite and beam - gas collisions . the former process corresponds to coherent photon - photon interactions which excite both nuclei , followed by neutron decay ( which activates the minimum - bias trigger detectors ) and accompanied by resonance(s ) production which decays into charged particles transverse to the beam direction . transverse particle multiplicities from upc events are typically @xmath98 ( _ e.g. _ @xmath99-meson decay ) and generally @xmath100 @xcite for @xmath101 at midrapidity . upc backgrounds are therefore restricted to the most - peripheral ( 90 - 100% ) centrality bin . upc yields should be approximately proportional to @xmath102 @xcite ( charge numbers for colliding ions 1 and 2 ) whereas beam - gas contamination should scale with beam current . other background events , _ e.g. _ mutual coulomb dissociation processes such as @xmath103 + a @xmath104 b + n for both nuclei , can be eliminated by requiring minimum transverse particle production . remaining background events will appear near the lower @xmath4 edge of the @xmath89 distribution . estimates of background contamination in minimum - bias trigger data at rhic range from @xmath105% @xcite to 6% @xcite overall for au - au at @xmath3 = 130 gev corresponding to 0 - 20% and 60% , respectively , of the hadronic collision event yield in the 90 - 100% centrality bin . most of the upc events occur at @xmath106 and can be eliminated by cuts on the number of transverse charged particles . the remaining background contributions for @xmath107 are less than the amounts listed above . for the present analysis background contamination was assumed to be dominated by upc events and to diminish in magnitude with collision energy and @xmath102 . in the present analysis background contamination was applied to the nominal 90 - 100% centrality bin assuming 3% , 2% and 1% overall contamination levels in the au - au data at @xmath3 = 200 , 130 and 62 gev , respectively . upc contamination for au - au collisions at 20 gev and for cu - cu at 200 and 62 gev was estimated to be negligible . however , calculations were done for the latter three systems assuming a 1% overall background contamination ( 10% contamination within the nominal 90 - 100% centrality bin ) in order to provide a reference for further systematic error estimation . the above sources of systematic uncertainty primarily affect the position , slope and shape of the @xmath89 distribution at the lower @xmath90 end - point . these changes impact the mcg model which must describe those data and , in turn , the collision geometry quantities like @xmath9 . the phenomenological method in appendix b was used to estimate the changes in the collision geometry quantities relative to the nominal values . the latter differences were taken as the estimated errors . background levels in collider experiments vary significantly depending on beam quality , beam current and interaction rate . excessive trigger backgrounds beyond that considered here may result from beam - gas interactions during periods of high integrated beam currents . collision event pile - up in the particle tracking detectors during periods of high luminosity adversely affect collision vertex finding . either condition may be so severe as to preclude access to the low multiplicity range of the minimum - bias distribution . even so , the power - law dependence and p - p end - point constraints enable accurate centrality estimates and collision geometry assignments to be made for the remaining minimum - bias data . charged particle trajectory reconstruction efficiencies in the large acceptance tracking detectors at rhic are typically 70 - 95% @xcite and decrease approximately linearly with particle density in the detectors by about 20% from most - peripheral to most - central collisions @xcite . uncertainties in the assumed track reconstruction efficiencies were estimated by comparing corrected data for @xmath108 versus @xmath0 between the rhic experiments for au - au collisions at 20 gev @xcite 130 gev @xcite and 200 gev @xcite . the comparisons indicate uncertainties in both the overall tracking efficiencies and in the dependence on particle density in the tracking detectors . the latter variation in efficiency from peripheral to central collisions is about @xmath109% . overall changes in tracking efficiency are compensated in the mcg model by multiplicative adjustments to parameter @xmath50 and therefore have no effect on the centrality measures reported here . changes in the assumed tracking efficiency dependence on @xmath4 affect the lower and upper end - point positions of the @xmath89 distribution and distort its shape . the distortions must be accounted for by the mcg model in order to determine the net effect on the collision geometry measures . the effects of the 8% uncertainty in the multiplicity dependence of the trajectory reconstruction efficiency were estimated by generalizing parameter @xmath50 in eq . ( [ eq1 ] ) to @xmath110 where @xmath111 and @xmath112 for most - central au - au collisions . the same value for @xmath113 was assumed for cu - cu . systematic errors were estimated as the differences between the nominal centrality bin averages and those obtained assuming @xmath111 . the choice to maximize multiplicity variation for most - central collisions is arbitrary . as a result systematic errors in centrality bin average multiplicities were not included in tables [ tableii]-[tablevii ] . however , this ambiguity does not affect the resulting systematic errors in the other centrality measures reported here . the combined systematic errors ( all components added in quadrature ) in both magnitude and relative percent ( given in parentheses ) are listed in tables [ tableii ] - [ tablevii ] for the six collision systems studied here . impact parameter uncertainty is about @xmath114% for au - au and @xmath115% for cu - cu . uncertainty in @xmath9 is about @xmath116% for au - au and @xmath117% for cu - cu for peripheral centralities , reducing to about @xmath118% or less for central collisions where full geometrical overlap of the two nuclei suppresses the dependence of @xmath0 on variations in the nuclear surface geometry . systematic errors in @xmath10 vary from about @xmath119% for au - au collisions and @xmath120% for cu - cu for central to peripheral collisions . errors in @xmath2 ( @xmath121% or less ) are suppressed due to covariation of @xmath9 and @xmath10 . individual contributions to the systematic errors in collision geometry bin averages @xmath8 , @xmath9 , @xmath10 and @xmath2 for au - au and cu - cu collisions at @xmath3 = 200 gev are summarized in tables [ tableviii ] and [ tableix ] , respectively . similar results were obtained for the other four collision systems . errors are given in percent where the absolute values of positive and negative errors were averaged together . the three numbers listed for each instance correspond to the average errors within the 60 - 100% , 20 - 60% and 0 - 20% centrality bins , respectively . the dominant errors are due to uncertainties in the matter density and the n - n total inelastic cross section . errors due to uncertainties in the analytic form of the phenomenological multiplicity fluctuation model are only significant for peripheral collisions . errors due to uncertainties in the multiplicity dependent particle track reconstruction efficiency and in the trigger and vertex reconstruction inefficiencies are negligible when constrained by p - p data as shown previously by trainor and prindle @xcite . percent uncertainties due to possible background contamination are listed for all six collision systems in table [ tablex ] . the largest errors occur in the most - peripheral centrality bin as expected . the absolute magnitudes of the errors decline rapidly with increasing centrality and are negligible for the centrality bins not listed in the table . reference errors for au - au at 20 gev and cu - cu at @xmath3 = 200 and 62 gev were based on an assumed 10% background contamination in the 90 - 100% centrality bin . estimates of total systematic error when background contamination differs from that assumed here can be obtained by removing the error contributions in table [ tablex ] from the total errors in tables [ tableii ] - [ tableiv ] for au - au collisions at 200 , 130 and 62 gev , respectively , and then adding ( in quadrature ) the appropriately scaled background errors from table [ tablex ] . for au - au collisions at 20 gev and cu - cu collisions at 200 and 62 gev the scaled errors from table [ tablex ] should be combined in quadrature with the total systematic errors in tables [ tablev ] - [ tablevii ] . if much larger backgrounds than those assumed here are encountered , then the present monte carlo results should not be scaled ; rather the errors should be recalculated . overall , the systematic errors in collision geometry bin averages for non - peripheral centralities are dominated by uncertainties in the nuclear geometry and @xmath13 . errors in the more peripheral bins are dominated by background contamination and ambiguities in the analytic form of the multiplicity fluctuation model . the power - law description of heavy - ion collision centrality developed by trainor and prindle @xcite prescribes simple , analytic parametrizations of the @xmath0 , @xmath1 , @xmath2 and @xmath4 dependences on total inelastic cross section fraction @xmath122 . for @xmath0 the running integral relation on @xmath123 is accurately given by @xcite ( 1 - /_0 ) & = & , [ eq7 ] where ( n_part/2)^ & = & ( n_part , min/2)^ /_0 + ( n_part , max/2)^ ( 1 - /_0 ) . [ eq8 ] similarly n_bin^ & = & n_bin , min^ /_0 + n_bin , max^ ( 1 - /_0 ) , [ eq9 ] where subscripts @xmath124 and @xmath125 refer respectively to the lower and upper half - max end - point positions of the @xmath126 and @xmath127 distributions . the above parametrizations are shown as the solid lines in the two upper panels of fig . [ fig3 ] in comparison with the present mcg model results ( solid dots ) from table [ tableii ] for au - au collisions at @xmath3 = 200 gev . the end - point parameters for the power - law parametrizations were @xmath128 , @xmath129 , @xmath130 and @xmath131 . the parametrization for @xmath132 is compared with the mcg result in the lower - left panel . the average multiplicities using eq . ( [ eq5 ] ) , the values for @xmath50 and @xmath19 from sec . [ subsec : twocomponent ] , and the above parametrizations for @xmath0 and @xmath2 ( solid curve ) is compared with the mcg result in the lower - right panel . the simple power - law parametrizations introduced in ref . @xcite quantitatively describe the mcg results , thus confirming their utility and precision . calculation of the analytic power - law parametrizations require the upper half - max end - point values for @xmath0 and @xmath1 from the mcg model . these quantities are listed in table [ tablexi ] for the six collision systems studied here . very peripheral collision data from the four rhic experiments have generally remained unpublished due to concerns with possibly large , and not well understood background contamination , trigger inefficiencies , and primary collision vertex finding inefficiencies . trainor and prindle @xcite originally showed , and the present analysis confirms , that the power - law dependence of the multiplicity frequency distribution data together with knowledge of the proton - proton multiplicity enables accurate centrality information to be obtained in spite of these uncertainties . nevertheless , upc events in principle cause the lower end - point portion of the multiplicity frequency distribution for minimum - bias a - a collisions to differ from the p - p limit and these backgrounds will contaminate the a - a data in the most - peripheral centrality bins . in this section we discuss additional analysis methods using data available to the rhic experiments to minimize backgrounds and to provide contamination level estimates for the accepted centrality bins . transverse particle production information via scintillators , silicon detectors , and/or calorimetry are available in the rhic experiments at the trigger level @xcite . these data record total energy deposition in the sensitive detector material for each ( minimum bias ) triggered collision event including contributions from a - a hadronic collisions , ultra - peripheral collisions , beam - gas and other backgrounds , etc . but do not include particle track finding and primary collision vertex finding inefficiencies . if the integrated yields from these detectors , denoted as @xmath133 , is approximately proportional to @xmath4 , then the frequency distribution will follow an approximate power - law such that the data can be usefully represented by @xmath134 versus @xmath135 , where @xmath136 is the number of triggers . mcg simulations for the integrated yields from the star central trigger barrel ( ctb ) @xcite plastic scintillator detector ( @xmath137 , @xmath24 azimuth coverage ) were done for p - p and au - au minimum - bias collisions at @xmath3 = 200 gev . details of this simulation are discussed in appendix c. the results for @xmath134 versus @xmath135 are shown in fig . both the p - p and au - au simulations reproduce the general shapes of measured star ctb minimum - bias trigger yields @xcite . the simulated au - au ctb yields follow the @xmath138 power - law distribution very well where the lower half - max point coincides with the mode of the p - p distribution . trigger inefficiency in au - au causes a depletion at low multiplicity as indicated by the green , dashed curve . over- or under - corrected yields result in the blue dotted curves . from fig . [ fig4 ] it is clear how the p - p trigger yield can be used to constrain corrections for trigger inefficiency in a - a . ( color online ) monte carlo simulation results for the minimum - bias trigger frequency distribution for transverse particle production integrated trigger detector yield quantity @xmath133 to the 1/4 power . simulated yields are shown for 1 m 200 gev au - au ( solid black curve , multiplied by 3 ) and 1 m proton - proton ( solid red curve ) collisions . the yield for au - au collisions when trigger inefficiency is large is illustrated by the lowest ( right - most ) green dashed curve ( hand - drawn sketch ) . over- and under - corrected yields are similarly illustrated by the upper ( left - most ) and middle dotted blue lines , respectively . background contamination contributes at lower multiplicity as illustrated by the dashed - dotted magenta curve ( hand - drawn sketch ) . , width=384 ] background triggers appear at lower values of @xmath133 where upc events typically produce of order two charged particles at mid - rapidity compared with the average from p - p collisions of about @xmath139 in @xmath137 and therefore appear as a peak or enhancement ( dashed - dotted curve in fig . [ fig4 ] ) below the mode of the p - p distribution . comparisons between the measured p - p and au - au @xmath134 distributions and the power - law simulations for au - au provide a reasonable means for defining additional event cuts to reduce background contamination . this information can also be used to estimate the contamination level for accepted events . measurements of the dependence of the lower end - point region of the @xmath134 distribution on integrated beam current , luminosity , and ion species would help disentangle background contributions from beam - gas collisions , event pile - up , and upc events . a similar power - law analysis using reconstructed particle tracks , but without a primary collision vertex requirement , would permit the vertex finding inefficiencies to be estimated and corrected . a monte carlo glauber and two - component multiplicity production model with fluctuations for high energy heavy ion collisions was used to describe minimum - bias multiplicity frequency distribution data from rhic . updated woods - saxon radii and diffusivities for the point matter densities of @xmath27au and @xmath26cu were determined using a combination of charge density measurements from electron scattering and theoretical hartree - fock predictions . the binary scattering parameter @xmath19 was estimated using a compilation of available rhic data from 20 to 200 gev . the values for @xmath19 obtained here display some energy dependence , although with large uncertainty . the model was fitted to the 130 gev au - au minimum - bias data @xcite in both the conventional semi - log format and in the more sensitive power - law inspired format introduced in ref . quantitative agreement between the present model and data was obtained . systematic errors in centrality bin averages for the geometrical quantities @xmath8 , @xmath9 , @xmath10 and @xmath11 were estimated based on charged particle multiplicity for @xmath23 and full @xmath24 azimuth acceptance . the sources of systematic error considered here included uncertainties associated with the model ( nuclear density , nucleon - nucleon inelastic total cross section , multiplicity production model analytic form and parameters ) and with the corrected minimum - bias multiplicity frequency distribution data ( background contamination , trigger and collision vertex finding inefficiencies , and particle trajectory reconstruction inefficiencies ) . the tp centrality analysis method was applied to the mcg predictions and error analysis for six collision systems relevant to the rhic heavy ion program . the analysis showed that significantly reduced errors in the collision geometry quantities result when the uncertainties in the minimum - bias multiplicity frequency distribution data are constrained by the empirical power - law behavior and the minimum - bias p - p collision data . the reduction in errors is in agreement with the original tp power - law analysis ; the resulting errors are in general smaller than previous estimates @xcite which did not utilize the power - law and p - p constraints . the accuracy of simple parametrizations of centrality dependent quantities developed by trainor and prindle @xcite was confirmed . we also discussed how particle production data at the trigger level ( _ e.g. _ scintillator hits and calorimeter energy depositions ) can be used within the power - law and p - p constraint methodology @xcite to define additional event cuts which minimize background contamination and which enable the detrimental effects of trigger inefficiencies to be reduced . the systematic errors presented in this paper demonstrate the accuracy of collision geometry quantities which can be achieved with rhic data by exploiting the power - law behavior of the a - a data and using constraints from minimum - bias p - p collision data . the results presented here assume that backgrounds from upc events , pileup , beam - gas collisions , etc . are not too large and that the minimum - bias trigger and collision vertex reconstruction efficiencies are not too small for peripheral collisions . it is intended that the mcg results presented here will serve as a useful resource for the rhic community and that the analysis method developed by trainor and prindle @xcite and applied in this paper , together with the trigger data analysis method discussed here , will enable accurate description of , and better access to the heavy ion peripheral collision data from rhic . the authors gratefully acknowledge helpful discussions , detailed comments and the plots in fig . [ fig3 ] from prof . t. a. trainor ( univ . of washington ) as well as helpful discussions and comments from drs . j. c. dunlop and zhangbu xu ( both from brookhaven national lab ) . this research was supported in part by the u. s. dept . of energy grant de - fg03 - 94er40845 . cccccc + centrality ( % ) & @xmath78 & @xmath8 & @xmath9 & @xmath10 & @xmath2 + 90 - 100 & 2.9 & @xmath140 & @xmath141 & @xmath142 & @xmath143 + 80 - 90 & 8.2 & @xmath144 & @xmath145 & @xmath146 & @xmath147 + 70 - 80 & 18.7 & @xmath148 & @xmath149 & @xmath150 & @xmath151 + 60 - 70 & 37.8 & @xmath152 & @xmath153 & @xmath154 & @xmath155 + 50 - 60 & 69.7 & @xmath156 & @xmath157 & @xmath158 & @xmath159 + 40 - 50 & 118.5 & @xmath160 & @xmath161 & @xmath162 & @xmath163 + 30 - 40 & 190.2 & @xmath164 & @xmath165 & @xmath166 & @xmath167 + 20 - 30 & 291.5 & @xmath168 & @xmath169 & @xmath170 & @xmath171 + 10 - 20 & 433.5 & @xmath172 & @xmath173 & @xmath174 & @xmath175 + 5 - 10 & 577.6 & @xmath176 & @xmath177 & @xmath178 & @xmath179 + 0 - 5 & 705.6 & @xmath180 & @xmath181 & @xmath182 & @xmath183 + cccccc + centrality ( % ) & @xmath78 & @xmath8 & @xmath9 & @xmath10 & @xmath2 + 90 - 100 & 2.7 & @xmath184 & @xmath185 & @xmath186 & @xmath187 + 80 - 90 & 7.4 & @xmath188 & @xmath189 & @xmath190 & @xmath191 + 70 - 80 & 16.6 & @xmath192 & @xmath193 & @xmath194 & @xmath195 + 60 - 70 & 33.2 & @xmath196 & @xmath197 & @xmath198 & @xmath199 + 50 - 60 & 60.1 & @xmath200 & @xmath201 & @xmath202 & @xmath203 + 40 - 50 & 100.4 & @xmath204 & @xmath205 & @xmath206 & @xmath207 + 30 - 40 & 157.8 & @xmath208 & @xmath209 & @xmath210 & @xmath211 + 20 - 30 & 237.7 & @xmath212 & @xmath213 & @xmath214 & @xmath215 + 10 - 20 & 348.7 & @xmath216 & @xmath217 & @xmath218 & @xmath219 + 5 - 10 & 459.5 & @xmath220 & @xmath221 & @xmath222 & @xmath223 + 0 - 5 & 557.6 & @xmath224 & @xmath225 & @xmath226 & @xmath227 + cccccc + centrality ( % ) & @xmath78 & @xmath8 & @xmath9 & @xmath10 & @xmath2 + 90 - 100 & 2.4 & @xmath228 & @xmath229 & @xmath230 & @xmath231 + 80 - 90 & 6.6 & @xmath232 & @xmath233 & @xmath234 & @xmath235 + 70 - 80 & 14.6 & @xmath236 & @xmath237 & @xmath238 & @xmath239 + 60 - 70 & 29.0 & @xmath240 & @xmath241 & @xmath242 & @xmath243 + 50 - 60 & 52.3 & @xmath244 & @xmath245 & @xmath246 & @xmath247 + 40 - 50 & 87.1 & @xmath248 & @xmath249 & @xmath250 & @xmath251 + 30 - 40 & 136.9 & @xmath252 & @xmath253 & @xmath254 & @xmath255 + 20 - 30 & 205.5 & @xmath256 & @xmath257 & @xmath258 & @xmath259 + 10 - 20 & 300.7 & @xmath260 & @xmath261 & @xmath262 & @xmath263 + 5 - 10 & 395.5 & @xmath264 & @xmath265 & @xmath266 & @xmath267 + 0 - 5 & 480.8 & @xmath268 & @xmath269 & @xmath270 & @xmath271 + cccccc + centrality ( % ) & @xmath78 & @xmath8 & @xmath9 & @xmath10 & @xmath2 + 90 - 100 & 1.9 & @xmath272 & @xmath273 & @xmath274 & @xmath275 + 80 - 90 & 4.8 & @xmath276 & @xmath277 & @xmath278 & @xmath279 + 70 - 80 & 10.5 & @xmath280 & @xmath281 & @xmath282 & @xmath283 + 60 - 70 & 20.5 & @xmath284 & @xmath285 & @xmath286 & @xmath287 + 50 - 60 & 36.2 & @xmath288 & @xmath289 & @xmath290 & @xmath291 + 40 - 50 & 59.2 & @xmath292 & @xmath293 & @xmath294 & @xmath295 + 30 - 40 & 91.6 & @xmath296 & @xmath297 & @xmath298 & @xmath299 + 20 - 30 & 135.9 & @xmath300 & @xmath301 & @xmath302 & @xmath303 + 10 - 20 & 196.1 & @xmath304 & @xmath305 & @xmath306 & @xmath307 + 5 - 10 & 255.4 & @xmath308 & @xmath309 & @xmath310 & @xmath311 + 0 - 5 & 309.2 & @xmath312 & @xmath313 & @xmath314 & @xmath315 + cccccc + centrality ( % ) & @xmath78 & @xmath8 & @xmath9 & @xmath10 & @xmath2 + 90 - 100 & 2.2 & @xmath316 & @xmath317 & @xmath318 & @xmath319 + 80 - 90 & 4.7 & @xmath320 & @xmath321 & @xmath322 & @xmath323 + 70 - 80 & 8.4 & @xmath324 & @xmath325 & @xmath326 & @xmath327 + 60 - 70 & 14.3 & @xmath328 & @xmath329 & @xmath330 & @xmath331 + 50 - 60 & 23.3 & @xmath332 & @xmath333 & @xmath334 & @xmath335 + 40 - 50 & 36.6 & @xmath336 & @xmath337 & @xmath338 & @xmath339 + 30 - 40 & 55.4 & @xmath340 & @xmath341 & @xmath342 & @xmath343 + 20 - 30 & 81.5 & @xmath344 & @xmath345 & @xmath346 & @xmath347 + 10 - 20 & 117.4 & @xmath348 & @xmath349 & @xmath350 & @xmath351 + 5 - 10 & 152.2 & @xmath352 & @xmath353 & @xmath354 & @xmath355 + 0 - 5 & 184.8 & @xmath356 & @xmath357 & @xmath358 & @xmath359 + cccccc + centrality ( % ) & @xmath78 & @xmath8 & @xmath9 & @xmath10 & @xmath2 + 90 - 100 & 1.8 & @xmath360 & @xmath361 & @xmath362 & @xmath363 + 80 - 90 & 3.8 & @xmath364 & @xmath365 & @xmath366 & @xmath367 + 70 - 80 & 6.7 & @xmath368 & @xmath369 & @xmath370 & @xmath371 + 60 - 70 & 11.2 & @xmath372 & @xmath373 & @xmath374 & @xmath375 + 50 - 60 & 17.9 & @xmath376 & @xmath377 & @xmath378 & @xmath379 + 40 - 50 & 27.7 & @xmath380 & @xmath381 & @xmath382 & @xmath383 + 30 - 40 & 41.3 & @xmath384 & @xmath385 & @xmath386 & @xmath387 + 20 - 30 & 60.0 & @xmath388 & @xmath389 & @xmath390 & @xmath391 + 10 - 20 & 84.9 & @xmath392 & @xmath393 & @xmath394 & @xmath395 + 5 - 10 & 108.9 & @xmath396 & @xmath397 & @xmath398 & @xmath399 + 0 - 5 & 132.2 & @xmath400 & @xmath401 & @xmath402 & @xmath403 + @ccccc error & + source & @xmath404(% ) & @xmath405(% ) & @xmath406(% ) & @xmath407(% ) + @xmath34 & 0.9 , 1.0 , 0.8 & 0.1 , 0.1 , 0.2 & 0.1 , 0.9 , 2.2 & 0.0 , 0.9 , 2.0 + @xmath35 & 1.7 , 1.6 , 1.5 & 3.2 , 1.6 , 0.3 & 5.2 , 5.1 , 2.6 & 2.1 , 3.5 , 2.3 + @xmath13 & 0.1 , 0.1 , 0.0 & 0.4 , 0.5 , 0.3 & 1.0 , 2.0 , 2.5 & 0.6 , 1.5 , 2.2 + @xmath19 & 0.0 , 0.0 , 0.1 & 0.2 , 0.0 , 0.0 & 0.3 , 0.0 , 0.0 & 0.2 , 0.1 , 0.0 + @xmath20 & 0.0 , 0.0 , 0.2 & 0.3 , 0.1 , 0.0 & 0.4 , 0.2 , 0.1 & 0.1 , 0.1 , 0.0 + @xmath408-form & 0.1 , 0.0 , 0.7 & 2.2 , 0.1 , 0.2 & 3.5 , 0.3 , 0.3 & 1.2 , 0.2 , 0.1 + trigger & 0.0 , 0.0 , 0.0 & 0.1 , 0.0 , 0.0 & 0.1 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 + vertex & 0.0 , 0.0 , 0.1 & 0.1 , 0.0 , 0.0 & 0.1 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 + tracking & 0.0 , 0.0 , 0.1 & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 + @ccccc error & + source & @xmath404(% ) & @xmath405(% ) & @xmath406(% ) & @xmath407(% ) + @xmath34 & 1.0 , 1.0 , 1.0 & 0.2 , 0.4 , 0.5 & 0.2 , 1.1 , 2.6 & 0.2 , 0.8 , 2.0 + @xmath35 & 1.9 , 1.8 , 1.7 & 2.1 , 1.8 , 0.8 & 3.2 , 4.3 , 3.2 & 1.2 , 2.5 , 2.3 + @xmath13 & 0.2 , 0.2 , 0.2 & 0.2 , 0.4 , 0.4 & 0.5 , 1.5 , 2.2 & 0.3 , 1.1 , 1.8 + @xmath19 & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 & 0.2 , 0.0 , 0.1 & 0.2 , 0.1 , 0.1 + @xmath20 & 0.1 , 0.1 , 0.3 & 0.8 , 0.2 , 0.1 & 1.2 , 0.4 , 0.2 & 0.4 , 0.2 , 0.1 + @xmath408-form & 0.3 , 0.1 , 1.9 & 2.8 , 0.1 , 0.5 & 4.7 , 0.5 , 0.6 & 1.5 , 0.5 , 0.3 + trigger & 0.0 , 0.0 , 0.1 & 0.2 , 0.1 , 0.0 & 0.2 , 0.1 , 0.0 & 0.1 , 0.0 , 0.0 + vertex & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 & 0.2 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 + tracking & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 & 0.0 , 0.0 , 0.0 + .[tablex ] percent changes in collision geometry bin average quantities relative to the reference values in tables [ tableii ] - [ tablevii ] due to background contamination of the peripheral collision events as explained in sec . [ sec : results ] . background contamination errors for centrality bins not listed were negligible . [ cols="^,^,^,^,^,^ " , ] @xmath28 average percent changes within the combined centrality bin . @xmath29 reference uncertainties assuming 10% background contamination in the number of collisions in the nominal 90 - 100% centrality bin , equivalent to 1% overall background . @ccc system & @xmath409 & @xmath410 + au - au 200 gev & 379.3 & 1166 + au - au 130 gev & 377.5 & 1082 + au - au 62 gev & 374.9 & 988 + au - au 20 gev & 373.4 & 926 + cu - cu 200 gev & 115.8 & 230 + cu - cu 62 gev & 112.7 & 196 + [ [ section ] ] - multiplicity frequency distributions for high energy minimum - bias proton - proton collisions are well described by a negative binomial distribution ( nbd ) @xcite given by _ nbd(n , n , k ) & = & ( c n+k-1 + k-1 + ) ( ) ^n ( 1 + n / k)^-k . [ eqa1 ] in eq . ( [ eqa1 ] ) @xmath411 and @xmath412 are random and mean multiplicities , the variance is @xmath413 , the variance excess relative to poisson statistics is @xmath414 , and the multiplicity fluctuation variance excess per final - state particle is @xmath415 . the latter expression serves as an operational definition of parameter @xmath416 . for p - p collisions both @xmath412 and @xmath417 increase approximately linearly with ln(@xmath418 ) @xcite as expected from pqcd cross sections for parton scattering . multiplicity fluctuation variance excess per final - state particle within a given acceptance is equal to the integral of two - particle correlations on relative momentum coordinates within the same acceptance @xcite . for au - au collisions at 130 gev @xcite and at 62 and 200 gev @xcite the integral of charged particle correlations for @xmath23 and @xmath24 azimuth acceptance is dominated by a two - dimensional ( 2d ) correlation peak at small relative opening angles . the magnitude of this correlation integral increases roughly with @xmath419 . multiplicity fluctuation variance excess per final - state particle was therefore approximated by & = & [ eqa2 ] where quantity @xmath420 $ ] is the variance excess for p - p collisions , @xmath421 is the average number of binary collisions per incident participant nucleon for the most - central ( 0 - 5% ) a - a data , and @xmath422 is taken to be the ratio of the 2d correlation peak amplitude for most - central a - a collisions to that for p - p collisions . values of @xmath415 for p - p collisions were obtained from ua5 @xcite using either measured values of @xmath416 or the ua5 energy dependent fit function for @xmath416 . @xmath412 for @xmath23 is given by the values of @xmath50 assumed in this analysis . simulations demonstrated that the values of @xmath416 obtained by ua5 with very large @xmath17 acceptance remained approximately constant at the smaller @xmath17 acceptance assumed in this analysis . values of quantities @xmath420 $ ] were 0.09 , 0.24 , 0.40 and 0.51 for @xmath45 = 20 , 62 , 130 and 200 gev , respectively . @xmath422 for au - au collisions at 20 , 62 , 130 and 200 gev were estimated to be 4.8 , 5.0 , 5.4 and 5.5 , respectively , from analysis of correlation data at 130 gev @xcite assuming ln@xmath45 scaling . the remaining values for cu - cu at 62 and 200 gev were respectively 3.4 and 3.7 assuming @xmath81 scaling . finally , for each simulated collision with event - wise @xmath2 and @xmath16 from eq . ( [ eq1 ] ) nbd distribution eq . ( [ eqa1 ] ) was sampled to obtain the event - wise multiplicity . [ [ section-1 ] ] - fit recovery of the nominal @xmath6 distribution following shifts in the nuclear density parameters and @xmath13 was achieved by constructing a perturbative correction , @xmath423 , which was added to the right - hand side of eq . ( [ eq1 ] ) . parameter shifts affect the collision distributions on @xmath0 and @xmath1 and in turn the @xmath424 and @xmath6 distributions . the fit recovery term is defined such that the nominal @xmath424 distribution is maintained when the frequency distribution on @xmath0 changes . running integrals of the distributions @xmath424 and @xmath425 produce a one - to - one correspondence between @xmath426 and @xmath427 as a function of the number of summed events . using the power - law representation this running integral relation is given by , _ |n_ch , min^1/4^|n_ch^1/4 d|n_ch^ 1/4 = _ 1^(n_part/2)^1/4 d(n^_part/2)^1/4 [ eq13 ] which defines a locus of ordered pairs @xmath428 $ ] , which can be compactly expressed as a discrete function defined by _ ch^1/4 & = & n^1/4_ch , d^ d [ ( n_part/2)^1/4 ] . [ eq14 ] in eqs . ( [ eq13 ] ) and ( [ eq14 ] ) subscripts @xmath429 and @xmath430 refer to the nominal ( reference ) distribution when @xmath431 and to the shifted distribution when @xmath432 . lower limit @xmath433 for the reference distribution . in order to recover the nominal @xmath424 distribution the correction term must be computed by the following difference , ( ) & = & n_ch , s r [ ( ) ^ ] - n_ch , s s [ ( ) ^ ] . [ eq15 ] throughout this fit recovery procedure the parameters of @xmath58 in eq . ( [ eq2 ] ) remained fixed . numerically stable results with one - million minimum - bias simulated collisions required coarse binning of the monte carlo distributions @xmath434 and @xmath435 prior to evaluation of eqs . ( [ eq13 ] ) - ( [ eq15 ] ) . when parameters @xmath34 and @xmath13 were shifted excellent fit recovery was achieved with centrality bins corresponding to 0 - 5% , 5 - 10% , 10 - 40% , 40 - 70% and 70 - 100% total cross section fractions . when diffusivity parameter @xmath35 was shifted the optimum binning was 0 - 5% , 5 - 10% , 10 - 20% , 20 - 40% , 40 - 60% , 60 - 80% and 80 - 100% . the power - law representation in eq . ( [ eq13 ] ) exploits the approximately uniform statistics throughout the domain of the integrands , making numerical stability of the results less problematic . variations in the density parameters and @xmath13 produced typical changes in @xmath78 as a function of centrality of order several percent , up to 5% . fit recovery in terms of @xmath78 was typically within a few tenths of a percent and always less than about 1% where the magnitude of @xmath423 relative to @xmath436 was less than about 1% , 4% and 1.5% for changes in @xmath34 , @xmath35 and @xmath13 , respectively , for all six collision systems . a similar method was used to account for shifts in @xmath6 representing the systematic uncertainties in trigger inefficiency , background contamination , and collision vertex reconstruction inefficiency . for these simulations @xmath437 remained fixed while @xmath438 varied . the required correction is given by , ( ) & = & n_ch , r s [ ( ) ^ ] - n_ch , r r [ ( ) ^ ] . [ eq16 ] [ [ section-2 ] ] - the monte carlo simulation model discussed in sec . [ sec : discussion ] for transverse particle production and energy deposition in the star ctb for p - p and au - au minimum bias collisions is described here . for p - p collisions the interaction point along the beam - line was randomly selected within @xmath3125 cm of the geometrical center of the star detector @xcite . the minimum - bias charged particle multiplicity was obtained by sampling the negative binomial distribution ( nbd ) fitted to the ua5 @xmath45 = 200 gev p - p data @xcite with parameters @xmath439 and @xmath440 . the @xmath5 , @xmath17 and particle species ( @xmath441 , k@xmath442 , proton or antiproton ) for each charged particle produced in the collision was sampled from the measured @xmath443 , @xmath444 , and particle species distributions , respectively . distribution @xmath443 was assumed to be proportional to @xmath445 , where @xmath446 , @xmath447 is the particle rest mass , and inverse effective temperature @xmath448 = 5.7 , 5.2 and 4.8 gev@xmath449 , for pions , kaons and protons , respectively . the @xmath448 values were obtained by fitting data @xcite within the approximate range @xmath450 gev . @xmath5 was restricted to be less than 2 gev/@xmath65 . distribution @xmath444 was obtained from the ua5 @xcite measurements for @xmath45 = 200 gev non - singly diffractive p - p collisions for @xmath451 . the particle species probabilities for pions , kaons and protons at low momentum were assumed to be 0.85 , 0.085 and 0.065 , respectively @xcite . particle trajectories were extended outward from the collision point and approximately transverse to the beam direction toward the star central trigger barrel ( ctb ) detector @xcite which was approximated by a uniform cylinder 220 cm in radius , coaxial with the beam line , and having full @xmath24 azimuthal coverage and total longitudinal length of 484 cm located symmetrically about the geometrical center of the star time projection chamber @xcite . charged particles were propagated in a 0.5 t solenoidal magnetic field as helices until they either intersected or missed the ctb . the simulated mean number of charged particles per unit pseudorapidity at @xmath452 intersecting the ctb was consistent with the measured @xmath50 from ua5 @xcite . energy deposition in the plastic scintillator material ( ch ) was estimated by sampling the landau distribution @xcite for the calculated most - probable energy loss @xcite for 1 cm thick plastic taking into account the range of crossing angles between the helical trajectories and the cylindrical detector . the areal density ( effective thickness ) along the path through the material was estimated to be ( 1.032 gm/@xmath453 @xcite where @xmath113 is the angle of incidence between the particle trajectory helix and a normal to the detector cylinder at the intersection point . the mean electron excitation ( ionization ) energies for carbon and hydrogen were assumed to be 80 ev and 20 ev , respectively @xcite . energy depositions in the carbon and hydrogen components of the detector material were added together . light attentuation in the plastic scintillator ( 380 cm attenuation length @xcite ) was included in the simulated ctb output signal . for each p - p collision the total output from each produced charged particle which deposited energy in the ctb scintillators was summed , resulting in a simulated result for the quantity @xmath133 in sec . [ sec : discussion ] . the p - p results in fig . [ fig4 ] correspond to one - million triggered events with non - zero ctb signal . the resulting trigger output frequency distribution , @xmath454 , served as input for the au - au trigger simulations described next . for au - au collisions the monte carlo glauber model presented here was used to generate an ensemble of @xmath3 = 200 gev simulated minimum - bias collisions . for each au - au collision the @xmath454 distribution was sampled @xmath455 times corresponding to the average number of p - p collisions required to generate mean multiplicity @xmath16 . the sampled values of @xmath133 were added to obtain the total ctb trigger output for each simulated au - au collision . the results in fig . [ fig4 ] correspond to one - million au - au triggers with non - zero ctb output . 99 glauber r j 1959 _ lectures in theoretical physics _ , edited by brittin w e and dunham l g ( new york : interscience ) 315 glauber r j 1967 _ high energy physics and nuclear structure _ , edited by alexander g ( amsterdam : north - holland ) 311 glauber r j 1970 _ high energy physics and nuclear structure _ , edited by devons s ( new york : plenum ) 207 . adler c _ et al _ ( star collaboration ) 2001 _ phys . lett . _ * 87 * 112303 adler c _ et al _ ( star collaboration ) 2002 _ phys . lett . _ * 89 * 202301 adams j _ et al _ ( star collaboration ) 2006 _ phys . _ c * 73 * 034906 ray l , coker w r , hoffmann g w 1978 _ phys . _ c * 18 * 2641 ray l 1979 _ phys . 1855 ray l , hoffmann g w , barlett m , mcgill j , amann j , adams g , pauletta g , gazzaly m , and blanpied g s 1981 _ phys _ c * 23 * 828 pauletta g _ et al _ 1981 _ phys . _ b * 106 * 470 hoffmann g w _ et al _ 1978 _ phys . _ b * 79 * 376 hoffmann g w _ et al _ 1980 _ phys . rev . _ c * 21 * 1488 back b b _ et al _ ( phobos collaboration ) 2001 _ phys . lett . _ * 87 * 102303 back b b _ et al _ ( phobos collaboration ) 2002 _ phys _ c * 65 * 061901(r ) back b b _ et al _ ( phobos collaboration ) 2004 _ phys . _ c * 70 * 021902(r ) adams j _ et al _ ( star collaboration ) 2005 _ phys . _ b * 616 * 8 adams j _ et al _ ( star collaboration ) 2006 _ phys . _ b * 637 * 161 adler s s _ et al _ ( phenix collaboration ) 2006 _ phys _ c * 74 * 024904

the properties of the kaons and antikaons in the nuclear medium have been the object of numerous investigations since the possibility of the existence of a kaon condensed phase in dense nuclear matter was pointed out @xcite . the enhancement of the @xmath0 yield in ni+ni collisions measured recently by the kaos collaboration at gsi @xcite can be explained by assuming the @xmath0 meson to feel a strong attraction in the medium @xcite . kaonic atom data , a compilation of which is given in ref . @xcite , also favor an attractive @xmath0 nucleus interaction . most of the recent theoretical works start from chiral lagrangians that reproduce the free space scattering properties which , in the case of @xmath8 scattering , are dominated by the presence of an isospin zero resonance , the @xmath6 . it is precisely the influence of this resonance what makes the @xmath9 interaction repulsive at threshold , while the lowest order born s - wave amplitude from the chiral lagrangian ( the tomozawa - weinberg term ) is attractive . evidently , one can not expect chiral perturbation theory to work close to a resonance . this is why all works studying the free space @xmath10 interaction from the chiral lagrangian either introduce the @xmath6 as an elementary field @xcite or generate it dynamically through the lippmann - schwinger @xcite or the bethe - salpeter @xcite equations . these latter approaches allow for a microscopic incorporation of the medium effects on the the @xmath5 interaction @xcite . for instance , pauli blocking on the intermediate nucleon states , makes the @xmath5 interaction density dependent and this , in turn , modifies the @xmath0 properties from those in free space . these medium modifications were already included long time ago in the context of brueckner - type many body theory @xcite to obtain the kaon - nucleus optical potential for kaonic atoms . the medium properties of kaons and antikaons have also been obtained from mean field theories , built within the framework of chiral lagrangians @xcite , based on the relativistic walecka - type model extended to incorporate strangeness in the form of hyperons or kaons @xcite , or using explicitly quark degrees of freedom @xcite . all the different approaches agree qualitatively in establishing that , in the medium , the @xmath11 feels a moderate repulsion and the @xmath0 a strong attraction . how large is this attraction is still somewhat controversial . recent phenomenological approaches @xcite based on fits to kaonic atom data find a @xmath0-nucleus potential of the order of @xmath12 mev in the nuclear center . however , no calculation that starts from the bare @xmath13 interaction predicts such an attraction , the values ranging from @xmath14 to @xmath15 mev . hopefully , heavy - ion reactions , that are sensitive to higher density regions , will help in elucidating these discrepancies . at the same time , it is necessary to develop theories that treat the intricacies associated to the mutual interaction between all the hadrons in the medium as accurately as possible . in the present work we perform a microscopic study of the the @xmath0 properties in nuclear matter by incorporating the medium modifications on the @xmath5 amplitude using the model of ref . @xcite , which was shown to reproduce the @xmath9 scattering observables very satisfactorily . as mentioned above , one source of density dependence is the pauli blocking on the nucleon states . this makes the @xmath5 interaction attractive and , on the other hand , shifts the resonance to higher energies @xcite . however , it was shown recently @xcite that a self - consistent calculation of the @xmath0 self - energy leaves the position of the resonance unchanged , due to a compensation of the repulsive @xmath6 shift with the attraction felt by the @xmath0 meson . the importance of these medium effects makes it interesting to investigate other medium modifications of the particles participating in building up the @xmath5 interaction . this is the aim of the present work . to this end , we include , in addition , the dressing of the pions in the @xmath16 , @xmath17 intermediate states , which couple strongly to the @xmath18 state . this is done through a pion self - energy that contains the effect of one- and two - nucleon absorption , conveniently modified to include the effect of nuclear short - range correlations . the medium effects on the nucleons and hyperons is considered via density - dependent mean - field binding potentials . after reviewing the free - space formalism in sect . [ sec : free ] , we describe , in sect . [ sec : medium ] , the appropiate modifications needed to incorporate the medium effects on the @xmath1 amplitude . the details of the kaon and pion self - energies are given in sect . [ sec : self ] . our results are discussed in sect . [ sec : results ] , where various approximations are compared . finally , sect . [ sec : conclusions ] summarizes our main results and conclusions . in this section we review the formalism used in ref . @xcite for describing @xmath1 scattering in free space . the effective chiral lagrangian formalism has been very successful in explaining the properties of meson - meson interaction at low energies @xcite , as well as those of the meson - baryon systems @xcite when the interaction is weak , as in the case of the s - wave @xmath19 and @xmath20 interaction . however , the @xmath21 system couples strongly to many other channels and generates a resonance below threshold , the @xmath22 . in this case the standard chiral scheme , an expansion in powers of the typical momenta involved in the process , fails to be an appropriate approach . in ref . @xcite a non perturbative scheme , consisting of solving a set of coupled - channel bethe - salpeter equations using the lowest order chiral lagrangian in s - wave , was shown to reproduce the low energy @xmath8 scattering data very satisfactorily with only one - parameter , the cut - off used to renormalize the loop integrals . a reinterpretation of these equations to the light of the inverse amplitude method can be seen in ref . @xcite , where it is shown that the effect of higher order chiral lagrangians can be reabsorbed with the choice of a suitable cut - off in some cases . the model of ref . @xcite follows closely that of ref . @xcite , where the success of the non - perturbative approach using the chiral lagrangians was first shown , but includes the complete set of @xmath23 meson and @xmath24 baryon octet states in the basis space of coupled channels . the use of this complete set , apart from providing the right su(3 ) symmetry in the limit of equal baryon masses and equal meson masses , was found to be essential for reproducing the experimental branching ratios at threshold with only the lowest order lagrangian . the lowest order chiral lagrangian , coupling the octet of pseudoscalar mesons to the octet of @xmath24 baryons , is @xmath25 \rangle \ , \end{aligned}\ ] ] where the symbol @xmath26 denotes the trace of su(3 ) matrices and @xmath27 \\ \gamma_{\mu } = \frac{1}{2 } ( u^\dagger \partial_{\mu } u + u \partial_{\mu } u^\dagger ) \\ u = u^2 = { \rm exp } ( i \sqrt{2 } \phi / f ) \\ u_{\mu } = i u ^\dagger \partial_{\mu } u u^\dagger \ . \end{array}\ ] ] the su(3 ) matrices for the mesons and the baryons are the following @xmath28 @xmath29 at lowest order in momentum the interaction lagrangian reduces to @xmath30 \rangle \ .\ ] ] the coupled channel formalism requires to evaluate the transition amplitudes between the different channels that can be built from the meson and baryon octets . for @xmath8 scattering there are 10 channels , namely @xmath9 , @xmath31 , @xmath32 , @xmath33 , @xmath34 , @xmath35 , @xmath36 , @xmath37 , @xmath38 and @xmath39 . in the case of @xmath40 scattering the coupled channels are : @xmath41 , @xmath42 , @xmath43 , @xmath44 , @xmath45 and @xmath46 . these amplitudes have the form @xmath47 where @xmath48 are the initial , final momenta of the baryons ( mesons ) . the explicit values of the coefficients @xmath49 can be found in ref . @xcite . at low energies we can neglect the spatial components and eq . ( [ eq : v - matrix ] ) simplifies to @xmath50 the coupled - channel bethe - salpeter equations in the center of mass frame read @xmath51 where the indices @xmath52 run over all possible channels and @xmath53 with @xmath54 and @xmath55 being , respectively , the baryon mass , baryon energy and meson mass in the intermediate state . although eq . ( [ eq : loop ] ) requires the half - off - shell amplitudes , it was shown in ref . @xcite that the off - shell part goes into renormalization of coupling constants . therefore , one can factorize the remainig on - shell components of @xmath56 and @xmath57 outside the integral of eq . ( [ eq : loop ] ) reducing the integral equation to a set of algebraic equations . the loop integral reads @xmath58 with @xmath59 . the model discussed here depends on the loop regularization cut - off , whose value @xmath60 mev was chosen to reproduce the @xmath8 scattering branching ratios at threshold @xcite . the weak decay constant is @xmath61 , a value lying in between the pion and kaon ones that was chosen to optimize the position of the @xmath6 resonance . in table [ tab:1 ] we summarize the predictions of the model for several scattering observables at threshold . the scattering cross sections , which are not used in the fit , were shown to be in good agreement with the low energy data @xcite and the model can safely be used up to 500 mev / c @xmath0 lab momentum @xcite up to some punctual discrepancy in the small @xmath62 cross section around 400 mev / c , where the d - wave @xmath63 resonance , not accounted for in the theory , shows up in the data . the dynamics of the @xmath18 interaction at low energies is dominated by the @xmath6 resonance , which in the approaches of refs . @xcite is interpreted as an isospin @xmath64 quasi - bound @xmath8 state . the @xmath6 resonance is slightly below the @xmath9 threshold and leads to a repulsive @xmath9 amplitude . although the isospin @xmath65 @xmath41 channel is attractive @xcite , the @xmath66 averaged amplitude , @xmath67 , is still repulsive and , according to the low energy theorem , the @xmath0 self - energy ( @xmath68 ) should also be repulsive . however , kaonic atom data suggest that , even at a small fraction of the normal nuclear matter density @xmath69 , the @xmath0 feels a strongly attractive potential . this implies a rapid transition from a repulsive @xmath5 interaction to an attractive one as density increases and a microscopic study of the @xmath0 properties in the medium can not be done in terms of the simple @xmath68 or impulse approximation . it is therefore necessary to consider the density dependence of the in - medium @xmath5 interaction , @xmath70 . one source of density dependence comes from the pauli principle , which prevents the scattering to intermediate nucleon states below the fermi momentum , @xmath71 . to incorporate this effect in the channels having an intermediate nucleon state , one must replace the free nucleon propagator by the in - medium one in the loop integral of eq . ( [ eq : gprop ] ) . the evaluation of the @xmath0 self - energy will require the knowledge of the @xmath10 @xmath57-matrix at momenta @xmath72 in the nuclear lab frame , where @xmath73 is a nucleon momentum from the fermi sea and @xmath74 is the momentum of the @xmath0 . the loop integral of eq . ( [ eq : gprop ] ) used in the description of the @xmath10 scattering data is evaluated and regularized by means of a cut - off in the center - of - mass frame . in order to be able to use the same formalism we boost the @xmath10 system to its center - of - mass frame and then evaluate the loop function . however , since pauli blocking is most easily implemented in the lab frame , we express the occupation number in terms of the momentum of the intermediate nucleon in the laboratory , @xmath75 . hence , we have @xmath76 where @xmath77 is the total four - momentum in the lab frame , @xmath78 and @xmath79 \vec{p } - \vec{q } \label{eq : lorentz}\ ] ] is the nucleon momentum in the lab frame corresponding to a momentum @xmath80 in the center - of - mass frame . as one can see from eq . ( [ eq : gpauli0 ] ) , the pauli blocking corrections are only operative at the nucleon pole . indeed , by collecting the terms proportional to @xmath81 in the nucleon propagator one finds @xmath82 , which sets the value of @xmath83 to be used in eq . ( [ eq : lorentz ] ) . this allows one to perform the integration over @xmath83 in eq . ( [ eq : gpauli0 ] ) analytically resulting in @xmath84 in practice , we have checked that the value @xmath85 to be used in eq . ( [ eq : lorentz ] ) can be replaced , with negligible changes , by the energy of the nucleon when both the meson and the baryon are placed on - shell @xmath86 the second term on the right hand side of eq . ( [ eq : gpauli ] ) , related to the hole part of the nucleon propagator , is purely real for the kinematical conditions studied in this work , since one always finds @xmath87 . all works studying the pauli blocking effects on the in - medium @xmath1 interaction @xcite have neglected the hole term of eq . ( [ eq : gpauli ] ) , a simplification that is well justified for low densities since the phase space for holes is much reduced with respect to that for particles . however , this effect can become important if one wants to find the properties of the @xmath3 at high densities as those achieved in heavy ion reactions or neutron stars . another source of density dependence is related to the fact that all the mesons and baryons participating in the intermediate loops interact with the nucleons of the fermi sea and , as a consequence , feel a binding potential which changes the threshold energy of the different channels . the binding effects on the baryons are taken within a mean - field approach consisting in adding , to the single particle energies in eqs . ( [ eq : gprop ] ) and ( [ eq : gpauli ] ) , a momentum - independent potential of the type @xmath88 , where @xmath89 @xmath90 is normal nuclear matter density . in the case of nucleons a reasonable value is @xmath91 mev as implied by numerous nuclear structure data . on the other hand , since the experimental binding energy of a @xmath92 particle in hypernuclei extrapolates to about 30 mev for the very heavy systems @xcite , we take the potential depth of the @xmath92 to be @xmath93 mev . finally , although there is very little information on the @xmath94 single particle states , the recent measurement of the @xmath94 binding energy of about 4 mev in @xmath95he @xcite indicates that the @xmath94 can certainly be bound in the nucleus . moreover , analysis of @xmath94-atom data by batty et al @xcite and from ref . @xcite found the data to be compatible with a potential of the type @xmath96 with @xmath97 about @xmath98 to @xmath99 mev . based on this evidence , we take @xmath93 mev for the @xmath94 potential . the nuclear medium effects on the mesons will be included through the corresponding self - energy . we will only consider the dressing of the @xmath4 and @xmath2 mesons since the @xmath100 and @xmath101 mesons appear in intermediate states that lie quite far above the @xmath8 threshold . in the next section , we show how the @xmath3 and @xmath2 self - energies are constructed . here , we will focus on how the dressed meson propagator is incorporated into the scheme . to this end , we write the meson propagator ( @xmath102 ) @xmath103 in the lehmann representation @xmath104 where the spectral density , defined as @xmath105 has been introduced . the free meson propagator in eq . ( [ eq : gpauli ] ) must now be replaced by the dressed one written as in eq . ( [ eq : dressed ] ) . while the @xmath83 integral proceeds as before , there now appears an additional integration over the variable @xmath106 running over all possible excited states to which a meson of momentum @xmath107 can couple . the new meson - baryon loop integral is ( @xmath108 ) @xmath109 where @xmath110 for the intermediate states involving hyperons ( @xmath16 and @xmath17 ) . if only pauli blocking effects are considered , the effective interaction @xmath111 is obtained by solving the coupled - channel bethe - salpeter equation using the meson - baryon propagator of eq . ( [ eq : gpauli ] ) for the intermediate loops involving nucleons and that of eq . ( [ eq : gprop ] ) otherwise . if one also incorporates the dressing of the mesons , one must solve the bethe - salpeter equation using the meson - baryon propagator of eq . ( [ eq : gmed ] ) . the model discussed in sect . [ sec : free ] for the @xmath1 interaction gives rise to a s - wave @xmath3 self - energy ( @xmath112 or @xmath113 ) @xmath114 \ , \label{eq : selfka}\ ] ] which is obtained by summing the in - medium @xmath5 interaction , @xmath115 ( @xmath116 ) , over the nucleons in the fermi sea . the values @xmath117 stand now for the energy and momentum of the @xmath3 in the lab frame . note that a self - consistent approach is required since one calculates the @xmath4 self - energy from the effective interaction @xmath118 which uses @xmath4 propagators which themselves include the self - energy being calculated . we also include a p - wave contribution to the @xmath4 self - energy coming from the coupling of the @xmath4 meson to hyperon particle - nucleon hole ( @xmath119 ) excitations . the @xmath0 meson can couple to @xmath120 , @xmath121 or @xmath122 and the @xmath113 to @xmath123 , @xmath124 or @xmath125 . the vertices @xmath126 are easily derived from the @xmath127 and @xmath128 terms of eq . ( 1 ) , expanding @xmath129 up to one meson field . using a non - relativistic reduction of the @xmath130 matrix , one finds @xmath131 where @xmath107 is the kaon momentum . we take @xmath132 and @xmath133 . the coefficients @xmath134 and @xmath135 are listed in table [ tab:2 ] . the p - wave self - energy for the @xmath0 meson reads @xmath136 where @xmath137 is the generalized lindhard function for different particle and hole masses in a medium of density @xmath138 , with @xmath139 or @xmath140 . as shown in the appendix of ref . @xcite , it reads @xmath141 where @xmath142 and @xmath143 an analogous expression is found for the @xmath113 meson . in symmetric nuclear matter , as is considered here , both self - energies are the same and reduce to @xmath144 where @xmath145 and @xmath146 are the generic @xmath147 and @xmath148 coupling constants , respectively . this p - wave contribution to the @xmath4 self - energy can become important for large momentum values . the solution of the new @xmath0 dispersion relation @xmath149 with @xmath150 , determines the effective mass , @xmath151 , and decay width , @xmath152 , of the @xmath0 meson in the medium . for the pion self - energy we take that of ref . @xcite which consists of a small momentum independent s - wave part , @xmath153 , plus a p - wave part , @xmath154 . the latter , constructed by allowing the pion to couple to particle - hole ( @xmath155 ) , @xmath156-hole ( @xmath157 ) and two - particle - hole ( @xmath158 ) excitations , can be written as @xmath159 where @xmath160 is a monopole form factor with cut - off @xmath161 gev , @xmath162 ( @xmath163 ) is the lindhard function for @xmath155 ( @xmath157 ) excitations and @xmath164 is a phenomenological function accounting for @xmath158 excitations . the @xmath164 contribution at @xmath165 was obtained from fits to pionic atom data @xcite . the extension to kinematical regions away from pionic atoms was done by multiplying the imaginary part by the available phase space of @xmath158 states for the given @xmath83 , @xmath107 values at density @xmath166 . the explicit expression can be found in ref . @xcite . the pion self - energy is further modified to include the effect of nuclear short - range correlations . this is accomplished by replacing the self - energy of eq . ( [ eq : pself ] ) by @xmath167 where @xmath168 is the sum of the @xmath169 , @xmath170 and @xmath164 functions , and @xmath171 is the usual landau - migdal parameter , which is taken slightly momentum dependent as in ref . @xcite . we have also considered the effects of including form factors at the @xmath172 vertices of eq . ( [ eq : kpwave ] ) . if a monopole form factor with @xmath173 gev normalized to 1 as @xmath174 is used , the changes observed are very small , at the level of 2% . since the value of @xmath175 for kaons is not so well known and the changes induced by the form factor are so small , we simply ignore it in the results presented below . in this section we discuss the results of our calculations . to facilitate the comparison with other works we will distinguish three types of approximations , all of which include the binding effects on the baryons : * _ pauli _ : the blocking of intermediate nucleon states below the fermi momentum @xmath71 is taken into account , but the mesons propagate as in free space . this approach uses the meson - baryon propagator of eq . ( [ eq : gpauli ] ) for the intermediate states involving nucleons and has been the standard medium effect studied in the literature @xcite . * _ in - medium kaons _ : apart from pauli blocking , the dressed @xmath7 propagator is included in the loops in a self - consistent manner . this approach , which uses eq . ( [ eq : gmed ] ) for the intermediate @xmath5 states , has recently been used in the study of the @xmath0 properties in the medium @xcite . * _ in - medium pions and kaons _ : the dressing of the pions is also considered and , therefore , eq . ( [ eq : gmed ] ) is used for the intermediate channels involving both @xmath4 and @xmath2 mesons . this is the main novelty of the present work , apart from considering also the baryon potentials . to assess the importance of dressing the mesons , we show first their spectral density . in fig . [ fig : specpi ] the spectral density of the @xmath2 mesons in nuclear matter at density @xmath176 is shown as a function of energy for several momenta . note that the scale has been adapted to the two higher momenta such that the structure of the spectral density is more clearly seen . the peak value of the @xmath177 mev / c spectral function is about 375 gev@xmath178 . as the momentum increases the position of the peak moves to higher energies . however , it increasingly deviates from the energy @xmath179 at which a delta - type spectral density would be located if the dressing of the @xmath2 meson was ignored . the in - medium peak positions are at 158 , 179 and 212 mev for @xmath177 , 200 and @xmath180 mev / c , respectively , while the corresponding free values are 172 , 244 and 331 mev . we also observe , to the left of the peaks , the typical structure of the @xmath155 excitations , which spreads the pionic strength over a wider energy range . the spectral funcion of a @xmath0 meson of zero momentum is shown in fig . [ fig : specka ] as a function of energy for the three types of approximations used in this work . the solid line shows results for nuclear matter density @xmath176 , the short - dashed line for @xmath181 and the long - dashed line for @xmath182 . the spectral functions shown in the upper panel correspond to the approximation in which only pauli blocking effects have been considered . for @xmath182 a two - mode excitation is clearly visible . the left peak corresponds to the @xmath0 pole branch , appearing at an energy smaller than its mass , @xmath183 , due to the medium effects which , as we will see , are already attractive at this density . the upper peak corresponds to the @xmath6-hole state , which is located above @xmath183 because of the shifting of the @xmath6 resonance to energies above the @xmath9 threshold due to the restrictions on phase space imposed by pauli blocking . as density increases , the @xmath0 feels an enhanced attraction and the @xmath0 pole peak moves to lower energies . at the same time the reduced phase space for the intermediate nucleon states makes the appearance of the resonance less likely . the @xmath6-hole peak moves to higher energies and loses strength , a reflection of the tendency of the @xmath6 to dissolve in the dense nuclear medium . these results are in complete agreement with those obtained by waas and weise @xcite . when the @xmath4 spectral function displayed in fig . [ fig : specka ] is insterted in the coupled - channel bethe - salpeter equation , the intermediate @xmath5 states are more spread out over energies . as a consequence , the resulting in - medium interaction , @xmath118 , and the new self - energy calculated from eq . ( [ eq : selfka ] ) , become softened and distributed over a wider range of energies , affecting in turn the distribution of the @xmath4 spectral strength . after a few iterations of this process , self- consistency is achieved and the resulting @xmath0 spectral function is displayed in the middle panel of fig . [ fig : specka ] . the two peak mode is barely visible now . as was noted by lutz @xcite , including the in - medium attraction felt by the @xmath0 through the use of the dressed propagator lowers the threshold for the @xmath5 states that had been increased by the pauli blocking on the nucleon . this has a compensatory effect and the resonance barely moves with respect to its free space value . the @xmath0 pole peak appears at similar or slightly smaller energies , but its width is larger , due to the strength of the intermediate @xmath5 states being distributed over a wider region of energies . therefore the @xmath0 pole and the @xmath6-hole branches merge into one another and can hardly be distinguished . finally , the self - consistent @xmath0 spectral function when the pions in the intermediate loops are also dressed is shown in the bottom panel of fig . [ fig : specka ] . the essential difference with respect to the other two approaches is the fact that the @xmath184 channels ( @xmath185 or @xmath94 ) start to be operative at lower energies due to the attraction felt by the pions in the medium . as seen by the long - dashed line , even at very small densities one no longer distinguishes the @xmath6-hole peak from the @xmath0 pole one . as density increases the attraction felt by the @xmath0 is much more moderate and the @xmath0 pole peak appears at higher energies than in the other two approaches . at normal nuclear matter density a peak starts to develop to the left of that corresponding to the @xmath0 pole . as will be shown next , this structure is associated to a cusp in the imaginary part of the @xmath0 self - energy at an energy corresponding to the position of the new in - medium @xmath186 threshold . the @xmath0 self - energy at zero momentum , @xmath187 , is displayed in fig . [ fig : selftot ] as a function of the @xmath0 energy for several densities and for the three approximations discussed here . the graphs on the left show the real part and those on the right the imaginary part . we have seen that the @xmath4 spectral strength is distributed over a wider range of energies and incorporates the attractive medium effects felt by the @xmath4 . this is why the self - energy shown in the middle panels ( _ in - medium kaons _ ) is smoother and wider than that displayed on the top ( _ pauli _ ) ones . note that @xmath188 in the middle panel is non - zero at energies as low as 300 mev . this corresponds to the @xmath189 and @xmath190 many - body channels that are opened at these energies and which are incoporated in the calculation via the @xmath0 p - wave self - energy . in the lower panels , where the pions are also dressed , we see that @xmath191 has bigger strength at these low energies . this can be understood in the following way : the @xmath192 channel is one of the important building blocks of the @xmath6 strength below the @xmath1 threshold . when the pions are dressed in the medium they couple strongly to @xmath155 and @xmath158 components , so that the @xmath0 is effectively coupling to intermediate @xmath193 , @xmath194 states with lower threshold than the @xmath192 channel . actually , the opening of this latter channel on top of the already existing @xmath193 and @xmath194 is visible through a cusp in @xmath195 around @xmath196 mev . the effect on the @xmath0 self - energy of the baryon binding potentials is basically a shift in the scale of energies , corresponding to the difference of potentials between the nucleon and the @xmath94 hyperon , which amounts to 20 mev at @xmath176 . the scattering properties of a @xmath4 in the nuclear medium will depend on the characteristics of the in - medium @xmath5 scattering amplitude . in particular , it is interesting to see if the resonant shape of the free @xmath8 scattering amplitude remains when the medium effects are incorporated . [ fig : ampltot ] shows the real part ( on the left ) and imaginary part ( on the right ) of the @xmath197 scattering amplitude for a total momentum @xmath198 as a function of @xmath199 . results are shown for several densities and for the three types of approximations . on the top panels we can see how , as density increases , the pauli blocking effects merely shift the resonance to higher energies , hence changing from being 27 mev below the @xmath8 threshold to being above it from a certain density on . note , however , that the resonance still appears , for each density , below the new threshold imposed by the pauli principle , i.e. @xmath200 . on the other hand , the resonance shape remains pretty much unaltered . when the dressing of the @xmath4 meson is also incorporated , we observe how the resonant shape smears out quite fast with increasing density . moreover , the resonance does not move up to higher energies but stays close to the free space value due to a cancellation between the repulsive pauli blocking effects and the attractive medium effects on the @xmath4 . these results are in qualitative agreement with the self - consistent calculation of lutz @xcite , although we seem to observe a faster dilution of the resonance with the medium density . finally , when the in - medium dressing of the pions are also included we obtain a quite similar pattern . the resonance dissolves very fast although it shows a tendency to move to higher energies . the isospin averaged in - medium scattering length defined as @xmath201 with @xmath202 , is shown if fig . [ fig : scatlen ] . the change of @xmath203 from negative to positive values indicates the transition from a repulsive interaction in free space to an attractive interaction in the medium . as shown by the dotted line , this transition happens at a density of about @xmath204 when only pauli effects are considered , in agreement with what was found in ref . @xcite . however , this transition occurs at even lower densities ( @xmath205 ) when one considers the dressing of the mesons in the description , as seen from the dashed ( _ in - medium kaons _ ) and full ( _ in - medium pions and kaons _ ) lines . the deviations from the approach including only pauli blocking or those dressing the mesons are quite appreaciable over a wide range of densities . the effective @xmath0 mass is shown in fig . [ fig : emass ] as a function of @xmath206 for the different approximations . note that the effective mass , defined as the solution of eq . ( [ eq : disp ] ) for @xmath207 , gives the position of the @xmath0 pole peak in the spectral function assuming a weak energy dependence of @xmath208 . in addition , one can only think of the @xmath209 as a quasi - meson moving with the modified @xmath210 mass if the spectral function shows a clear and narrow enough @xmath0 pole peak , which happens when @xmath211 is sufficiently small . by looking at the @xmath0 spectral function of fig . [ fig : specka ] one sees that these conditions are not always fulfilled , especially when the dressing of the mesons is incorporated . keeping this in mind , we can still analyze the effective mass as a simplified way to assess how important is the attraction felt by the @xmath0 meson in the medium . morover , this analysis will allow to compare our results with those obtained by other calculations . as can be seen in the zoomed area on the top right of the figure , the in - medium @xmath0 mass increases slightly from the free value but quickly gains attraction as density increases . the transition from repulsion to attraction occurs around @xmath212 for the _ pauli _ approximation ( dotted line ) while the two other approximations that incorporate the dressing of the @xmath4 ( dashed line ) and also that of the pion ( full line ) show attraction already at around @xmath213 . however , the density dependence is stronger with the _ pauli _ approximation which gives rise to a lower value of @xmath210 from @xmath214 onwards . it turns out that , in spite of the essentially different shapes of the @xmath0 spectral function , the position of the @xmath0 pole is very similar for the _ pauli _ and the _ in - medium kaons _ approximations for densities higher than @xmath69 . consequently , the calculated effective masses decrease similarly . however , when the dressing of the pions is also included , the attraction is substantially more moderate and the effective mass seems to level off at high densities around the value achieved already at @xmath69 of @xmath215 mev . this would correspond to a @xmath0 optical potential of @xmath216 mev . pauli _ and _ in - medium kaons _ approximations give rise to a a more attractive value of the effective mass at @xmath69 , namely @xmath217 mev which corresponds to a @xmath218 mev . it is worth mentioning that all the calculations , either microscopic @xcite or based on mean field theories @xcite predict effective masses at @xmath69 which range between 375 mev and 425 mev . the corresponding @xmath0 potential at the center of the nucleus would then be in between @xmath14 and @xmath15 mev . therefore , no theory is able to explain the large attraction of @xmath219 obtained from the best available fit to kaonic atom data @xcite . however , we should warn that the data of @xmath0 atoms are only sensitive to the @xmath0 self - energy at the low effective densities felt by the kaons that explore basically the nuclear surface . the values obtained by the fit in ref . @xcite at @xmath176 are an extrapolation of a preassumed functional dependence on @xmath166 . it is possible to obtain equally good fits with other functionals that give the same value for the low densities explored by the kaons but different extrapolations at @xmath176 . in fact , quite good fits to the @xmath0 atoms data are also obtained with the batty potential of ref . @xcite which is of the order of @xmath220 mev . we also note that the brueckner - type calculations of ref . @xcite obtained a shallow @xmath0-nucleus potential , of the order of @xmath221 mev at the center of @xmath222c , and predicted reasonably well the data available at that time . with the values obtained here for @xmath210 and similar ones that we obtain for neutron matter the phenomenon of @xmath0 condensation appears very unlikely , since there is not enough attraction for the @xmath0 energy to become smaller than the electron chemical potential in beta stable neutron star matter . the implications of our results for the in - medium @xmath0 self - energy on kaonic atoms is currently being analyzed @xcite and gives results compatible with the existing kaonic atom data . that analysis , based on a local density approximation , requires the knowledge of the optical potential , defined as @xmath223 where @xmath224 is the on - shell momentum value that fulfills eq . ( [ eq : disp ] ) for a given value of @xmath106 . the real and imaginary parts of the optical potential for the _ in - medium pions and kaons _ approximation are shown in fig . [ fig : optpotpika ] for three energy values : @xmath225 mev ( dotted line ) , @xmath226 ( solid line ) and @xmath227 mev ( dashed line ) . the self - energy shown by the solid lines is the relevant one for studies of low lying @xmath0 atomic states , while the results displayed by the dotted line would be the ones to use for very deeply bound states in the pb region . the dashed lines show an example of the @xmath0 optical potential at positive energies and would be the one to consider to treat the @xmath0 distortions in @xmath0-nucleus scattering reactions around 25 mev . while the real part of the optical potential becomes less attractive with increasing energy , the imaginary part does not show a too strong energy dependence in the range of energies explored . this behavior can be easily inferred from the @xmath0 self - energies shown at the bottom of fig . [ fig : selftot ] . in this work we have studied the @xmath0 properties in nuclear matter using a chiral unitary approach for the @xmath13 interaction in s - wave that incorporates the medium effects microscopically . pauli blocking acting on the intermediate nucleon states modifies the @xmath13 interaction from that in free space and gives rise to a @xmath0 spectral function that shows two distinctive peaks at very low density . the lower one corresponds to the the position of the new @xmath0 pole , located below the @xmath0 mass from already quite low densities on , which indicates that the @xmath0 feels an attraction in the medium . the higher peak is associated to the in - medium @xmath6 resonance that appears above the free @xmath228 threshold due to the repulsive effect induced by blocking the intermediate nucleon states below the fermi momentum . as density increases , the @xmath0 feels more attraction and the @xmath6 keeps moving to higher energies . when the @xmath0 self - energy is incorporated self - consistently into the scheme , there is a compensation between the attraction felt by the @xmath0 and the repulsive pauli blocking shift . as a result , the resonance appears at a similar location as that in free space . however , its width is much larger and it tends to dissolve with increasing density due to a weaker in - medium @xmath13 interaction which now is determined from dressed @xmath4 propagators which are more spread out over energies . in the present work we also include the dressing of the pions in the intermediate @xmath229 , @xmath192 states to which the @xmath230 system can couple . we use a pion self - energy that includes the coupling to @xmath155 , @xmath157 and @xmath158 excitations modified by the effect of short - range @xmath231 correlations . the fact that now the pions also have a spreading width makes the in - medium @xmath230 interaction even smoother . the @xmath0 feels less attraction and the @xmath6 resonance , which is shifted slighty upwards from its free space position , dissolves even faster with density . our approach gives rise to a @xmath0 self - energy at normal nuclear density which has about half the attraction of that obtained with other theories and approximation schemes and would make the phenomenon of kaon condensation very unlikely . the self - energy obtained is , however , compatible with the existing data on kaonic atoms . this work is partially supported by dgicyt contract numbers pb95 - 1249 and pb96 - 0753 , and by the eec - tmr program under contract no . ct98 - 0169 . kaplan and a.e . nelson , phys . b175 ( 1986 ) 57 r. barth et al . , 78 ( 1997 ) 4007 w. cassing , e.l . bratkovskaya , u. mosel , s. teis and a. sibirtsev , nucl . a614 ( 1997 ) 415 ; e.l . bratkovskaya , w. cassing and u. mosel , nucl . a622 ( 1997 ) 593 g.q . li , c .- h . lee and g.brown , nucl . a625 ( 1997 ) 372 ; _ ibid _ , phys . 79 ( 1997 ) 5214 e. friedman , a. gal , c.j . batty , nucl . a579 ( 1994 ) 518 c .- h . lee , g.e . brown , d.p . min and m. rho , nucl . a385 ( 1995 ) 481 ; c .- h . lee , phys . 275 ( 1996 ) 255 n. kaiser , p.b . siegel and w. weise , nucl . a594 ( 1995 ) 325 ; n. kaiser , t. waas and w. weise , nucl . a612 ( 1997 ) 297 e. oset and a. ramos , nucl . a635 ( 1998 ) 99 v. koch , phys . b337 ( 1994 ) 7 t. waas , n. kaiser and w. weise , phys . b365 ( 1996 ) 12 ; _ ibid . _ b379 ( 1996 ) 34 t. waas and w. weise , nucl . a625 ( 1997 ) 287 m. lutz , phys . lett . b426 ( 1998 ) 12 m. alberg , e.m . henley and l. wilets , ann . ( 1976 ) 43 g. mao , p. papazoglou , s. hofmann , s. schramm , h. stcker and w. greiner , phys . c , in press . nucl - th/9806068 j. schaffner and i.n . mishustin , phys . c53 ( 1996 ) 1416 ; + j. schaffner - bielich , i.n . mishustin and j. bondorf , nucl . a625 ( 1997 ) 325 k. tsushima , k. saito , a.w . thomas and s.v . wright , phys . lett . b429 ( 1998 ) 239 j. gasser and h. leutwyler , nucl . b250 ( 1985 ) 465 a. pich , rep . prog . phys . 58 ( 1995 ) 563 g. ecker , prog . 35 ( 1995 ) 1 v. bernard , n. kaiser and u.g . meissner , int . e4 ( 1995 ) 193 j.a . oller , e. oset and j.r . pelez , phys . 80 ( 1998 ) 2452 d.n . tovee et al . , nucl . b33 ( 1971 ) 493 r.j . nowak et al . , nucl . b139 ( 1978 ) 61 a.d . martin , nucl . b179 ( 1981 ) 33 m. iwasaki et al . 78 ( 1997 ) 3067 j.a . nacher , e. oset , h. toki and a. ramos , submitted to phys . b h. band , t. motoba and j. ofka , int . j. mod . a5 ( 1990 ) 4021 t. nagae et al . 80 ( 1998 ) 1605 c.j . batty et al , phys . lett . 74b ( 1978 ) 27 , _ 87b ( 1979 ) 324 e. oset , p. fernndez de crdoba , l.l . salcedo and r. brockmann , phys . reports 188 ( 1990 ) 79 o. meirav , e. friedman , r.r . johnson , r. olszewski and p. weber , phys . c40 ( 1989 ) 843 a. ramos , e. oset and l.l . salcedo , phys . c50 ( 1994 ) 2314 c.j . batty , nucl . a372 ( 1981 ) 418 s. hirenzaki , y. okumura , h. toki , e. oset and a. ramos , in preparation .

supersymmetry serves as the most promising extension of the standard model . in the short future supersymmetry will be verified experimentally through the lhc experiments . the elegance of the theory and the simplifications that introduces is of great importance however supersymmetry must be broken in our four dimensional world . there exist various mechanisms of supersymmetry breaking . in this article we are interested in dynamical breaking of supersymmetry induced from a toroidal compact dimension . theories with one toroidal compact dimension resemble ( mathematically ) field theories at finite temperature . it is known that supersymmetry is spontaneously broken at finite temperature @xcite , a fact that is closely related to the boundary conditions that fermions and bosons obey in the thermal compact dimension . specifically supersymmetry is spontaneously broken due to the periodicity of bosonic degrees of freedom and anti - periodicity of the fermionic degrees of freedom . field theories at finite temperature are conceptually related to field theories with compact dimensions . thus it is easily understood that the boundary conditions of the fields in the compact dimensions control the breaking of supersymmetry . in general , when studying supersymmetric theories in flat spacetime , the background metric is ordinary minkowski . spacetime topology affect the boundary conditions of the fields that are integrated in the path integral . given a class of metrics , several spacetime topologies are allowed . here we shall focus on a model that has @xmath3 topology underlying the spacetime , @xmath4 refers to a spatial dimension . the specific topology is a homogeneous topology of the flat clifford klein type @xcite . non trivial topology , implies non trivial field configurations , that enter dynamically in the action . this non triviality enters in the action through the boundary conditions and as is well known the boundary conditions are controlled from the topology . the effective potential is a strong order parameter indicating when supersymmetry is broken . the appearance in the effective potential of vacuum terms which have different coefficients for fermions and bosons lead to the fact that the effective potential of the theory has no longer its minimum at zero and thus , supersymmetry is spontaneously broken . this quite general phenomenon can only be avoided in field theories with compact spatial dimensions , if in some way these vacuum terms are cancelled @xcite . indeed the combination of the allowed boundary conditions , as we show later , can save spontaneous supersymmetry breaking at finite volume . particularly this is due to the fact that we can have periodic fermions and anti - periodic bosons . this can not be avoided at finite temperature due to the restricted boundary conditions as we already mentioned , except in cases where fermions have a complex chemical potential @xcite as in pure @xmath5 yang mills theories with adjoint fermions at finite temperature @xcite , namely , @xmath6 normally a question arises while following the above considerations . why should someone care in avoiding spontaneous supersymmetry breaking at finite volume ? this is because when supersymmetry spontaneously breaks ( like at finite temperature ) then supersymmetry ceases to be a controllable symmetry of the theory , since it is always broken and not dynamically , but by definition . one would like to have control on the way that supersymmetry breaks ( especially in our case that the model spacetime we work is 4 dimensional and supersymmetry must be a symmetry and be broken dynamically . this can be avoided in theories with higher dimensions where supersymmetry can hold in the higher dimensional space and break in our world ) . in this paper we shall study a 4 dimensional n=1 supersymmetric model at one loop , in topology @xmath3 . we shall find the allowed field configurations that are determined from the topology and construct in a correct way the lagrangian . the calculation of the effective potential follows , through which we shall find when supersymmetry breaks and when does not . after that we shall add in the lagrangian non holomorphic and hard supersymmetry breaking terms . this terms break supersymmetry hardly . however for some values of the couplings and of the masses , interesting phenomena occur . particularly it is found that in some cases when the volume of the compact space is small supersymmetry is not broken and when the compact radius exceeds a critical value , supersymmetry breaks dynamically . we shall give a cosmological implication of this case which resembles first order cosmological phase transitions in the early universe . also other terms have a curious but worthy of mentioning effect . for these terms another strange phenomenon occurs for specific values of the couplings and masses . in detail for large values of the circumference of the compact dimension , supersymmetry is unbroken in contrast to the previous case and breaks after a critical length , as the compact dimension magnitude decreases . although this is not so interesting from a phenomenological point of view ( in four spacetime dimensions ) , it worths mentioning ( and may have application in the physics of extra dimensions ) . there exist similar works in the literature but with the difference that the symmetry under study is not supersymmetry but a global symmetry ( continuous or discrete ) . at finite temperature in some cases broken @xmath7 symmetries become restored at high temperatures . also unbroken symmetries at small temperatures may break at high temperatures ( a phenomenon known as inverse symmetry breaking ) . the last may have cosmological implications . in our case the same occurs but with supersymmetry in place of the symmetry and for a compact dimension playing the temperature s role ( also note that roughly the high temperature limit is closely related to the small length limit through the transformation @xmath8 . actually through the last transformation we can relate the two limits where this is possible ) . one interesting feature of this resemblance is the similarity of the terms in the lagrangian that trigger these phenomena in both cases . also these terms appear in the new inflationary models and help in the procedure of reheating the universe after inflation . we shall present and describe everything in detail in the forthcoming sections . also let us mention that our calculations will be in 1-loop level and within the perturbative limits with @xmath9 , where @xmath10 is the largest mass scale in the theory and @xmath11 is the circumference of the compact dimension . also within the four dimensional setup we use , renormalizability of masses and couplings is ensured when @xmath12 . in section 1 we review the mathematical setup needed for field theories with non trivial topology . the resemblance with extra dimensional theories is pointed out . in section 2 we describe the n=1 supersymmetric model we shall use and calculate the effective potential in the case of the compact dimension having infinite length and after that at finite volume . in section 3 we add several supersymmetry breaking terms and study in detail their effect on the vacuum energy of the model . in section 4 we review the continuous symmetry restoration , symmetry non - restoration and inverse symmetry breaking at finite temperature and point out the resemblance of these with our case ( a resemblance that stems from the interactions of the scalar sector ) . in section 5 we present a cosmological application of one of our results and in section 6 a short discussion with the conclusions follow . the existence of non trivial field configurations ( in terms of boundary conditions ) due to non trivial topology ( twisted fields ) , was first pointed out by isham @xcite and then adopted by other authors @xcite . in the spacetime of our case , the topological properties of @xmath13 are classified by the first stieffel class @xmath14 which is isomorphic to the singular ( simplicial ) cohomology group @xmath15 because of the triviality of the @xmath16 sheaf . it is known that @xmath17 classifies the twisting of a bundle . specifically , it describes and classifies the orientability of a bundle globally . in our case , the classification group is @xmath18 and , we have two locally equivalent bundles , which are however different globally _ i.e. _ cylinder - like and moebius strip - like . the mathematical lying behind , is to find the sections that correspond to these two bundles , classified completely by @xmath19 @xcite . the sections we shall consider are real scalar fields and majorana spinor fields which carry a topological number called moebiosity ( twist ) , which distinguishes between twisted and untwisted fields . the twisted fields obey anti - periodic boundary conditions , while untwisted fields periodic in the compactified dimension . usually ( inspired by field theory at finite temperature ) one takes scalar fields to obey periodic and fermion fields anti - periodic boundary conditions , disregarding all other configurations that may arise from non trivial topology . we shall consider all these configurations . let @xmath20 , @xmath21 and @xmath22 , @xmath23 denote the untwisted and twisted scalar and twisted and untwisted spinor fields respectively . the boundary conditions in the @xmath24 dimension are , @xmath25 and @xmath26 for scalar fields and @xmath27 and @xmath28 for fermion fields , where @xmath29 stands for the remaining two spatial and one time dimension which are not affected by the boundary conditions . spinors ( both dirac and majorana ) , still remain grassmann quantities . we assign the untwisted fields twist @xmath30 ( the trivial element of @xmath31 and the twisted fields twist @xmath32 ( the non trivial element of @xmath18 ) . recall that @xmath33 ( @xmath34 ) , @xmath35 ( @xmath36 ) , @xmath37 ( @xmath38 ) . we require the lagrangian to be scalar under @xmath19 thus to have @xmath30 moebiosity . the topological charges flowing at the interaction vertices must sum to @xmath30 under @xmath39 . for supersymmetric models , supersymmetry transformations impose some restrictions on the twist assignments of the superfield component fields @xcite . now which fields can acquire vacuum expectation value ? grassmann fields can not acquire vacuum expectation value ( vev ) since we require the vacuum value to be a scalar representation of the lorentz group . thus , the question is focused on the two scalars . the twisted scalar can not acquire non zero vev @xcite , consequently , only untwisted scalars are allowed to develop vev s . in the literature , twisted fields have frequently been used , for example in the scherk - schwarz mechanism @xcite , where the harmonic expansion of the fields is of the form : @xmath40 the @xmath10 parameter incorporates the twist mentioned above . this treatment is closely related to automorphic field theory @xcite in more than 4 dimensions ( which is an alternative to the one used by us ) . the scherk - schwarz mechanism is a well known mechanism that generates supersymmetry breaking to our 4 dimensional world after dimensional reduction and is frequently used for compactifications in extra dimensional models . concerning the automorphic field theory , let us quote here a different approach to the above . due to the compact dimension we can use generic boundary conditions for bosons and fermions in the compact dimension . these are,@xmath41with , @xmath42 , @xmath43 , @xmath44 . the values @xmath45 correspond to periodic and antiperiodic bosons respectively while @xmath46 corresponds to periodic and anti - periodic fermions ( for details see @xcite ) . the model we shall present is described by the global @xmath47 , @xmath48 supersymmetric lagrangian , @xmath49_{d}+[\phi ^{+}\phi ] _ { d}+[\frac{m_{1}}{2}\phi ^{2}+\frac{g_{1}}{6}\phi ^{3}+\frac{m}{2}\phi _ { 1}^{2}+g\phi \phi _ { 1}^{2}]_{f}+{\rm{h.c } } , \label{lagra}\]]where @xmath50 , @xmath51 are chiral superfields and the superpotential from which the interaction part of the lagrangian arises is @xmath52_{f}$ ] . in the above , @xmath53is a left chiral superfield . it contains the untwisted scalar field components and the untwisted weyl fermion . although the untwisted scalar is complex , we shall use the real components which will be the representatives of the sections of the trivial bundle classified by @xmath54 . moreover,@xmath55is another left chiral superfield containing the twisted scalar field and the twisted weyl fermion . writing down ( [ lagra ] ) in component form , we get ( writing weyl fermions in the majorana representation):@xmath56notice that moebiosity is conserved at all interaction vertices _ i.e. _ equals @xmath30 . the moebiosity of @xmath20 and @xmath57 is @xmath30 while for @xmath58 and @xmath59 is @xmath60 . using the @xmath18 cyclic group properties we see that the lagrangian ( [ test ] ) has moebiosity @xmath30.the complex field @xmath20 can be written in terms of real components as @xmath61 , where @xmath62 ( @xmath63 is the classical value ) . thus , @xmath64 and @xmath65 are real untwisted field configurations belonging to the trivial element of @xmath66 and satisfying periodic boundary conditions in the compactified dimension . the twisted scalar field can be written as @xmath67 , since , this field , being a member of the non trivial element of @xmath68 can not acquire a vev . notice we gave a vev for an untwisted boson , this is useful in order to find the minimum of the effective potential minimizing it in terms of @xmath69 . the masses of the two majorana fermion fields and the four bosonic fields at tree order are calculated to be:@xmath70 in ( [ mass ] ) @xmath71 , @xmath72 are the masses of the untwisted bosons ( @xmath73 and @xmath74 respectively ) , @xmath75 , @xmath76 are the masses of the twisted bosons ( @xmath77 and @xmath78 ) and , finally , @xmath79 , @xmath80 are the untwisted majorana fermion and twisted majorana fermion masses respectively ( @xmath81 and @xmath82 ) . the general tree level result for theories with rigid supersymmetry in terms of chiral superfields is satisfied ( see @xcite ) _ i.e. _ : @xmath83also , the following relations hold true:@xmath84since twisted scalars can not acquire vacuum expectation value , supersymmetry is not spontaneously broken at tree level , like in the o raifeartaigh models . indeed the auxiliary field equations,@xmath85imply that @xmath86 and @xmath87 and consequently @xmath88 , thus , at tree level , no spontaneous supersymmetry breaking occurs . we now proceed by assuming that the topology is changed to @xmath3 , while the local geometry remains minkowski . the metric reads : @xmath90 with @xmath91 and @xmath92 with the points@xmath93and @xmath94 periodically identified . the boundary conditions for the fields are:@xmath95we wick rotate the time direction @xmath96 thus giving the background metric the euclidean signature @xcite . the twisted fermions and twisted bosons will be summed over odd matsubara frequencies , while the untwisted fermions and untwisted scalars will be summed over even matsubara frequencies @xcite . adopting the @xmath97 renormalization scheme @xcite the euclidean effective potential at one loop level reads:@xmath98 \\ & -2\ln [ k^{2}+\frac{4\pi^{2 } n^{2}}{l^{2}}+m_{f_{1}}^{2}]+\ln [ k^{2}+\frac{% 4\pi^{2 } n^{2}}{l^{2}}+m_{b_{2}}^{2 } ] \notag \\ & -2\ln [ k^{2}+\frac{\pi^{2 } ( 2n+1)^2}{l^{2}}+m_{f_{2}}^{2}]+\ln [ k^{2}+\frac{% \pi^{2 } ( 2n+1)^2}{l^{2}}+m_{t_{1}}^{2 } ] \notag \\ & + \ln [ k^{2}+\frac{\pi^{2 } ( 2n+1)^2}{l^{2}}+m_{t_{2}}^{2}]{\bigg ) . } \notag\end{aligned}\]]@xmath99includes the tree and the one loop corrections for infinite length,@xmath100-% \frac{3}{2 } ) \\ & + m_{b_{2}}^{4}(\ln [ \frac{m_{b_{2}}^{2}}{\mu ^{2}}]-\frac{3}{2}% ) + m_{t_{1}}^{4}(\ln [ \frac{m_{t_{1}}^{2}}{\mu ^{2}}]-\frac{3}{2}% ) + m_{t_{2}}^{4}(\ln [ \frac{m_{t_{2}}^{2}}{\mu ^{2}}]-\frac{3}{2 } ) \notag \\ & -2m_{f_{1}}^{4}(\ln [ \frac{m_{f_{1}}^{2}}{\mu ^{2}}]-\frac{3}{2}% ) -2m_{f_{2}}^{4}(\ln [ \frac{m_{f_{2}}^{2}}{\mu ^{2}}]-\frac{3}{2}){\bigg ) } , \notag\end{aligned}\]]and @xmath101 is the renormalization scale , being of the order of the largest mass @xcite . furthermore , we shall assume that @xmath102 which is required for the validity of perturbation theory @xcite . it is well known that when one considers only twisted fermions and untwisted bosons in @xmath3 ( like in thermal field theories ) , vacuum contributions @xmath103@xmath104 do not cancel and supersymmetry is spontaneously broken . the non - cancellation occurs because bosons and fermions satisfy different boundary conditions . in our model the field content is such that cancellation of vacuum contributions is being enforced , after having included all topologically inequivalent allowed field configurations . this situation is similar to finite temperature calculations . the question if supersymmetry is broken or not requires to check the zero modes of the vacuum state @xcite . it is easy to see why in conventional finite temperature field theories and their conceptional analogues @xmath3 topological field theories supersymmetry is broken . the vacuum state , in the wess zumino in @xmath3 , does contain one bosonic zero mode . in our case this does not occur because we have equal vacuum zero modes ( twisted spinors , do not have a zero mode ) . consequently , in our model , we expect that supersymmetry will not be spontaneously broken @xcite ( for a detailed discussion we recommend the paper of fujikawa @xcite ) . the leading order contribution to the one loop effective potential for small @xmath11 values is given by @xcite : @xmath105)(2m_{f_{1}}^{4}-m_{b_{1}}^{4}-m_{b_{2}}^{4})}{% 1024\pi ^{4}}+\frac{(\gamma + \ln [ \frac{2}{\pi } % ] ) ( 2m_{f_{2}}^{4}-m_{t_{1}}^{4}-m_{t_{2}}^{4})}{64\pi ^{4}}}\notag \\ & { + \frac{(2m_{f_{1}}^{3}-m_{b_{1}}^{3}-m_{b_{2}}^{3})}{384l\pi ^{3}}-\frac{% ( 2m_{f_{1}}^{2}-m_{b_{1}}^{2}-m_{b_{2}}^{2})}{768\pi ^{2}l^{2}}+\frac{% ( 2m_{f_{2}}^{2}-m_{t_{1}}^{2}-m_{t_{2}}^{2})}{384\pi ^{2}l^{2}}}\notag\\ & { + \frac{2m_{f_{1}}^{4}\ln [ lm_{f_{1}}]-m_{b_{2}}^{4}\ln [ lm_{b_{2}}]-m_{b_{1}}^{4}\ln [ lm_{b_{1}}]}{1024\pi ^{4}}}\notag \\ & { + \frac{2m_{f_{2}}^{4}\ln [ lm_{f_{2}}]-m_{t_{2}}^{4}\ln [ lm_{t_{2}}]-m_{t_{1}}^{4}\ln [ lm_{t_{1}}]}{64\pi ^{4}}}\notag \\ & { -\frac{(2m_{f_{1}}^{4}\ln [ \frac{m_{f_{1}}^{2}}{\mu ^{2}}]-m_{b_{2}}^{4}\ln [ % \frac{m_{b_{2}}^{2}}{\mu ^{2}}]-m_{b_{1}}^{4}\ln [ \frac{m_{b_{1}}^{2}}{\mu ^{2}}])}{64\pi ^{2}}}\notag\\ & { -\frac{(2m_{f_{2}}^{4}\ln [ \frac{m_{f_{2}}^{2}}{\mu ^{2}}]-m_{t_{2}}^{4}\ln [ % \frac{m_{t_{2}}^{2}}{\mu ^{2}}]-m_{t_{1}}^{4}\ln [ \frac{m_{t_{1}}^{2}}{\mu ^{2}}])}{64\pi ^{2}}.\notag}\end{aligned}\ ] ] since relation ( [ sp ] ) holds , the terms proportional to @xmath106 cancel @xcite . also , the minimum of the potential vanishes at @xmath1070 and supersymmetry is preserved . indeed , expanding ( [ pot ] ) for small @xmath63 we get : @xmath108 in figure [ susyzero ] we plot the effective potential for the upper perturbative limit @xmath109 . the other numerical values are chosen to be : @xmath110200 , @xmath1117000 , @xmath1120.001 , @xmath1130.09 , @xmath1147000 . let us now introduce into the lagrangian ( [ test ] ) various hard supersymmetry breaking terms of the form @xmath115 and @xmath116 , where @xmath117 and @xmath118 are dimensionless couplings ( and recall @xmath119 ) . this terms , being non holomorphic and hard , break supersymmetry explicitly ( also these terms have moebiosity zero ) . indeed the addition of such terms re - introduces quadratic divergences in the theory , namely , @xmath120 with @xmath121 a relevant upper cut - off of the theory and @xmath122 , @xmath123 boson and fermion couplings . since @xmath124 develops a vev , the fields coupled to it will acquire an additional mass of the form @xmath125 and @xmath126 . we can add various combinations of the allowed terms . there exist two class of phenomena occurring , depending on the supersymmetry breaking terms we use . the first type and , from a phenomenological point of view more interesting , resembles the first order phase transition picture in thermal field theory . actually , regardless that there exist hard supersymmetry breaking terms , supersymmetry is unbroken for small values of the length of the compact dimension @xmath11 . the minimum of the potential is zero at @xmath88 , @xmath127 . as the length of the compact dimension increases , a second non supersymmetric minimum is created after a critical length is reached . then phenomena may occur that can be described by the theory of metastable vacua ( the reader may find useful the papers @xcite where similar issues are discussed in @xmath2 but for a non supersymmetric @xmath128 theory . again non trivial topology induces similar behavior of the effective potential ) . the second type of phenomena describes a theory that supersymmetry is unbroken when the length of the compact dimension is large and as the radius decreases , supersymmetry breaks spontaneously after a critical length . thus supersymmetry is broken only for small lengths . although this is rather curious for a four dimensional model , we shall present it because it may have application to extra dimensional models . we introduce into the lagrangian ( [ test ] ) an interaction term between the two untwisted scalar fields , of the form @xmath130 . since @xmath119 , the scalar field @xmath65 will acquire an additional mass term of the form @xmath125 . this way , the masses of the fields now become : @xmath131 as expected , supersymmetry is now broken and relation ( [ str ] ) becomes , @xmath132one can see that the supersymmetric minimum at @xmath88 is still preserved . indeed , @xmath133 can be written as:@xmath134we can see that in the continuum limit ( infinite @xmath11 ) , the supersymmetric vacuum becomes metastable and a second non supersymmetric vacuum appears . including finite size corrections , we see that for small @xmath135 the effective potential has a unique supersymmetric minimum at @xmath88 . as @xmath135 increases , a second minimum develops , which becomes supersymmetric at the critical value @xmath136 . when @xmath137 the second minimum is non supersymmetric and becomes energetically more preferable than the supersymmetric one @xcite . this said behavior of the potential is valid whenever @xmath138 and for @xmath139 . using the same numerical values as before , we plot the effective potential for @xmath1400.5 , first in the continuum limit ( figure [ continuumnonsusy ] ) , and then including @xmath135 dependent corrections ( figure [ nonsusy ] ) . let us discuss the above results . @xmath141 , @xmath142 are couplings among the untwisted superfields , @xmath141 corresponding to the supersymmetry breaking term . if the @xmath141 interaction is stronger than @xmath142 and if the mass ( @xmath143 ) of the twisted superfield is larger than the untwisted one ( @xmath144 ) , then the following phenomenon occurs . for small length @xmath135 of the compact dimension , supersymmetry is not broken ( figure [ nonsusy ] ) . as @xmath135 grows larger , a second minimum appears which is not supersymmetric ( @xmath137 ) . there exists a small barrier separating the two minima ( figure [ nonsusy ] ) , and there exists the possibility of quantum barrier penetration between them . this resembles the first order phase transition picture of thermal field theories . upon closer examination , we can see that in the continuum limit , the supersymmetric vacuum becomes metastable and a second non supersymmetric vacuum appears . including finite size corrections , we see that for small @xmath135 the effective potential has a unique supersymmetric minimum at @xmath1450 . as @xmath135 increases , a second minimum develops , which becomes supersymmetric at the critical value @xmath136 . when @xmath137 the second minimum breaks supersymmetry and becomes energetically more preferable than the supersymmetric one @xcite . this said behavior of the potential is always valid whenever @xmath138 and for @xmath139 . using the same numerical values as before , we plot the effective potential for @xmath1400.5 , first in the continuum limit ( figure [ continuumnonsusy ] ) , and then including @xmath135 dependent corrections ( figure [ nonsusy ] ) . let us try something different now . we add an interaction among a twisted boson and the untwisted boson that acquires vev , namely @xmath147 . since @xmath119 the twisted boson @xmath148 will have additional contribution to it s tree order mass . the masses now read , @xmath149 as expected @xmath150 , since supersymmetry is hard broken . an interesting phenomenon occurs for this term and for a class of other terms as we shall see . in detail , when the length of the compact dimension is small , supersymmetry is broken and becomes restored when the radius increases ( and overcomes a critical length @xmath151 ) . this is strange and rather counterintuitive to what would be expected from a phenomenologically correct four dimensional theory . however we describe it since it might be useful to extra dimensional physics . also , as we shall see in the next section , the whole behavior resembles the inverse symmetry breaking of continuous symmetries at finite temperature . let us call it inverse supersymmetry breaking for brevity . this said behavior can appear when @xmath152 is of the order of @xmath153 or for values smaller , that is @xmath154 when only the term @xmath146 appears in the lagrangian ( recall that in the previous subsection , the metastable vacua phenomena of the previous subsection occurred when @xmath155 ) . this whole phenomenon is well seen in figure [ inverse1 ] . we used the following numerical values , @xmath110200 , @xmath1117000 , @xmath1120.001 , @xmath1130.09 , @xmath1147000 and @xmath156 . as can be seen in figure [ inverse1 ] the phenomenon looks like a second order phase transition with the length of the compact dimension playing the role of the temperature . no barrier appears between the vacua at @xmath88 and at @xmath157 . the study was limited to perturbation preserving values of @xmath11 . as we see for large @xmath11 ( @xmath158 ) supersymmetry is unbroken and start s to break at @xmath159 . as the length decreases , the breaking is more profound . the two non supersymmetric vacua are not equivalent . [ table1 ] [ cols="^,^",options="header " , ] + we tried to find how @xmath151 changes under a change of @xmath152 . in the table we present the values of @xmath152 and the corresponding values of @xmath151 , and in figure [ plotl3lc ] we plot the dependence . in figure [ fitl3lc ] we fit the curve with a continuous function . the dots are the values that appear in the table , while the continuous line corresponds to the function @xmath160 . thus the dependence of @xmath152 as a function of @xmath151 is roughly , @xmath161 and @xmath151 ] ] we shall not present here the detailed study of all the allowed terms ( this will be done in @xcite ) . we shall discuss only the main results . a term of the form @xmath163 always breaks supersymmetry and non of the previous phenomena occurs . as we shall see in the next section this is similar to symmetry non restoration where a continuous symmetry is broken and never get s restored . the combined addition of terms @xmath164 ( that is interaction of the untwisted scalar @xmath165 with a twisted fermion and twisted boson ) causes inverse supersymmetry breaking as previous . the conditions that must hold in order this occurs are the same as before ( @xmath166 ) and in addition @xmath167 . for this condition the behavior of supersymmetry breaking is well described from figure [ inverse1 ] . also for @xmath168 the @xmath169 dependence is similar to that of figure [ fitl3lc ] . particularly the @xmath169 dependence in this case , is described by , @xmath170 which is the same as before . this dependence is the same for all the cases we studied . thus this motivates us to think that there is a universal behavior of @xmath152 as a function of @xmath151 . of course all effects we described in this section appear in one loop level and within perturbative limits . thus there exist the danger that all these effects are an artifact of perturbation theory . however the theory is at some level supersymmetric and one loop corrections may be adequate enough @xcite . a detailed study should include higher loop corrections . in the next section similar problems - considerations are encountered and discussed . in this section we review some conceptually similar phenomena to the above . the difference is that symmetries are studied at finite temperature and the symmetry is not supersymmetry but a continuous global @xmath171 or a discrete @xmath172 . as we shall see symmetry non - restoration and inverse symmetry breaking phenomena occur naturally when similar terms to @xmath173 , @xmath174 appear in the lagrangian . symmetry non - restoration means that a symmetry broken at @xmath175 never gets restored at high temperatures . inverse symmetry breaking means that an unbroken symmetry at @xmath175 may be spontaneously broken at high temperature . these phenomena occur in field theories when cross interactions are included among the scalar fields similar to the bosonic hard supersymmetry breaking term @xmath129 . similar to this term scalar interactions and also yukawa terms like the ones of the previous sections are frequently used in the theory of reheating after inflation . actually this similarities motivated us to use such terms in order to see what their effect would be on supersymmetry breaking . we shall discuss these in the end of this section . first let us describe the inverse symmetry breaking phenomenon . consider a theory with real scalar fields @xmath176 and @xmath177 described by the @xmath171 globally symmetric lagrangian , @xmath178 where @xmath176 and @xmath177 be real scalars with @xmath179 and @xmath180 components . in the above lagrangian one of the global @xmath181 symmetries may break at high temperature if the @xmath182 coupling takes negative values . thus one of the two scalar fields @xmath176 or @xmath177 may acquire a non zero vacuum expectation value . thus at high temperature and for certain values of the parameters , the initial @xmath171 is broken to @xmath183 . this was called inverse symmetry breaking and was first point out by weinberg @xcite and extensively studied by many authors @xcite . in the case @xmath184 and @xmath185 the initial @xmath171 symmetry reduces to a @xmath172 symmetry and similar arguments hold , in connection to the breaking of one of the two discrete symmetries ( for details see @xcite ) symmetry non - restoration was used in @xcite to solve the monopole problem in the @xmath186 gut . monopoles are usually produced during phase transitions at temperatures of the order @xmath187gev . as was proposed in @xcite , the symmetric phase of @xmath186 was never realized , no matter how high the temperature becomes . in that paper the interaction term @xmath188 appearing in the scalar kibble - higgs sector is responsible for the non - restoration of the @xmath186 symmetry . actually the scalar interaction of @xmath189 and @xmath190 gives negative contributions to the thermal masses and one of those becomes negative . once this happens the corresponding higgs field maintains a vacuum expectation value for high temperatures and the symmetry is never restored . this phenomenon occurs for certain values of the parameters ( see @xcite ) . however these results are very sensitive and are altered when someone includes two loop calculations @xcite . in the same spirit there are arguments based on large n calculations which can show that symmetry non - restoration is an artifact of perturbation theory . for an interesting discussion on this see @xcite and references therein . a cross term of the form @xmath191 is used in non relativistic models in condensed matter physics , for example in the coupled two field bose gases . however the effect of this does not break any of the initial symmetry patterns @xcite . in conclusion the intuitive approach to all phenomena at finite temperature consists of the statement that symmetries broken at small temperatures become restored at high temperature ( in the same class belong finite volume theories ) . many field theory models exist that belong to this class . counter intuitive phenomena occur in field theories with rich scalar sector . especially if the multi - scalar sectors interact weakly with negative couplings then the phenomena known as inverse symmetry breaking or symmetry non - restoration occur . this happens at high temperature and refers to the spontaneous breaking of a symmetry at high temperature . usually the symmetry is a global @xmath171 or for the case of symmetry non - restoration a continuous , like @xmath186 . although inverse symmetry breaking is counterintuitive , nature has provided us with examples that systems are more symmetric at low temperatures than at high temperature . for example the rochelle salt which at low temperatures is orthorhombic and after a critical temperature becomes monoclinic . similar phenomena occur in liquid crystals ( for more examples see @xcite and references therein ) . the process of reheating after inflation is one of the most important features of the new inflationary scenario @xcite . the process of reheating is necessary in order that the inflating vacuum like state of the universe transforms to a hot friedmann universe state . during the reheating process a massive scalar field gives it s vacuum energy to lighter fermions and bosons . the lagrangian governing this process is , @xmath192 the role of the inflaton fields is played from the scalar field @xmath193 . the inflaton field decays to the particles @xmath194 and @xmath165 due to the interaction terms @xmath195 and @xmath196 . note that we used similar terms in order to break supersymmetry hard and all the effects we seen in the previous section are due to these interaction terms . also , in order reheating takes place , the condition @xmath197 must hold ( remember that similar conditions hold in the supersymmetric model we studied previously . there @xmath198 was the tree order mass of the untwisted scalar field and @xmath10 the tree order mass of the twisted scalar . one of the conditions we used is that the untwisted sector has greater mass than the twisted sector , namely @xmath199 . also the untwisted fermion has mass m ) . so with the interactions @xmath200 and @xmath196 the initially concentrated energy ( during inflation ) to the field @xmath193 is transferred to particle creation through the processes @xmath201 and @xmath202 , and the universe thermalizes @xcite . the cross interaction terms of the form @xmath203 are used in some versions of hybrid inflation . let us briefly mention another application of cross interactions between scalar fields . these terms are used to the thermal inflation scenario . this is a modified version of the new inflation and old inflation scenario @xcite . in the thermal inflation scenario , a lagrangian that contains a term @xmath203 this term is necessary in order a phase transition takes place . actually there is bump in the effective potential that is solely created from this interaction term and the phase transition is strongly first order . the universe supercools in the false vacuum and after a critical temperature tunnels to the true vacuum through bubble nucleation . at this point thermal inflation ends ( for details see @xcite ) . before closing this section we conclude that cross terms between scalar fields and yukawa interactions are frequently used in particle models and these terms make the effective potential locally unstable thus triggering first order phase transitions , reheating and other processes . also at high temperatures and for specific values of the parameters , these terms may cause inverse symmetry breaking or symmetry non restoration . our study involved these terms but the phenomena studied where related to supersymmetry breaking and inverse supersymmetry breaking when the space has a compact dimension . so the resemblance between the two setups is quite clear . we shall discuss on this later on . in this section a brief qualitative application ( although fictitious ) of one of the above results is presented . consider a toy universe that has just come out of it s strong gravity period and it s particle content ( matter ) is described by ( [ test ] ) with the addition of the hard supersymmetry breaking term @xmath204 . the back - reaction of gravity on field theory is considered small ( _ i.e._field theory calculations made in the previous part considering flat background , are consistent and thus the metric fluctuations are negligible ) . this toy universe s expansion is described by : @xmath205 a homogeneous clifford - klein metric ( known as bianchi i cosmological model ) , with @xmath206 , @xmath207 , @xmath208 , as in ( [ flat ] ) . in ( [ metric ] ) , @xmath209 , @xmath210 , @xmath211 , describe the scale factors of the two infinite and of the compact dimension respectively . also we assume , @xmath212 with @xmath213 . figure [ nonsusy ] motivates us to think as follows : at small lengths of the compact dimension the toy universe s ground state is the supersymmetric vacuum although we had broken supersymmetry using a hard term , something usually unexpected . as the circumference of the compact dimension grows , the toy universe acknowledges the presence of the other true vacuum ( in terms of it s quantum one loop effective potential ) and at some point ( bubbles of the new vacuum create within the false vacuum ) quantum penetrates to the other vacuum , the non supersymmetric one . therefore , at small lengths of the compact dimension , supersymmetry was unbroken and as the radius grows , supersymmetry breaks . it seems that using a compact dimension in the present model , supersymmetry breaks dynamically after some critical radius of the compact dimension , although supersymmetry is expected to be broken for all lengths ( this would exactly be the effect of a hard supersymmetry breaking term ) . let us now do some toy cosmology on this toy universe . @xmath214 is the minimum of the effective potential at the origin ( note @xmath127 ) , and @xmath215 the minimum after quantum barrier penetration ( the non supersymmetric vacuum ) . we assume this toy universe has a cosmological constant which is chosen to be @xmath216 ( a choice which shall be explained below ) . note that @xmath217 . the friedmann equation describing it s evolution is : @xmath218 referred to the @xmath219 dimension ( we omit the analysis on the other dimension and to the compact one since it is similar ( [ cond ] ) . for details see @xcite ) . in the early post quantum gravity period , this toy universe is at the @xmath214 vacuum state , the energy density is @xmath220 . the friedmann equation reads : @xmath221 and without getting into much detail ( see @xcite ) , an inflationary solution ( corresponding to a flat universe ) follows in all space dimensions , being of the form , @xmath222 with @xmath223 . note that the rate of the expansion is the same for all dimensions . during the inflation period of this toy universe , its quantum vacuum state is the supersymmetric vacuum ( false vacuum ) , until for some length quantum tunnelling occurs ( due to one loop quantum effects ) , and the new vacuum state is @xmath215 , the new minimum of the effective action . during the quantum vacuum penetration , the energy release ( something like latent heat ) @xcite is of the order @xmath224 which thermalizes the matter content at a temperature @xmath225 , with @xmath226 @xmath227 and @xmath225 characterizing the phase transition point . after thermalization , the energy density is @xmath228 and the friedmann equation reads : @xmath229 ( we fixed @xmath230 in order to cancel the value of @xmath215 ) . so after vacuum penetration the toy universe follows a radiation dominated expansion ( note that the maximum temperature ever reached was the thermalization temperature @xmath225 @xcite ) . note that the above picture has many similarities with the strongly supercooled first order phase transitions of the early universe ( old inflationary scenario ) . let us point out its main features . start with a toy universe filled with fermions and bosons interacting in a non supersymmetric way ( due to explicit hard breaking ) . the toy universe is at a supersymmetric vacuum ( unexpectedly ) when it s magnitude ( specifically the compact dimension magnitude ) is small , but as it evolves spatially ( inflation in our setup ) quantum penetrates to a non supersymmetric vacuum , which is energetically preferable . so at the early toy universe s epoch , supersymmetry was not broken ( at least the vacuum quantum state did not realize broken susy ) , although the matter content of it , interacts in a non supersymmetric way , but supersymmetry dynamically breaks ( through quantum tunnelling ) @xcite when the toy universe evolves at larger sizes . let us note here that in order this toy universe is realistic , one must deal with defects ( monopoles , domain walls ) and with the cosmological experimental observations that do not suggest non trivial topology in the spatial dimensions . maybe domain walls may be avoided in first order phase transitions but if we want to include guts in this universe we can not avoid defects ( only if the temperature after the quantum penetration is smaller compared to the temperature that the defects are created , then defects maybe be avoided ) . even if one deals with defects , the non trivial topology problem remains , so the magnitude of the compact dimension must be larger than the particle horizon ( which can be achieved during inflation ) . in this paper we studied a simple supersymmetric model in @xmath2 spacetime topology . we discussed how topology can affect the boundary conditions of the fields and we seen that in @xmath2 bosons and fermions can have periodic and anti - periodic boundary conditions along the compact dimension . also we discussed how supersymmetry is broken spontaneously and how this can be avoided in terms of the boundary conditions that the fields obey . we confirmed these by calculating the effective potential of the theory . * when an interaction among the two untwisted scalars is added , a term of the form @xmath129 , supersymmetry remains unbroken for small values of the radius of the compact dimension , and as the length increases , breaks after a critical length . this occurs when @xmath155 . this phenomenon resembles first order phase transitions of finite temperature field theories . also we applied this to a toy cosmological model , which described the evolution of a universe with an initial cosmological constant and filled with the aforementioned fields . * the addition of an interaction among the scalar of the untwisted sector that develops a vev and the twisted scalars or the twisted fermions results in a very peculiar phenomenon . particularly when certain conditions hold ( similar to the aforementioned ) supersymmetry is broken for small lengths and as the radius increases , becomes restored after a critical length . this resembles conceptually second order phase transitions . the last is similar to inverse symmetry breaking phenomena at finite temperature , that appear in field theories with rich scalar sector . in that case a symmetry unbroken at low temperatures may break at high temperature . in our case at small lengths supersymmetry breaks while remains unbroken for large radius values . we called this inverse supersymmetry breaking for brevity . the terms in the lagrangian that trigger inverse symmetry breaking are the same that trigger inverse supersymmetry breaking . also these terms appear in the theory of reheating and in the thermal inflation . in the study we realized that there is a universality in the @xmath169 dependence . we shall present these in detail in @xcite . finally let us discuss the physical significance of inverse supersymmetry breaking . this phenomenon is not so appealing to a four dimensional theory . what would be expected is that supersymmetry should be unbroken for small values of the radius of the compact dimension and breaks dynamically at large distances . what happens here is the converse . for large values of the radius , supersymmetry is unbroken and breaks dynamically for small values of the radius . however this would be interesting for a five dimensional model . imagine a theory where supersymmetry is unbroken for large radius of the compact dimension and breaks dynamically for small values of the compact dimension . it is an interesting task to find what this mechanism ( which is basically coupling interplay between interactions ) has to say for the radius stabilization mechanism of extra dimensional models . k. kirsten , e. elizalde , phys . b365 , 72 ( 1996 ) ; k. kirsten , j. phys . a26 , 2421 ( 1996 ) ; e. elizalde , k. kirsten , yu . kubyshin , z. phys . c70 , 159 ( 1996 ) ; e. elizalde , k. kirsten , j. math . 35 , 1260 ( 1994 ) yu . p. goncharov , a. a. bytsenko , phys . b163 , 155 ( 1985 ) ; phys . b168 , 239 ( 1986 ) ; phys . b169 , 171 ( 1986 ) ; phys . b160 , 385 ( 1985 ) ; class . grav.8:l211 , 1991 ; class . grav . 8:2269 , 1991 ; class.quant . grav . 4:555 , 1987 ; nucl . b271 , 726 ( 1986 ) e. elizalde , a romeo , rev . phys . 1 , 113 ( 1989 ) ; e. elizalde , j. phys . a39 , 6299 , 2006 ; e. elizalde , `` ten physical applications of spectral zeta functions '' , springer ( 1995 ) ; e. elizalde , j. math . phys . 35,6100 ( 1994 ) ya . b. zeldovich , a. a. starobinsky , sov . 10 , 135 ( 1984 ) ; a. a. starobinsky , phys . 91b , 99 ( 1980 ) ; s. r. coleman , phys . d15 , 2929 ( 1977 ) ; c. g. callan , s. r. coleman , phys . d16 , 1762 ( 1977 ) ; l. van . hove , phys . rep . 137 , 11 ( 1988 ) , nucl . b207 , 15 ( 1982 ) ; ashok das , finite temperature field theory , world scientific 1997 ; d. bailin and a. love , supersymmetric gauge field theory and string theory , institute of physics publishing 2003

the interaction among oscillators in a system not only leads them into a cooperative dynamics but also often quenches their oscillations . there are two dynamically different oscillation quenching phenomena which are termed as amplitude death ( ad ) and oscillation death ( od ) @xcite . in ad , the amplitude of oscillation quenches to zero , whereas od is caused by quenching in the frequency of oscillation @xcite . it is also defined that ad occurs via the stabilization of a homogeneous steady state ( hss ) while od occurs via the stabilization of an inhomogeneous steady state @xcite . the mechanisms underlying these two quenching phenomena have been identified recently and the results show that the parametric mismatch @xcite , dynamic @xcite , time delay @xcite and nonlinear couplings @xcite underly ad , while od occurs mainly because of the symmetry breaking coupling in the system @xcite . recently , diverse routes of transition from the ad to od have also been reported @xcite . experimental and theoretical studies show definitive evidence of oscillation quenching in realistic systems ranging from biological @xcite , chemical @xcite , electronic @xcite and laser @xcite systems to climate @xcite systems . such oscillation quenching in many cases leads to undesirable situations . in the interaction between neuronal dynamics and brain metabolism , the decrease in cerebral metabolic rate , coupled with the stabilizing properties of atp - gated potassium channels , leads to a burst suppression in the eeg pattern which symbolizes inactivated brain @xcite . this type of suppression results in hypothermia , coma and ohtahara syndrome , a type of early infantile encephalopathy and is also observed during deep levels of anesthesia . the suppression of normal sinus rhythm of pacemaker cells causes cardiac arrest @xcite . owing to the fatal consequences due to oscillation suppression , interesting efforts have been undertaken to retrieve / resurrect oscillations of the system @xcite . in @xcite , the oscillation death in diffusively coupled oscillators has been found to be eliminated through a spatial disorder in the form of parametric mismatch and in ref . @xcite processing delay is used to revoke oscillations successfully in delay coupled systems . regarding the above mentioned issues , in this article , we demonstrate that the problem can also be well resolved by providing a suitable feedback in the system . the latter can be found to be present in most of the natural systems , including neural networks @xcite , genetic networks @xcite , vision systems @xcite , etc . the vital role of feedback in controlling the dynamics of the given system and the control over synchronization @xcite are already known , which can be seen in a variety of fields ranging from electronics @xcite , biology to quantum information @xcite . for example , the feedback control of deep brain simulation has been found to be the most effective treatment for chronic neural diseases like essential tremor , dystonia and parkinson s disease @xcite . also , the feedback generated by the voltage - gated ion channels in neural cells is found to be crucial in generating neural signals @xcite . in this article , we show the applicability of the feedback technique in resurrecting oscillations in a wide range of systems . we show both numerically and analytically that the addition of feedback destabilizes the stable attractors which results in a wiping out of the oscillation quenching and inducing a resurrection of oscillations . in addition , by considering a rather general system , we prove analytically the above destabilizing nature over the ad state . further , it will be more important to develop an adaptable mechanism thereby improving the ones available in the literature at present . this is because , for example to use the available parametric mismatch method , one needs to tune the internal parameters of the system , while the processing delay also depends on the underlying process of the system where that process may be unknown in many situations so as to hinder the efficiency . in contrast , the feedback method suggested here can be given more easily which is already in practice under different contexts such as in deep brain simulation @xcite . in addition to the above adaptable nature of feedback , with the aid of numerical and analytical studies , we show the important fact that this method does not impose a restriction that the output of all the oscillators need to be fedback . from the output of only a few of the system oscillators , we show that in typical systems the resurrection of oscillations can be achieved easily . the structure of the paper is as follows . in section [ genr ] we present the general form of the system that we consider . in section [ diffu ] , we illustrate the role of feedback on two important oscillation quenching scenarios , namely the symmetry breaking coupling and the parametric mismatch in a system of diffusively coupled stuart - landau oscillators , through numerical analysis . in section [ anal ] , we present suitable analytical support of the numerical results based on an appropriate linear stability analysis . in section [ dynam ] , we illustrate the role of the considered feedback in indirectly coupled or dynamically coupled stuart - landau oscillators . a realistic chemical oscillator model , namely the brusselator model , is considered in section [ brusec ] . we have also proved the applicability of the technique in more general situations over amplitude death state in appendix [ gen_an ] . in addition , in the appendix [ more ] we illustrate our method with different coupling schemes and with different models such as van der pol oscillator , rsseler system and so on . appendix [ app2_an ] includes the details in obtaining the boundary curves of the ad region which were given in sec . a summary of our results and conclusions are presented in section [ conclu ] . consider a system of coupled dynamical systems , @xmath0 where @xmath1 characterizes the dynamics of the isolated @xmath2-th system , @xmath3 is a @xmath4 dimensional state vector of the system @xmath2 , @xmath5 is the @xmath6 coupling matrix of the network , @xmath7 are respectively the uniform coupling and feedback strengths , @xmath8 is a coupling function and @xmath9 is the feedback term which can be written as @xmath10 . here @xmath11 is simply a @xmath12 constant matrix and @xmath13 characterizes the feedback and it depends on the state vectors of the system . such a dependence of @xmath14 on the state vectors of the system may be linear ( example : @xmath15 ) or nonlinear ( example : @xmath16 ) , where @xmath17 s represent weight factors which can take values from @xmath18 to @xmath19 and @xmath20 is a suitable number . in our following study , we consider the form of @xmath14 as @xmath15 . in the appendix [ gen_an ] , we have considered rather general forms for @xmath21 and @xmath22 and shown analytically that the trivial ad state which appears in the system could be wiped out through the strengthening of @xmath23 so as to resurrect oscillations . ) with respect to the introduced two different forms of feedback ( i ) @xmath24 and ( ii ) @xmath25 , @xmath26 , for the choice @xmath27 and @xmath28 . ( b)for two choices of @xmath29 , @xmath30 and @xmath31 , the change in the hb point by the increment of @xmath32 ( which appears in the expression of @xmath17 ) with @xmath33 in ( [ sys1 ] ) . inset in ( b ) projects the od and oscillatory ( os ) regions in the ( @xmath34 ) space for @xmath35 . ( c),(d ) and ( e ) : temporal behaviors of the system in the os , od and revoked os state respectively for the values of ( @xmath34 ) @xmath36 ( @xmath37 ) , ( @xmath38 ) , and ( @xmath39 ) . the color bar in figs . ( c ) , ( d ) and ( e ) gives information on the value of @xmath40 corresponding to different colors.,width=362 ] to start with , we use the paradigmatic model known as the coupled stuart - landau oscillators for the purpose of illustration for the advocated feedback method . it is well known that the dynamical equation defining the stuart - landau oscillator can be obtained from a general ordinary differential equation near a hopf bifurcation point @xcite . as the hopf bifurcation arises widely in the literature , the stuart - landau oscillator helps to model a variety of systems in different areas ranging from biology @xcite to lasers @xcite and is also used in the reaction - diffusion process @xcite . this model is often used in neural networks to model spiking neurons @xcite . the first reason for using the model in neural networks is that the periodically spiking neurons have an exponentially stable limit cycle attractor and secondly the real part of the complex amplitude of the stuart - landau oscillator can describe the membrane voltage in the neurons and the imaginary part can be related to the recovery variable embedding the effects of the other variables of physiological neuron models @xcite . thus , in the literature we can find the use of this model in studying the effects of synchronization and desynchronization in neural networks @xcite and also various collective dynamical states , including chimeras @xcite . in our study , we also include other useful models such as the van der pol oscillator , rssler system and brusselator model . the corresponding results are briefly indicated in the appendix b and sec [ brusec ] . now , we first consider a system of coupled stuart - landau oscillators which is characterized by @xmath41 where @xmath42 , @xmath43^{t}$ ] , @xmath44 is the kronecker delta ( @xmath44 @xmath45 , if @xmath46 and @xmath47 , if @xmath48 ) and * i * represents the identity matrix , in eq . ( [ sys1 ] ) . here the diffusive coupling acts only on the first half of the evolution equations ( @xmath49- variable alone ) which breaks the rotational symmetry @xcite and consequently induces oscillation death in the system . to elucidate clearly the role of feedback in ( [ sys1 ] ) , we first consider the case @xmath27 , with @xmath50 . in the case @xmath51 , @xmath52 , we note that the individual systems in ( [ sys1 ] ) show limit cycle oscillations with @xmath53 . the introduction and strengthening of diffusive coupling ( @xmath54 ) stabilizes the symmetric pair of nontrivial equilibrium points @xmath55 @xmath56 , where @xmath57 , @xmath58 , @xmath59 and @xmath60 , through a sub - critical hopf bifurcation @xcite which gives rise to od . with the introduction of feedback @xmath61 , these nontrivial equilibrium points soon lose their stability via a super - critical hopf bifurcation . in our study , we introduced such a feedback in two ways : ( i ) @xmath62 and ( ii ) @xmath63 , @xmath26 ( again here @xmath64 denotes the kronecker delta ) . first by setting @xmath62 , for different values of @xmath29 we traced the hopf bifurcation points and these points are collectively shown as the hb curve@xmath65 in fig . [ am1](a ) . the region lying under this curve is an od region ( denoted by od@xmath65 ) and the parametric region above this curve is free from od and corresponds to oscillatory states ( os ) . similarly for @xmath63 , @xmath26 , the od region ( od-@xmath66 which includes od-@xmath19 also ) and the curve of hopf bifurcation points ( hb curve@xmath67 ) have been shown in fig.[am1](a ) , which show that the uniform distribution of @xmath17 as in ( i ) helps to redeem from the od state sooner than in the case ( ii ) . by extending the constituents of the network to @xmath33 , we have verified that this technique can work as well with larger @xmath68 . further , we have checked whether the feedback needs contributions from all the constituents of the network . this is vital as in a practical situation we can not assure or impose all the constituents to contribute to the feedback . thus , we have distributed @xmath17 s as @xmath69 , @xmath70 , where @xmath32 determines the number of contributing components of the network . by varying @xmath32 , we have drawn the hb curves separating the od state with the os state for two different values of @xmath29 , @xmath30 and @xmath31 in fig . [ am1](b ) . interestingly , these curves demonstrate clearly that the contribution from even a single oscillator is sufficient to revoke oscillations in the network . secondly , the critical value of @xmath23 ( hb point ) above which the os state arises gets decreased sharply with that of @xmath32 . these facts prove that this technique can work well regardless of the number of oscillators present in the network and the number of them which contributes towards the feedback . for simplicity , we chose @xmath35 in the following studies . in the system ( [ sys1 ] ) endowed with a parametric mismatch of @xmath71 , @xmath72 and @xmath27 . ( c ) the reduction of ad regions in the ( @xmath73 ) space with the increase of @xmath23 and by keeping @xmath74 and @xmath27 . ( d ) and ( e ) the emergence of os state from ad and od states in @xmath33 case for @xmath71 and @xmath72 which is obtained for two different set of initial conditions . ad region in fig . ( e ) is also in conformity with the analytical results presented in the sec . [ anal].,width=362 ] the inset of fig . [ am1](b ) depicts the information about od and os regions with @xmath33 oscillators . correspondingly , in figs . [ am1](c ) , [ am1](d ) and [ am1](e ) , we have captured the temporal behaviors of the system ( in the @xmath75-variables ) for different sets of ( @xmath76 ) , for a finite time interval after leaving out sufficiently large transients . the first one ( fig . [ am1](c ) ) shows the temporal behavior at an os state of the system when @xmath52 and @xmath77 , where the value of @xmath29 is not sufficient to induce od . now increasing @xmath29 to @xmath78 ( while keeping @xmath52 ) , the subsequent figure ( fig . [ am1](d ) ) shows the quenching of this oscillation . now switching @xmath23 on , the temporal behavior in fig . [ am1](e ) shows the resurrected oscillations for @xmath79 . rubchinsky and sushchik @xcite have introduced a disorder in the form of parametric mismatch which revokes oscillations in ( [ sys1 ] ) . on the other hand , just like symmetry breaking which is predominant in inducing od , the parametric mismatch is also a key candidate that induces ad in the system . recently , koseska _ et al . _ in @xcite have shown that an increase in this inhomogeneity not only induces ad but also od , whereas the feedback mechanism that we consider here does not show such a behavior in the system . this feature provides a definitive advantage over parametric mismatch . now , we augment the system with a parametric mismatch . for the mismatch in the parameter @xmath80 , @xmath81 , in the above coupled stuart - landau oscillators ( [ gen])-([sys1 ] ) with @xmath27 , we find the existence of both ad ( @xmath82 ) and od ( @xmath30 ) while @xmath52 . now , from these death states the transitions towards os state by @xmath23 are demonstrated in figs . [ bif1](a ) and [ bif1](b ) , which show the destabilization of both the ad and od states via supercritical hopf bifurcations . further , the role of @xmath23 over @xmath83 and @xmath29 is more clearly demonstrated in fig . [ bif1](c ) , where the colored islands denote the ad regions for the values of @xmath23 @xmath84 and @xmath85 . one can check that similar phenomenon occurs to od regions also which we do not depict here explicitly . now considering the case of @xmath33 globally coupled oscillators with @xmath86 for @xmath87 and @xmath88 for @xmath89 in ( [ sys1 ] ) , we have demonstrated the reduction in the ad and od regions in figs . [ bif1](d ) and [ bif1](e ) . the above two figures are plotted for two different sets of initial conditions . among them , the ad region in fig . [ bif1](e ) is the analytically relevant region which has been obtained in the next section . in the next section , the above obtained numerical results on the system ( [ sys1 ] ) are verified through analytical results wherever possible . also , we illustrate the applicability of the technique to several situations in appendix [ more ] , where we considered ( i ) repulsive link : stuart - landau ( sl ) oscillator , ( ii ) conjugate coupling : sl oscillator , ( iii ) repulsive link : van der pol oscillator , ( iv ) directly and indirectly coupled rssler system and ( v ) other chaotic oscillators ( sprott and lorenz systems ) . ) space for @xmath27 oscillators in system ( [ sys1 ] ) with @xmath90 . with parametric mismatch case : fig . ( b ) ad regions of system ( [ sys1 ] ) for @xmath91 , @xmath74 and @xmath92 bounded by the curves defined in ( [ condi ] ) . fig . ( c ) analytically obtained boundaries of ad regions in ( @xmath83 , @xmath29 ) for different values of @xmath23 ( = @xmath93 ) , where @xmath92 . ( d ) hb curves for the three different cases ( i ) @xmath92 , ( ii ) @xmath94 and ( iii ) @xmath95 for @xmath91 and @xmath72.,width=366 ] in this section , we present relevant analytical confirmations of the numerical results corresponding to the system ( [ sys1 ] ) for the cases with and without parametric mismatch discussed earlier in sec . [ diffu ] . in addition through the obtained analytical results , we show the effectiveness of the technique with the feedback contribution coming from a few number of oscillators in the network . in the absence of any parametric mismatch , as pointed out earlier the system in ( [ sys1 ] ) has a trivial equilibrium point @xmath96 : @xmath97 and two pairs of non - trivial equilibrium points as @xmath98 where @xmath60 and @xmath99 . the linear stability of these fixed points is determined by the eigenvalues of the jacobian matrix @xmath100 where @xmath101 , @xmath102 , @xmath103 . while @xmath52 , the eigenvalues of @xmath104 corresponding to the trivial equilibrium point @xmath96 are @xmath105 the eigenvalues corresponding to the equilibrium points @xmath106 and @xmath107 are given by @xmath108 similarly , @xmath109 and @xmath110 have the set of eigenvalues @xmath111 from eqs . ( [ 0eig]-[2eig ] ) we can note that among the five equilibrium points , @xmath106 and @xmath107 are found to have all their eigenvalues satisfying the condition @xmath112<0 $ ] for the parametric range @xmath113 while @xmath114 , @xmath115 while @xmath116 and @xmath117 while @xmath118 and thus they are stable in this range . on the other hand the other equilibrium points @xmath96 , @xmath119 and @xmath120 can never become stable for any choice of parametric values . thus the stabilization of @xmath121 and @xmath122 essentially gives rise to oscillation death in the system . now we introduce the feedback in such a way that @xmath123 and @xmath124 of @xmath14 in ( [ sys1 ] ) take the values @xmath125 and @xmath126 . with such a choice , we find that the stability determining eigenvalues corresponding to the equilibrium points change as @xmath127 the above equation shows that an increase in @xmath23 can destabilize the equilibrium points @xmath121 and @xmath122 through a hopf bifurcation . in the cases where @xmath128 , for all values of @xmath29 the hopf bifurcation occurs at @xmath129 whereas in the case of @xmath130 , if @xmath131 , the hopf bifurcation occurs at @xmath132 where @xmath133 is given in ( [ c11 ] ) . if @xmath134 , the hopf bifurcation occurs as in ( [ etac1 ] ) . the other equilibrium points @xmath96 , @xmath119 and @xmath120 are found to remain unstable . for the case of @xmath28 , the curve of hopf bifurcation points ( @xmath135 ) separating the death regimes with the oscillatory regimes is shown in fig . [ anal1](a ) which matches exactly with the one obtained numerically ( hb curve - @xmath19 in fig . [ am1](a ) ) . the analytical treatment of the other case corresponding to the hb curve-2 can also be investigated in a similar manner , though the results can not be written down in such a transparent manner . so we do not present the details here . next with the introduction of a parametric mismatch in the system ( [ sys1 ] ) , we prove the validity of the technique for the trivial ad state analytically for the @xmath27 case . as mentioned earlier in sec . [ parame ] , an increase in the value of the parametric mismatch parameter @xmath83 causes the stabilization of the trivial equilibrium point @xmath97 and thus introduces ad in the system . to destabilize the latter , we first introduce the feedback in such a way that @xmath136 , where @xmath137 . as before , through a linear stability analysis of the system ( [ sys1 ] ) with parametric mismatch included , we look for the stable regions of the equilibrium point @xmath97 , which can be studied from the characteristic equation of the linear eigenvalue problem of the system as @xmath138 ) with @xmath33 oscillators in the os , od and revoked os states respectively , for the values of ( @xmath34 ) @xmath36 ( @xmath139 ) , ( @xmath140 ) , and ( @xmath141 ) . here , we have taken @xmath142 , @xmath143 , @xmath144 , @xmath145 , @xmath72 and @xmath146 . the color bars in figs . ( a ) , ( b ) and ( c ) represent the values of the variables @xmath40.,width=362 ] the coefficients @xmath147 , @xmath148 , @xmath149 and @xmath150 in the above equation are given by @xmath151 , \nonumber \\ a_{02}&=&\frac{1}{4}\left[(4(1-\epsilon)+\epsilon \omega_1 \omega_2 ) \tilde{a}-(\epsilon-4)(\omega_1 ^ 2 a_2+\omega_2 ^ 2 a_1 ) \right ] , \nonumber \\ a_{03}&=&(1+\omega_1 ^ 2)(1+\omega_2 ^ 2)-\frac{\epsilon}{2 } ( 2+\omega_1 ^ 2+\omega_2 ^ 2 ) . \label{coeff}\end{aligned}\ ] ] here @xmath152 . since the characteristic equation for the eigenvalues is quartic in nature , we use the well known routh - hurwitz criteria @xcite to obtain the stable ad regions of the system . by doing so , we find that the ad regions are bounded by the curves @xmath153 the details of obtaining the boundary curves from the r - h criteria are presented in the appendix [ app2_an ] , and the ad region bounded by the curves ( [ condi ] ) has been shown in fig . [ anal1](b ) for @xmath125 and @xmath126 . [ anal1](c ) which portrays the boundaries of the ad regions in the ( @xmath73 ) space clearly shows that the analytical results match nicely with that of the numerical results given in fig . [ bif1](c ) . next by varying the nature of oscillators contributing towards feedback we have plotted fig . [ anal1](d ) , where we considered three cases ( i ) @xmath125 , @xmath126 ( both the oscillators contributing ) ( ii ) @xmath125 , @xmath154 ( only the high frequency oscillator contributing ) and ( iii ) @xmath155 , @xmath126 ( only the low frequency oscillator contributing ) . the analytically obtained hopf bifurcation curves for all the above three cases have been presented in fig . [ anal1](d ) . from the figure , we can note that for the first two cases ( i ) and ( ii ) the os state gets revoked from the ad state for even small values of @xmath23 and the hopf bifurcation curves of these two cases are closer to each other . but , in the case where the low frequency oscillator alone is contributing , we find that comparatively higher values of @xmath23 are needed to revoke oscillations . this shows that for a quicker resurrection of oscillations , the feedback from the high frequency oscillator is preferable . however , the resurrection of oscillations is possible even if one of the oscillators is contributing towards the feedback . next , we extend our studies on the revocation of oscillations from the ad state for the case of @xmath33 oscillators , where the parametric mismatch in the system is introduced in such a way that the system has two groups of oscillators . the first group contains @xmath156 oscillators with @xmath157 , @xmath158 and the second group has @xmath159 oscillators with @xmath160 , @xmath161 with @xmath162 and @xmath163 . also , we consider that among the @xmath156 oscillators in the first group only the output of a sub - group of @xmath164 oscillators is fed back , in other words , the @xmath17 s of @xmath14 in ( [ sys1 ] ) take the values as @xmath165 similarly in the second group of @xmath159 oscillators only the output of @xmath166 oscillators is fedback or @xmath167 the total number of oscillators contributing towards feedback is @xmath168 . the results corresponding to @xmath169 oscillators presented earlier in sec . [ parame ] clearly demonstrate the appearance of ad state in the system . the temporal behavior of the system in the original os state , ad state and revoked os state are shown in fig . [ sync ] , which shows the coherent nature among the oscillators in the first and second group . next to study the case of the @xmath33 oscillators analytically , we first try to reduce the problem to a simpler level . ) space for the values of ( @xmath170)@xmath36 @xmath171 , @xmath172 and @xmath173 . ( b ) shows the curves of @xmath174 for various combinations of ( @xmath164 , @xmath166 ) with fixed @xmath175 and @xmath176 . ( c ) fixing @xmath177 or @xmath178 , the values of @xmath174 for different values of @xmath166 or @xmath164 is plotted with @xmath176 . ( d ) shows the value of @xmath174 for different values of @xmath164 and @xmath166 for @xmath179 , where the values of @xmath174 are represented by the color function . in the above figures , we make the choice @xmath169 , @xmath146 and @xmath72,width=359 ] due to the existence of coherence among the oscillators in the first and second group , we represent the state of the oscillators in the first group by @xmath180 and the state of the oscillators in the second group by @xmath181 @xmath182 where @xmath183^{t}$ ] and @xmath184 , @xmath57 . the coefficients in the feedback take the form @xmath185 and @xmath186 , where @xmath187 , which is the ratio of oscillators present in the first group , and @xmath188 is the ratio of oscillators present in the second group . @xmath189 and @xmath190 are the ratio of the oscillators contributing towards feedback from the first and second groups , respectively . thus @xmath191 and @xmath192 , where @xmath164 and @xmath166 are respectively the number of oscillators in the first and second groups that are contributing towards feedback . the stability of the system ( [ sys_app ] ) around the equilibrium point @xmath97 can be studied similar to the previous case and one can obtain the characteristic equation for the eigenvalues of the above system ( [ sys_app ] ) as @xmath193 the coefficients in the above equation are given by @xmath194 where @xmath195 . as in the previous case , using the r - h criteria we determine the ad regions of the system . consequently , we find that the ad regions are bounded by the curves defined by @xmath196 next , using the above relations we find the ad regions in the different cases of the system . we consider here only two cases , ( i ) @xmath197 , ( ii ) @xmath198 . _ case-1 : @xmath169 . _ in this case , the population of the high frequency oscillators ( @xmath156 ) and the population of the low frequency oscillators ( @xmath159 ) are equal ( note that @xmath162 ) . first , we check the consistency of the obtained analytical results with the numerical results . for @xmath169 , we have plotted the boundaries of the ad regions obtained from analytical and numerical studies for different values of @xmath164 and @xmath166 in fig . [ an_ne](a ) . ( note that for @xmath199 and @xmath200 , the numerical results have been given in fig . [ bif1](e ) of sec . [ parame ] ) . the latter shows the consistency between the numerical and analytical results . now , we look at the preferential feedback configuration for quicker resurrection of oscillations . for the purpose , we fix the total number of oscillators contributing towards feedback as @xmath201 and vary the number of oscillators contributing from the first group ( @xmath164 ) and from the second group ( @xmath166 ) . fig . [ an_ne](b ) shows the boundaries of the ad regions for different values of @xmath164 ( also @xmath202 ) , where we can observe that on increasing the contributions from the high frequency oscillators ( or @xmath164 ) , the resurrection of oscillations occur for lower values of @xmath23 . again in fig . [ an_ne](c ) , we first fixed @xmath177 and plotted the critical value of @xmath23 needed for resurrection of oscillations ( @xmath174 ) for different values of @xmath166 . similarly , we fixed @xmath178 and plotted @xmath174 for different values of @xmath164 in the same fig . [ an_ne](c ) . from the figure , we find that a considerable decrease in the value of @xmath174 occurs only when @xmath164 is varied . the above results show that a feedback from the high frequency oscillators is more preferable for a quicker resurrection of oscillations . they are also evident from the fig . [ an_ne](d ) , where @xmath174 is represented by a color function . in the figure , one can find that the rate of change in the value of @xmath174 is larger along the @xmath164 direction . _ case-2 : @xmath198 . _ just as the technique shows preference over the feedback of high frequency oscillators , it shows dependence over the population in the two groups of oscillators . to illustrate the above , we considered two situations ( i ) @xmath156 @xmath203 @xmath159 , ( ii ) @xmath156 @xmath204 @xmath159 . [ an_ne2](a ) has been plotted for the case ( i ) where @xmath205 and @xmath206 . similarly fig . [ an_ne2](b ) corresponds to the case ( ii ) where @xmath207 and @xmath208 . the values of @xmath174 for different values of @xmath164 and @xmath166 are presented in fig . [ an_ne2 ] . comparing fig . [ an_ne2](a ) with fig . [ an_ne2](b ) , we find that the values of @xmath174 are larger in the case of @xmath209 than that in the case @xmath210 . thus in the case where lower frequency oscillators are highly populated than the high frequency oscillators , we require stronger feedback to revoke oscillations . for different values of @xmath211 for the case ( a ) @xmath210 ( @xmath205 , @xmath206 ) , ( b ) @xmath212 ( @xmath207 , @xmath208 ) , with @xmath146 and @xmath72 . , width=366 ] following the studies on directly coupled systems , we turn to check the validity of the proposed scheme to indirectly coupled oscillatory systems . for this purpose , we consider a collective system coupled to a dynamic environment , where the combined set of dynamical equations is characterized by ) space corresponding to the system ( [ sys2 ] ) with @xmath27 and @xmath213 , ( b ) the reduction in the ad and od regions in the system with @xmath33 coupled oscillators in the ( @xmath76 ) space for @xmath213 . ( c ) , ( d ) , ( e ) and ( f ) the corresponding temporal behavior of the system ( [ sys2 ] ) of @xmath33 oscillators in the ad , od and the revoked oscillatory state.,width=362 ] @xmath214 here , the state of the system along with the environment is defined by the state vector @xmath215^t$ ] , where the variables @xmath216 and @xmath40 correspond to the system and @xmath217 represent the environment . while @xmath52 , the increase in @xmath29 causes a stabilization of the trivial equilibrium point ( @xmath218)=(@xmath219 ) and a further increase in @xmath29 stabilizes a pair of nontrivial equilibrium points defined by ( @xmath218)=(@xmath220 ) , where @xmath221= @xmath222 , @xmath223 , @xmath224 and @xmath225 via pitchfork bifurcation @xcite . the introduction of the feedback , @xmath61 , destabilizes both the ad and od states via hopf bifurcation . such a reduction in the territories of ad and od states is illustrated in the ( @xmath34 ) space in fig.[bif2](a ) . the curve made up of the hopf bifurcation points separates out the ad and od regions with the os region . by extending @xmath68 to @xmath226 , fig . [ bif2](b ) depicts the ad and od regions of the system . when @xmath52 , the ad state which arises by increasing @xmath29 is found to disappear with an increase of @xmath29 . but for larger @xmath29 , we find the appearance of the od state . now by switching @xmath23 on , both the ad and od states are shown to be wiped out simultaneously . the temporal behaviors of the system in the ad state and the resurrected oscillatory state which arise through the enhancement of @xmath23 are shown in figs . [ bif2](c ) and [ bif2](d ) . similarly , figs . [ bif2](e ) and [ bif2](f ) show the behavior of the system at the od state and the resurrection of oscillations by an increase in @xmath23 . ) for @xmath227 and @xmath228 . ( b ) the occurrence of od through the coupling of the system for @xmath229 and @xmath52 , and ( c ) the resurrection of oscillations by introducing feedback @xmath230.,width=362 ] in this section , we consider an interesting coupled chemical oscillator modeled by the brusselator model @xcite . we consider the case where two identical cells are coupled . in such a case , the functions in eq . ( [ gen ] ) describing the system are given by @xmath231 in the absence of the coupling , the system shows stable limit cycle oscillations which has been illustrated in fig . [ br](a ) . by introducing the coupling , the system tends to an inhomogeneous steady state @xcite . for example , for @xmath227 , @xmath228 and @xmath229 , the system tends to an inhomogeneous steady state which has been illustrated in fig . [ br](b ) . in such a realistic example , by introducing the feedback , we found that the oscillations are revoked by increasing @xmath23 and it has been illustrated in fig . [ br](c ) . from a knowledge of the role of feedback in controlling the dynamics and the coherent activities of the system such as synchronization @xcite , we have here analyzed whether it can control oscillation quenching tendencies . for this purpose , we have demonstrated the effect of feedback over quenching induced by parametric mismatch and symmetry breaking ( the key candidates for inducing ad and od ) , indirect coupling and for some more cases ( given in appendix [ more ] ) , through numerical as well through analytical studies wherever possible . further , through analytical studies on ad state of a more general system ( appendix [ gen_an ] ) , we found the general applicability of the mechanism where the feedback resurrects oscillations from the ad state . in the case of od , a proper / suitable form of linear feedback would be helpful to resurrect oscillations or one can also explore the role of nonlinear feedback in such nontrivial od states . from the results obtained for different cases , we find that the trivial ad state is found to be destabilized through hopf bifurcation whereas the nontrivial od states are found to be destabilized even through saddle node type bifurcation ( see appendix [ more ] ) . in addition to the adaptability of the technique in practical situations , we have illustrated here one more important feature of the technique namely that it does not put any restriction over the number of oscillators contributing towards feedback . even with the feedback from a few number of oscillators we can break the death state of the system and thus provide an attractive methodology in practical situations . considering a two population network , the contribution from the high frequency oscillators are found to be more preferable compared to the feedback from the low frequency oscillators . the work of vkc forms part of a research project sponsored by insa young scientist project . the work forms part of an irhpa project of ml , sponsored by the department of science technology ( dst ) , government of india , who is also supported by a dae raja ramanna fellowship . sk thanks the department of science and technology ( dst ) , government of india , for providing a inspire fellowship . in this appendix , we show the applicability of the feedback technique over the ad state of a general two coupled system . for this purpose , we assume @xmath27 and the general forms for @xmath21 , @xmath232 and @xmath22 in eq . ( [ gen ] ) , where @xmath21 can be chosen as a polynomial in @xmath233 , @xmath234 in the above , @xmath235 and @xmath236 are nonlinear functions in @xmath216 and @xmath40 and the constants @xmath237 , @xmath238 , @xmath239 and @xmath240 are system parameters . the function @xmath241 can be written as @xmath242 where @xmath243 is a nonlinear function in @xmath244 that introduces nonlinear coupling in the system . the coupling matrix @xmath22 can be taken as @xmath245 where @xmath246 , @xmath247 and @xmath248 . the systems are coupled through both direct and conjugate variables , where @xmath249 introduces direct coupling , @xmath250 and @xmath251 introduce conjugate coupling . @xmath252 , @xmath189 and @xmath190 are coupling strengths . the system has a trivial equilibrium point at @xmath97 , which may become stable due to the parametric mismatch in the system or due to the coupling in the system . for example , in @xcite the conjugate coupling in the system induces ad even when the oscillators are identical . the feedback can be given as @xmath253 , where one can also add nonlinear terms in the feedback , if needed : @xmath254 first , considering the case of coupled identical oscillators @xmath255 , @xmath256 , @xmath257 and @xmath258 ( @xmath57 ) , the linearization around the trivial equilibrium point @xmath97 can be done . the eigenvalues corresponding to the case can be obtained easily ( note that the nonlinear terms in ( [ fgen ] ) and ( [ hgen ] ) do not play any role in the linearized equation for the trivial equilibrium point ) . the obtained eigenvalues corresponding to the case are of the form @xmath259 in the above @xmath260 and @xmath261 are eigenvalues corresponding to the trivial equilibrium point of the system when the feedback is absent ( @xmath52 ) . they are given by @xmath262 depending on the values of the system and coupling parameters , the real part of the eigenvalues @xmath263 and @xmath264 in ( [ 12 ] ) and ( [ lamgen ] ) are positive or negative while @xmath52 . when all the eigenvalues in ( [ 12 ] ) and ( [ lamgen ] ) have negative real parts , the stabilization of the equilibrium point gives rise to ad in the system . from eq . ( [ 12 ] ) , we notice that the increase in @xmath23 causes the eigenvalues @xmath265 to be more positive . thus a destabilization of the equilibrium point @xmath97 occurs or it wipes off ad . if the equilibrium point was unstable while @xmath52 , the increase in @xmath23 never stabilizes the equilibrium point . thus the above analysis makes clear the role of @xmath23 in destabilizing the attractor at @xmath97 . we can observe a similar effect even in the case where parametric mismatch is present ( the case where @xmath266 , @xmath267 , @xmath268 and @xmath269 ) in the system . the eigenvalues of the system can be obtained by solving the equation @xmath270 where @xmath271 . as @xmath148 , @xmath149 , @xmath150 are not simple in their form , we do not present them here . although the eigenvalues @xmath272 obtained from ( [ eigll ] ) are not of the simple form , the stability of the equilibrium point @xmath97 in the different parametric regions can be found through the routh - hurwitz ( r - h ) criteria . from the r - h criteria , an equilibrium point is said to be stable only when all the conditions given below are satisfied by the coefficients in the eigenvalue equation ( [ eigll ] ) . the r - h criteria are given as @xmath273 if the coefficients in the characteristic eigenvalue equation ( [ eigll ] ) fail to satisfy any one of the condition given above , the equilibrium point becomes unstable . in this aspect , we consider one of the simpler condition in ( [ acon ] ) , namely @xmath274 . the condition @xmath274 is broken when @xmath275 , thus this clearly shows that an increase in @xmath23 destabilizes the equilibrium point @xmath97 . further , more clear analytical illustration on the role of @xmath23 in the parameter mismatched system is given in sec . [ par_ann2 ] and [ par_ann100 ] with stuart - landau model as an example . as the above type of proof for the non - trivial od state is too cumbersome , we have illustrated the role of feedback over the state with more examples in the body of the paper as well in appendix [ more ] both numerically and analytically ( in some cases ) . from the above illustrations , one can also notice that in the case of ad , the nonlinear feedback terms can not play any role ( as they lose their significance in the linearized limit ) and do not provide any control over it , whereas in the case of od , the nonlinear feedback also can provide a control over it . we consider the case of two stuart - landau oscillators coupled diffusively with a repulsive link ( @xmath27 ) , as studied in @xcite . the functions characterizing this equation have the forms , ) with @xmath23 for @xmath213 . ( b ) and ( c ) show the corresponding transition routes from ad and od states to os state . ( d ) the emergence of os regions in the system ( [ sys4 ] ) with the increase of @xmath23 for @xmath276 , @xmath213 and @xmath277 in ( [ sys4 ] ) . , width=355 ] @xmath278 the eigenvalues corresponding to the trivial equilibrium point @xmath97 of the system are @xmath279 we demonstrate the destabilization of this ad state as well as the od state corresponding to the system ( [ sys3 ] ) with respect to the feedback in fig . [ re](a ) , where we can find that an increase in @xmath23 causes the reduction in ad and od regions of the system . the transition route followed by the system , as it transits to oscillatory state , is shown in figs . [ re](b ) and [ re](c ) . these figures show that the oscillations are resurrecting from the ad state via hopf bifurcation as in the previous cases , whereas the resurrection of oscillations from od state occurs through saddle node bifurcation . ) in the ( a ) ad state for ( @xmath34 ) @xmath36 ( @xmath280 ) , ( b ) the revoked os state for ( @xmath34 ) @xmath36 ( @xmath281 ) , ( c ) od state for ( @xmath34 ) @xmath36 ( @xmath282 ) , and ( d ) the revoked os state for ( @xmath34 ) @xmath36 ( @xmath283 ) . , width=359 ] recently , ad and od in the same system with large number of oscillators ( @xmath68 ) coupled globally has been seen in @xcite , whose equation is defined by @xmath284 even with @xmath276 oscillators in system ( [ sys4 ] ) , we have shown the reduction in ad and od regions of the system in fig . [ re](d ) . the temporal behavior of the system for different values of @xmath29 and @xmath23 are shown in figs . [ re2](a)-(d ) . next , we consider the case of @xmath27 oscillators coupled through a conjugate coupling described by @xmath285 this system has a trivial equilibrium point at @xmath286 : @xmath97 and has pairs of non - trivial equilibrium points for @xmath287 , @xmath288 : @xmath289 ) , where @xmath290 , @xmath291 and @xmath292 : @xmath293 ) , where @xmath294 , @xmath295 in which @xmath296 . the trivial equilibrium point @xmath286 has the eigenvalues @xmath297 for @xmath52 , the equilibrium point @xmath286 is unstable for all values of @xmath29 and for @xmath298 a saddle node type bifurcation occurs which stabilizes @xmath299 and @xmath300 as shown in fig . [ con](a ) . for @xmath61 , the eigenvalues corresponding to the equilibrium point @xmath286 are still unstable , and thus the system is still free of ad . then , the equilibrium points @xmath299 and @xmath300 are also destabilized through hopf bifurcations which is demonstrated in fig . [ con](b ) . then , the od regions of the system in the ( @xmath34 ) space is shown in fig . [ con](c ) , which clearly demonstrates the destabilization of od with the introduction of feedback . in the case of conjugate coupling . ( b ) the destabilization of the od state with the introduction of feedback via hopf bifurcation ( @xmath301 ) . ( c ) the reduction in the od regions of the system in the ( @xmath34 ) space for @xmath302.,width=359 ] next , we illustrate the role of feedback in the case two van der pol oscillators coupled diffusively through a repulsive link . the corresponding dynamical equations are defined through @xcite @xmath303 the ad and od regions of the system in the ( @xmath34 ) space are given in fig . [ vd ] which show the reduction in the ad and od regions with respect to @xmath23 . the transition from ad to oscillatory state occurs through a hopf bifurcation . on the other hand considering the transition from od to oscillatory state , the od state is transformed to ad through inverse pitchfork bifurcation and the oscillatory state arises from the ad state through hopf bifurcation . this shows that the role of the feedback in setting oscillations back in the system is not restricted to any particular oscillator . in the next example , we show that the feedback can destabilize oscillation quenching scenario even in chaotic oscillators . ) with respect to @xmath29 and @xmath23 for @xmath304 . ( b ) the transition route from od state to os state via ad state.,width=355 ] we consider the case of the @xmath27 coupled chaotic rssler system @xcite defined by @xmath305 here , the oscillators are coupled directly by diffusive type coupling and are also coupled indirectly to an environment defined by the variable @xmath306 . this system ( [ ros ] ) has only non - zero equilibrium points which include @xmath288 : @xmath307 @xmath36 @xmath308 , where , @xmath309 . for the parametric choice @xmath310 , @xmath311 , @xmath312 and @xmath313 , the equilibrium point @xmath300 is stable . the od regions corresponding to the above equilibrium points are given in figs . [ rtt](a ) and [ rtt](b ) with respect to the direct coupling strength ( @xmath29 ) and with respect to the indirect coupling strength ( @xmath272 ) , respectively . the above figures clearly show that the feedback is applicable even for the case of chaotic oscillators . ) ( a ) in the ( @xmath314 ) space for @xmath315 , ( b ) in the ( @xmath76 ) space for @xmath316 . in both the figures , we fixed @xmath310 , @xmath311 , @xmath317 and @xmath318.,width=321 ] to illustrate further the role of feedback in chaotic oscillators , we consider the @xmath27 coupled sprott and lorentz oscillators which are defined respectively by + ( i)two coupled sprott systems with repulsive link @xcite , @xmath319 in the case of ( a ) the sprott system ( [ spr ] ) for @xmath320 and ( b ) the lorenz system ( [ lor ] ) for @xmath321 , @xmath322 and @xmath323.,width=355 ] and ( ii)diffusive coupling among two lorenz oscillators @xcite , @xmath324 the ad and od regions corresponding to the coupled sprott and lorenz systems ( ( [ spr ] ) and ( [ lor ] ) ) are given in fig . [ s ] , which again confirm that the feedback wipes out ad and od in the system . the coefficients in the characteristic eigenvalue equation ( [ eveq ] ) are given by @xmath151 , \nonumber \\ a_{02}&=&\frac{1}{4}\left[(4(1-\epsilon)+\epsilon \omega_1 \omega_2 ) \tilde{a}-(\epsilon-4)(\omega_1 ^ 2 a_2+\omega_2 ^ 2 a_1 ) \right ] , \nonumber \\ a_{03}&=&(1+\omega_1 ^ 2)(1+\omega_2 ^ 2)-\frac{\epsilon}{2 } ( 2+\omega_1 ^ 2+\omega_2 ^ 2 ) . \label{coeff_ap}\end{aligned}\ ] ] here @xmath152 . the eigenvalue equation in ( [ eveq ] ) can be solved directly to get the stable regions of the steady state @xmath97 . on the other hand , we can use the r - h criteria to have a closer look at the stable regions of the system ( [ sys1 ] ) . according the r - h criteria , the stable region corresponds to the region in which @xmath325 ) are satisfied is depicted in the above figure . the light - blue shaded region ( i ) which lies under the light - blue curve ( curve -@xmath19 ) corresponds to @xmath274 , the region under the pink curve ( curve -@xmath66 ) which is denoted by region ( ii ) corresponds to @xmath326 , green shaded region ( iii ) corresponds to @xmath327 . the regions satisfying the conditions ( ii ) and ( iii ) in ( [ rhcri ] ) are the gray shaded region ( iv ) ( the region under curve -@xmath328 ) and red shaded region ( v ) ( the region under curve -@xmath329 ) , separately . the other condition @xmath330 is satisfied everywhere in the considered region of ( @xmath76 ) . the region in which all the conditions in ( [ rhcri ] ) are satisfied is denoted by blue shaded region ( vi ) ( the region under curve -@xmath331).,width=359 ] now we consider the above criteria one by one and obtain the required stable region of the steady state . + ( i)a . @xmath274 : this condition will be satisfied when @xmath332 this simple condition promises that above the value of @xmath333 , the trivial equilibrium point can never be stable and thus oscillation can be revoked by an increase in the value of @xmath23 . the region in the @xmath334 space in which the above condition is satisfied is denoted by i in fig . [ ap_fig ] . + ( i)b . @xmath330 : when @xmath335 , if the condition ( @xmath330 ) is satisfied for @xmath52 then it will be satisfied for all values of @xmath23 . if the condition @xmath330 is not satisfied while @xmath52 then the condition will not be satisfied for any value of @xmath23 . when @xmath336 , the condition @xmath330 will be satisfied only for the values of @xmath23 given by @xmath337 , \nonumber \\ \eta&<&\frac{2}{\tilde{a}}\left[\epsilon-3 - \sqrt{\epsilon^2 - 3 \epsilon+3-\omega_1 ^ 2-\omega_2 ^ 2}\right ] . \label{ic}\end{aligned}\ ] ] this condition @xmath330 is satisfied in the whole region considered in fig . [ ap_fig ] . + ( i)c . @xmath326 : when @xmath338 , if the condition is satisfied for @xmath52 then it will continue to be satisfied for all values of @xmath23 , if the condition is not satisfied for @xmath52 , by the variation of @xmath23 also the condition remains to be unsatisfied . when @xmath339 , it will be satisfied only for the values of @xmath23 given by @xmath340 the region of @xmath334 in which the above condition is satisfied is denoted by region ( ii ) which lies under pink curve ( or curve -@xmath66 ) in fig . [ ap_fig ] . + ( i)d . @xmath327 : similar to the previous conditions , if @xmath341 , the condition will be satisfied when @xmath342 otherwise it will be satisfied for all values of @xmath23 only if the condition is satisfied for @xmath52 . the region in which the above condition is satisfied is shown in fig . [ ap_fig ] as region iii . + ( ii ) @xmath343 : we can find that @xmath344 is a cubic polynomial in @xmath23 , and the region in which the above condition is satisfied is given by gray shaded region ( iv ) . the real roots of @xmath23 satisfying the equation @xmath345 forms the boundary of the region . + ( iii ) @xmath346 : the region satisfying this condition is shown by red shaded region ( v ) , whose boundary is the solution of the quintic equation @xmath347 . from fig . [ ap_fig ] , we can find that the region satisfying all the criteria given in ( [ rhcri ] ) is the region bounded between the curves @xmath348 the region is denoted by blue shaded region vi in fig . [ ap_fig ] . in this region , the equilibrium point @xmath97 is found to be stable or ad occurs in the region . 99 g. saxena , a. prasad , and r. ramaswamy , phys . rep . * 521 * 205 ( 2012 ) . a. koseska , e. volkov , and j. kurths , phys . rep . * 531 * 173 ( 2013 ) . a. koseska , e. volkov , and j. kurths , phys . lett . * 111 * 024103 ( 2013 ) . t. banerjee and d. ghosh , phys . e * 89 * 062902 ( 2014 ) . k. konishi , phys . e * 68 * 067202 ( 2003 ) . w. zou , d. v. senthilkumar , j. duan , and j. kurths , phys . e * 90 * 032906 ( 2014 ) . f. m. atay , phys . 91 , 094101 ( 2003 ) . w. zou , d. v. senthilkumar , y. tang , y. wu , j. lu , and j. kurths , phys . e * 88 * 032916 ( 2013 ) . m. lakshmanan and d.v . senthilkumar , _ dynamics of nonlinear time - delay systems _ , ( springer , berlin , 2010 ) . a. prasad , m. dhamala , b.m . adhikari , and r. ramaswamy , phys . e * 81 * 027201 ( 2010 ) . w. zou , d. v. senthilkumar , a. koseska , and j. kurths , phys . e * 88 * 050901(r ) ( 2013 ) . c. r. hens , p. pal , s. k. bhowmick , p. k. roy , a. sen and s. k. dana , phys . e * 89 * 032901 ( 2014 ) . g. b. ermentrout and n. kopell , siam j. appl . math * 50 * 125 ( 1990 ) . r. curtu , _ physica d _ * 239 * 504 ( 2010 ) . m. dolnik and m. marek , j. phys . chem * 92 * 2452 ( 1988 ) . m. toiya , v. k. vanag , and i. r. epstein , angew . , int . ed . * 47 * , 7753 ( 2008 ) . m. heinrich , t. dahms , v. flunkert , s. w. teitsworth , e. schll , new j. phys * 12 * 113030 ( 2010 ) . d. ruwisch , m. bode , d. volkov , and e. volkov int . j. bifurcation chaos appl . * 9 * 1969 ( 1999 ) . b. gallego and p. cessi , j. clim . * 14 * 2815 ( 2001 ) . s. ching , p. l. purdon , s. vijayan , n. j. kopell , and e. n. brown , pnas , * 109 * 3095 ( 2012 ) . j. jalife , r. a. gray , g. e. morley , and j. m. davidenko , chaos * 8 * 79 ( 1998 ) . l. rubchinsky and m. sushchik , phys . e * 62 * 6440 ( 2000 ) . w. zou , d. v. senthilkumar , m. zhan , and j. kurths , phys . lett * 111 * 014101 ( 2013 ) . s. kak , circuits , systems , and signal processing * 12 * 263 ( 1993 ) . a. becskei , b. sraphin , and l. serrano , embo . j. * 20 * 2528 ( 2001 ) . s. draghici , int . j. neural syst . * 8 * 113 ( 1997 ) . v. k. chandrasekar , j. h. sheeba , and m. lakshmanan , chaos * 20 * 045106 ( 2010 ) m. g. rosenblum and a. s. pikovsky , phys . * 92 * 114102 ( 2004 ) . o. v. popovych , c. hauptmann , and p. a. tass , phys . lett . * 94 * 164102 ( 2005 ) . g. yuan , x. zhang , and z. wang , optik - int . j. light electron opt . * 125 * 1950 ( 2014 ) . g. f. franklin , j. d. powell , and a. emami - naeini , _ feedback control of dynamical systems _ ( addison - wesley , 3rd ed . , uk , 1994 ) . j. e. gough , r. gohm , and m. yanagisawa , phys . rev . a * 78 * 062104 ( 2008 ) . s. lloyd , phys . a * 62 * 022108 ( 2000 ) . s. little and p. brown , ann . n. y. acad . sci * 1265 * 9 ( 2012 ) . s. santaniello , g. fiengo , and l. glielmo , _ in proceedings of the ieee international conference on control application ( cca 2008 ) _ ( ieee , new york , 2008 ) pp . 666 - 671 . j. a. anderson , _ an introduction to neural networks _ , ( a bradford book , 3rd ed . , ( 1997 ) ) . y. kuromato , _ chemical oscillations , waves and turbulence _ , ( springer , berlin , 1984 ) . c. u. choe , t. dahms , p. hvel , and e. schll , phys e * 81 * 025205(r ) ( 2010 ) . t. aoyagi , phys . . lett . * 74 * 4075 ( 1995 ) . s. uchiyama , physica a * 391 * 2807 ( 2012 ) . n. tukhlina and m. rosenblum , j. biol . * 34 * 301 ( 2008 ) . o. dhuys , i. fischer , j. danckaert , and r. vicente , phys . e * 83 * 046223 ( 2011 ) . k. ldge , _ nonlinear laser dynamics _ , ( wiley - vch , germany , 2011 ) p. r. bandyopadhyay and a. m. hellum , sci . * 4 * 6650 ( 2014 ) . e. m. izhikevich , _ dynamical systems in neuroscience : the geometry of excitability and bursting _ , ( mit press , cambridge , 2007 ) . n. tukhlina , m. rosenblum , a. pikovsky , and j. kurths , phys . rev . e * 75 * 011918 ( 2007 ) . o. v. popovych , c. hauptmann , and p. a. tass , biol . cybern . * 95 * 69 ( 2006 ) . k. pyragas , o. v. popovych , and p. a. tass , euro . . lett . * 80 * 40002 ( 2007 ) . j. h. sheeba , v. k. chandrasekar , and m lakshmanan , phys . lett * 103 * 074101 ( 2009 ) . g. c. sethia and a. sen , phys . lett * 112 * 144101 ( 2014 ) . l. schmidt and k. krischer , phys . lett * 114 * 034101 ( 2015 ) . a. zakharova , m. kapeller , and e. schll , phys . lett . * 112 * 154101 ( 2014 ) . a. koseska , e. volkov , and j. kurths , europhys . lett . * 85 * 28002 ( 2009 ) . a. koseska , e. volkov , and j. kurths , chaos * 20 * 023132 ( 2010 ) . m. lakshmanan , s. rajasekar , _ nonlinear dynamics : integrability , chaos and patterns _ , ( springer - verlag , berlin , 2003 ) k. bar - eli , physica d * 14 * 242 ( 1985 ) . c. r. hens , o. i. olusola , p. pal , and s. k. dana , phys . e * 88 * 034902 ( 2013 ) . m. nandan , c. r. hens , p. pal , and s. k. dana , chaos * 24 * 043103 ( 2014 ) . v. resmi , g. ambika , and r. e. amritkar , phys . e * 84 * 046212 ( 2011 ) .

costas loop is a classical phase - locked loop ( pll ) based circuit for carrier recovery and signal demodulation @xcite . the pll is an automatic control system which is designed to generate an electrical signal ( voltage ) , the frequency of which is automatically tuned to the frequency of the input ( reference ) signal . various pll based circuits are widely used in modern telecommunications , computer architectures , electromechanical systems ( see , e.g. @xcite ) . nowadays among the applications of costas loop there are global positioning systems ( see , e.g. , @xcite ) , wireless communication ( see , e. g. , @xcite ) and others ( @xcite ) . dynamic behavior of the pll and the costas loop has been described extensively in the literature @xcite , and a number of key parameters has been defined that describe its lock - in and lock - out characteristics . when the pll is initially out of lock , two different types of acquisition processes can occur , either the so - called lock - in process or the so - called pull - in process . the first of those is a fast process , i.e. the acquisition takes place within at most one beat note of the difference between reference frequency @xmath0 and initial vco ( voltage controlled oscillator ) frequency @xmath1 , cf . figure [ f1 - 1 ] for signal denotations . the frequency difference for which such a fast acquisition process takes place corresponds to the lock - in range @xmath2 , and the duration of the locking process is called lock time @xmath3 . when the difference between reference and vco frequency is larger than the lock - in range but less than the pull - in range @xmath4 , a slow acquisition process occurs . the time required to get acquisition is called pull - in time @xmath5 . in case of the pll all these acquisition parameters can be approximated by characteristic parameters of the pll , i.e. from natural frequency @xmath6 and damping factor @xmath7 . to the authors knowledge such acquisition parameters have not been analytically derived for the different types of costas loops . it seems that most authors only described the static properties of the costas loop such as the derivation of the phase error in the locked state and the like . based on methods developed earlier for the pll , the authors could now derive similar expressions for all relevant acquisition parameters of the costas loop . this enables the designer to determine the lock - in and pull - in ranges , and to estimate the duration of the corresponding processes . because the systems considered are highly non linear , exact computation of such parameters is very difficult or even impossible . therefore it is necessary to introduce a number of simplifications . this implies that the obtained results are only approximations , in some cases rather crude approximations . as will be shown in the following sections there are different types of costas loops . the very first of these loops has been described by j. costas in 1956 @xcite and was primarily used to demodulate amplitude - modulated signals with suppressed carrier ( dsb - am ) . the same circuit was used later for the demodulation of bpsk signals ( binary phase shift keying ) @xcite . with the advent of qpsk ( quadrature phase shift keying ) this costas loop was extended to demodulate qpsk signals as well . these two types of costas loop operated with real signals . in case of bpsk , the input signal @xmath8 is a sine carrier that was phase modulated by a binary signal , i.e. @xmath9 where @xmath0 is the ( radian ) carrier frequency , and @xmath10 can have two different values , either @xmath11 or @xmath12 , or two arbitrary equal and opposite values @xmath13 and @xmath14 , where @xmath15 can be any value . in case of qpsk , two quadrature carriers are modulated by two modulating signals , i.e. @xmath16 where @xmath17 and @xmath18 can both have two equal and opposite values @xmath13 and @xmath14 . it is obvious that in both cases the input signal is a real quantity . in the following these two types of costas loop will be referred to as `` conventional costas loops '' . much later , costas loops have been developed that operate not on real signals , but on pre - envelope signals @xcite . these types of costas loops will be referred to as `` modified costas loop '' in the following sections . the block diagram shown in figure [ f1 - 2 ] explains how the pre - envelope signal is obtained . the real input signal @xmath8 is applied to the input of a hilbert transformer [ 2 ] , [ 5 ] . the output of the hilbert transformer @xmath19 is considered to be the imaginary part of the pre - envelope signal , i.e the pre - envelope signal is obtained from @xmath20 the costas loops operating with pre - envelope signals will be referred to as `` modified costas loops '' , cf . sections [ s4 ] and [ s5 ] . because there are different types of costas loops the acquisition parameters must be derived separately for each of these types . this will be performed in the following sections . in order to see how good or bad the obtained approximations , we will develop simulink models for different types of costas loops and compare the results of the simulation with those predicted by theory . the operation of the costas loop is considered first in the locked state with zero phase difference ( see figure [ costas_locked ] ) , hence the frequency of the carrier is identical with the frequency of the vco . by the input signal @xmath8 is the product of a transferred binary data and the harmonic carrier @xmath21 with a high frequency @xmath22 . since the costas loop is considered to be locked , the vco orthogonal output signals are synchronized with the carrier ( i.e. there is no phase difference between these signals ) . the input signal is multiplied ( multiplier block ( @xmath23 ) ) by the corresponding vco signal on the upper branch and by the vco signal , shifted by @xmath24 , on the lower branch . therefore on the multipliers outputs one has @xmath25 consider the low - pass filters ( lpf ) operation . [ as - twice - frequency ] signals components , whose frequency is about twice the carrier frequency , do not affect the synchronization of the loop ( since they are suppressed by the low - pass filters ) . [ as - lpf - initstate ] initial states of the low - pass filters do not affect the synchronization of the loop ( since for the properly designed filters , the impact of filter s initial state on its output decays exponentially with time ) . [ as - lpf - data ] the data signal @xmath10 does not affect the synchronization of the loop . assumptions [ as - twice - frequency],[as - lpf - initstate ] , and [ as - lpf - data ] together lead to the concept of so - called _ ideal low - pass filter _ , which completely eliminates all frequencies above the cutoff frequency ( assumption [ as - twice - frequency ] ) while passing those below unchanged ( assumptions [ as - lpf - initstate],[as - lpf - data ] ) . in the classic engineering theory of the costas loop it is assumed that the low - pass filters lpf are ideal low - pass filters . since in figure [ costas_locked ] the loop is in lock , i.e. the transient process is over and the synchronization is achieved , by assumptions [ as - twice - frequency],[as - lpf - initstate ] , and [ as - lpf - data ] for the outputs @xmath26 and @xmath27 of the low - pass filters lpf one has @xmath28 thus , the upper branch works as a demodulator and the lower branch works as a phase - locked loop . since after a transient process there is no phase difference , a control signal at the input of vco , which is used for vco frequency adjustment to the frequency of input carrier signal , has to be zero : @xmath29 . in the general case when the carrier frequency @xmath22 and a free - running frequency @xmath30 of the vco are different , after a transient processes the control signal at the input of vco has to be non - zero constant : @xmath31 , and a constant phase difference @xmath32 may remain . consider the costas loop before synchronization ( see figure [ costas_before_sync ] ) . here the phase difference @xmath33 varies over time , because the loop has not yet acquired lock ( frequencies or phases of the carrier and vco are different ) . in this case , using assumption [ as - twice - frequency ] , the signals @xmath34 and @xmath35 can be approximated as @xmath36 approximations depend on the phase difference of signals , i.e. two multiplier blocks ( @xmath23 ) on the upper and lower branches operate as phase detectors . the obtained expressions with @xmath37 coincide with well - known ( see , e.g. , @xcite ) phase detector characteristic of the classic pll with multiplier / mixer phase - detector for sinusoidal signals . by assumptions [ as - lpf - initstate ] and [ as - lpf - data ] the low - pass filters outputs can be approximated as @xmath38 since @xmath39 , the input of the loop filter ( lf ) is @xmath40 such an approximation is called a _ phase detector characteristic of the costas loop_. since an ideal low - pass filter is hardly realized , its use in the mathematical analysis requires additional justification . thus , the impact of the low - pass filters on the lock acquisition process must be studied rigorously . the relation between the input @xmath41 and the output @xmath42 of the loop filter has the form @xmath43 where @xmath44 is a constant matrix , the vector @xmath45 is the loop filter state , @xmath46 are constant vectors , h is a number . the filter transfer function has the form : @xmath47 the control signal @xmath42 is used to adjust the vco frequency to the frequency of the input carrier signal @xmath48 here @xmath30 is the free - running frequency of the vco and @xmath49 is the vco gain . the solution of with initial data @xmath50 ( the loop filter output for the initial state @xmath50 ) is as follows @xmath51 where @xmath52 is the impulse response of the loop filter and @xmath53 is the zero input response of the loop filter , i.e. when the input of the loop filter is zero . [ as - lf - initstate ] zero input response of loop filter @xmath54 does not affect the synchronization of the loop ( one of the reasons is that @xmath54 is an exponentially damped function for a stable matrix @xmath44 ) . consider a constant frequency of the input carrier : @xmath55 and introduce notation @xmath56 then assumption [ as - lf - initstate ] allows one to obtain the classic mathematical model of pll - based circuit in signal s phase space ( see figure [ f2 - 1b ] ) : @xmath57 for the locked state a linear pll model can be derived , which is shown in figure [ linear - pll ] . this model is useful for approximation of hold - in range . in the locked state both reference and vco frequencies are approximately the same , hence the input of the lowpass filter is a very low frequency signal . therefore the lowpass filter can be ignored when setting up the linear model of the costas loop . the linear model is made up of three blocks , the phase detector pd , the loop filter lf and the vco . in digital costas loops the vco is replaced by a dco ( digital controlled oscillator ) . this will be discussed in later sections . for these building blocks the transfer functions are now defined as follows . * phase detector ( pd ) . * in the locked state , the phase error @xmath58 is very small so by we can write @xmath59 with @xmath60 called phase detector gain . @xmath61 note that the uppercase symbols are laplace transforms of the corresponding lower case signals . * loop filter ( lf ) . * for the loop filter we choose a pi ( proportional + integral ) filter whose transfer function has the from @xmath62 this filter type is the preferred one because it offers superior performance compared with lead - lag or lag filters . * vco*. the transfer function of the vco is given by @xmath63 where @xmath49 is called vco gain . consider another non linear model of costas loop in figure [ delay - costas ] ( delay model ) . here we use assumtions 13 ( initial states of filters are omitted , double - frequency terms are completely filtered by lpfs , and @xmath10 does nt affect synchronization ) and filters lpfs are replaced by the corresponding phase - delay blocks @xmath64 . outputs of low - pass filters are @xmath65 where @xmath66 then after multiplication of @xmath26 and @xmath27 we have @xmath67 and the output @xmath42 of the loop filter satisfies the following equations @xmath68 equations of costas loop in this case are @xmath69 for lpf transfer functions @xmath70 phase shift is equal to @xmath71 . therefore is equal to the following system @xmath72 where @xmath73 . equation is hard to analyze both numerically and analytically , however this model is still useful . in the following discussion it is used to approximate pull - in range and pull - in time . for this purpose we need to simplify delay model shown in figure [ delay - costas ] . consider block diagram in figure [ f2 - 4a ] . the lowpass filters ( lpf ) used in both @xmath74 an @xmath75 branches are assumed to be first order filters having transfer function . as will be demonstrated later the corner frequency of these filters must be chosen such that the data signal i is recovered with sufficient accuracy , i.e. the corner frequency @xmath76 must be larger than the symbol rate . typically it is chosen twice the symbol rate , i.e. @xmath77 with @xmath78 symbol rate and @xmath79 . the output signal @xmath80 of the multiplier in the @xmath74 branch consists of two terms , one having the sum frequency @xmath81 and one having the difference frequency @xmath82 . because the sum frequency term will be suppressed by the lowpass filter , only the difference term is considered . the same holds true for signal @xmath83 in the @xmath75 branch . it will show up that the range of difference frequencies is markedly below the corner frequency @xmath76 of the lowpass filter . hence the filter gain will be nearly 1 for the frequencies of interest . as will also be shown later the phase at frequency @xmath84 can not be neglected . the lowpass filter is therefore represented as a delay block whose transfer function has the value @xmath85 , where @xmath86 is the phase at frequency @xmath87 . the delayed signals @xmath88 and @xmath89 are now multiplied by the product block at the right in the block diagram . consequently the output signal @xmath41 of this block will have a frequency of @xmath90 . this signal is now applied to the input of the loop filter lf . its transfer function has been defined in ( [ loop - filter - tf ] ) . the corner frequency of this filter is @xmath91 . because the phase of the loop filter can not be neglected , it is represented as a delay block characterized by @xmath92 where @xmath93 is the phase of the loop filter at frequency @xmath90 . the analysis of dynamic behavior becomes easier when the order of some blocks in figure [ f2 - 4a ] is reversed ( see figure [ f2 - 4b ] ) , i.e. when we put the multiplying block before the lowpass filter . because the frequency of signal @xmath41 in figure [ f2 - 4a ] is twice the frequency of the signals @xmath88 and @xmath89 , the phase shift created by the lowpass filter at frequency @xmath90 is now twice the phase shift at frequency @xmath87 . the lpf is therefore represented here by a delay block having transfer function @xmath94 . we can simplify the block diagram even more by concatenating the lowpass filter and loop filter blocks . the resulting block delays the phase by @xmath95 . this is shown in figure [ f2 - 4c ] . the output signal @xmath42 of this delay block now modulates the frequency generated by the vco . to compute pull - in time we need to consider costas loop model in figure [ f2 - 1b ] with averaged signals of phase detector output @xmath96 and filter output @xmath97 ( see figure [ f2 - 7 ] ) . the model is built from three blocks . the first of these is labeled `` phase - frequency detector '' . we have seen that in the locked state the output of the phase detector depends on the phase error @xmath58 . in the unlocked state , however , the average phase detector output signal @xmath98 is a function of frequency difference as will be shown in next section ( eq . ( [ 20 ] ) ) , hence it is justified to call that block `` phase - frequency detector '' . as we will recognize the pull - in process is a slow one , i.e. its frequency spectrum contains low frequencies only that are below the corner frequency @xmath99 of the loop filter , cf . ( [ loop - filter - tf ] ) . the loop filter can therefore be modeled as a simple integrator with transfer function @xmath100 therefore @xmath101 the frequency @xmath1 of the vco output signal is defined as @xmath102 where @xmath30 is the free running frequency and @xmath49 is the vco gain . now we define the instantaneous frequency difference @xmath87 as @xmath103 substituting and ( [ 24 ] ) into ( [ 22 ] ) finally yields @xmath104 consider qpsk costas loop operation ( see figure [ costas_after_sync ] ) for the sinusoidal carrier and vco in lock state for the same initial frequencies @xmath105 . by , the input qpsk signal has the form @xmath106 where @xmath107 is the transmitted data , @xmath21 and @xmath108 are sinusoidal carriers , @xmath109 phase of input signal . the outputs of the vco are @xmath110 and @xmath111 . after multiplication of vco signals and the input signal by multiplier blocks ( @xmath23 ) on the upper @xmath74 branch one has @xmath112 on the lower branch the output signal of vco is multiplied by the input signal : @xmath113 here from an engineering point of view , the high - frequency terms @xmath114 and @xmath115 are removed by ideal low - pass filters lpfs ( see assumption [ as - twice - frequency ] in previews section ) . in this case , the signals @xmath26 and @xmath27 on the upper and lower branches can be approximated as @xmath116 apart from considered case there are two possible cases : 1 ) the frequencies are different or 2 ) the frequencies are the same but there is a constant phase difference . consider costas loop before synchronization ( see figure [ qpsk_before_sync ] ) in the case when the phase of the input carrier @xmath117 and the phase of vco @xmath118 are different : @xmath119 in this case , using assumption 1 , the signals @xmath26 and @xmath27 on the upper and lower branches can be approximated as @xmath120 after the filtration , both signals , @xmath34 and @xmath35 , pass through the limiters ( sgn blocks ) . then the outputs of the limiters @xmath121 and @xmath122 are multiplied with @xmath27 and @xmath26 , respectively . by assumption 2 and corresponding formula , the difference of these signals @xmath123 can be approximated as @xmath124 with @xmath125 . here @xmath126 is a piecewise - smooth function depends on @xmath127 at the points @xmath128 . ] shown in figure [ qpsk - char ] . ] the resulting signal @xmath129 , after the filtration by the loop filter , forms the control signal @xmath42 for the vco . to derive _ mathematical model in the signal space _ describing _ physical model _ of qpsk costas loop one takes into account and : @xmath130 however equations are nonlinear and non autonomous with discontinuous right - hand side , which are extremely hard to investigate . therefore , the study of is outside of the scope of this work . to derive linear model , we consider and the corresponding figure [ qpsk - char ] . the curve looks like a `` chopped '' sine wave . the costas loop can get locked at four different values of @xmath58 , i.e. with @xmath131 , @xmath132 , or @xmath133 . to simplify the following analysis , we can define the phase error to be zero wherever the loop gets locked . moreover , in the locked state the phase error is small , so we can write @xmath134 i.e. the output signal of the adder block at the right of figure [ qpsk_before_sync ] is considered to be the phase detector output signal @xmath96 . the phase detector gain is then @xmath135 it is easily seen that the linear model for the locked state is identical with that of the costas loop for bpsk , cf . figure [ linear - pll ] . because only small frequency differences are considered here , the lowpass filters can be discarded . the transfer functions of the loop filter and of the vco are assumed to be the same as in the case of the costas loop for bpsk , hence these are given by eqs . ( [ loop - filter - tf ] ) and ( [ 6 ] ) . similar to bpsk costas loop , it is reasonable to consider delay model of qpsk costas loop ( see figure [ delay - qpsk - costas ] ) . filters lpfs are replaced by the corresponding phase - delay blocks @xmath136 . the outputs of low - pass filters are @xmath137 then @xmath41 can be approximated as @xmath138 consider the loop filter transfer function . equations of delay model of qpsk costas loop in this case are @xmath139 the non linear model of the costas loop for qpsk is developed on the basis of the non linear model we derived for the costas loop for bpsk , cf . figure [ f2 - 4c ] . here again the order of lowpass filters and the blocks shown at the right of figure [ costas_after_sync ] is reversed . this results in the model shown in figure [ f3 - 4]a . in the block labeled `` b '' the function blocks at the right of figure [ f3 - 4]a have been integrated , cf . figure [ f3 - 4]b . the output signal @xmath96 of block b is the `` chopped '' sine wave as shown in figure [ f3 - 2 ] . its fundamental frequency is 4 times the frequency difference @xmath140 . the lowpass filters and the loop filter have been concatenated in the block labeled `` lpf + lf '' at the right of figure [ f3 - 4]a . referring to figure [ costas_after_sync ] signals @xmath80 and @xmath83 are passed through lowpass filters . as in the case of the costas loop for bpsk we assume here again that the difference frequency @xmath87 is well below the corner frequency @xmath76 of the lowpass filters , hence the gain of the lowpass filters is nearly 1 at @xmath141 . because the phase shift must not be neglected , we represent the lowpass filter by a delay , i.e. its frequency response at @xmath141 is @xmath142 where @xmath86 is the phase of the lowpass filter . due to the arithmetic operations in block `` b '' ( cf . figure [ f3 - 4 ] ) the frequency of the @xmath96 is quadrupled , which implies that the phase shift at frequency @xmath143 becomes @xmath144 . the frequency response of the loop filter at @xmath145 is given by @xmath146 where @xmath93 is the phase of the loop filter at frequency @xmath147 . hence the cascade of lowpass filter and loop filter can be modeled by the transfer function @xmath148)$ ] as shown in figure [ f3 - 4]a . the block diagram of the modified costas loop for bpsk is shown in figure [ f4 - 1 ] . the input signal is given by @xmath149 where @xmath150 is initial phase . the input signal is first converted into a pre - envelope signal , as explained in section [ ss1.1 ] . the output signal of the hilbert transformer is @xmath151 = m_1(t)\sin(\omega_1t+\theta_1).\ ] ] note that because the largest frequency of the spectrum of the data signal @xmath10 is much lower than the carrier frequency @xmath0 , the hilbert transform of the product @xmath152 $ ] equals @xmath153 $ ] [ 5 ] . the pre - envelope signal is obtained now from @xmath154).\ ] ] the exponential in eqn . ( [ 56 ] ) is referred to as a `` complex carrier '' . in figure [ f4 - 1 ] complex signals are shown as double lines . the solid line represents the real part , the dotted line represents the imaginary part . to demodulate the bpsk signal , the pre - envelope signal is now multiplied with the output signal of the vco , which is here a complex carrier as well . the complex output signal of the vco is defined as @xmath155).\ ] ] in the locked state of the costas loop both frequencies @xmath0 and @xmath1 are equal , and we also have @xmath156 . hence the output signal of the multiplier @xmath157 is @xmath158)\approx m_1(t),\ ] ] i.e. the output of the multiplier is the demodulated data signal @xmath10 . to derive the linear model of this costas loop , it is assumed that @xmath159 and @xmath160 . the output signal of multiplier @xmath157 then becomes @xmath161).\ ] ] in the complex plane ] this is a phasor having magnitude @xmath162 and phase @xmath163 , as shown in figure [ f4 - 2 ] . two quantities are determined from the phase of phasor @xmath164 , i.e. the demodulated data signal i and the phase error @xmath58 . the data signal is defined as @xmath165),\ ] ] i.e. when the phasor lies in quadrants i or iv , the data signal is considered to be + 1 , and when the phasor is in quadrants ii or iii , the data signal is considered to be -1 . this means that i can be either a phasor with phase 0 or a phasor with phase @xmath132 . these two phasors are plotted as thick lines in figure [ f4 - 2 ] . the phase error @xmath58 is now given by the difference of the phases of phasor @xmath164 and phasor @xmath74 , as shown in figure [ f4 - 2 ] , i.e. @xmath58 is determined from @xmath166 the product @xmath167 is computed by multiplier @xmath168 in figure [ f4 - 2 ] . the block labeled `` complex @xmath169 mag , phase '' is used to convert the complex signal delivered by @xmath168 into magnitude and phase . the magnitude is not used in this case , but only the phase . it follows from eqn . ( [ 61 ] ) that the phase output of this block is the phase error @xmath58 , hence the blocks @xmath170 , sgn , and complex @xmath169 mag , phase represent a phase detector with gain @xmath171 . the phase output of block complex @xmath169 mag , phase is therefore labeled @xmath96 . figure [ linear - pll ] shows the complete linear model of the modified costas loop for bpsk . the transfer functions of the loop filter and vco have been defined in eqs . ( [ loop - filter - tf ] ) and ( [ 6 ] ) . note that with this type of costas loop there is no additional lowpass filter , because the multiplication of the two complex carriers ( cf . ( [ 58 ] ) ) does not create the unwanted double frequency component as found with the conventional costas loops . figure [ f5 - 1 ] shows the block diagram of the modified costas loop for qpsk . the reference signal @xmath8 is defined by @xmath172 the hilbert transformed signal is then given by @xmath173 and the pre - envelope signal then becomes @xmath174 this can be rewritten as @xmath175+j\sin[\omega_1t+\theta_1]=\\ & = ( m_1(t)+jm_2(t))\exp(j[\omega_1t+\theta_1 ] ) . \end{aligned}\ ] ] herein the term @xmath176 is complex envelope , and the term @xmath177 is complex carrier . the vco generates another complex carrier given by . the multiplier @xmath157 creates signal @xmath164 that is given by @xmath178).\ ] ] when the loop has acquired lock , @xmath179 , and @xmath180 , so we have @xmath181 hence the output of @xmath157 is the complex envelope . in the locked state , the complex envelope can take four positions , as shown in figure [ f5 - 2 ] . when there is a phase error , @xmath164 deviates from the ideal position , as demonstrated in the figure . the phase error @xmath58 then is the angle between @xmath164 and the closest of the four possible positions . when @xmath164 is in quadrant i , e.g. , phasor @xmath182 is considered the estimate of the complex envelope . when @xmath164 is in quadrant ii , the estimate of the complex envelope is @xmath183 etc . the estimates @xmath74 and @xmath75 are taken from the output of sgn blocks , cf . figure [ f5 - 1 ] . the phase error is obtained from @xmath184\ ] ] where @xmath185 is the conjugate of the complex envelope . multiplier @xmath168 delivers the product @xmath186 , and the block `` complex @xmath169 mag , phase '' is used to compute the phase of that complex quantity . note that the magnitude is not required . the blocks @xmath157 , sgn , inverter , @xmath168 , and complex @xmath169 mag , phase form a phase detector having gain @xmath171 . the phase output of block complex @xmath169 mag , phase is therefore labeled @xmath96 . in the complex plane , scaledwidth=50.0% ] figure [ linear - pll ] shows the completed linear model of the modified costas loop for qpsk , which is the same as for bpsk . the transfer functions of the loop filter and vco have been defined in eqs . ( [ loop - filter - tf ] ) and ( [ 6 ] ) . in the classic books on phase - locked loops @xcite such concepts as hold - in pull - in lock - in and other frequency ranges for which pll can achieve lock were introduced . usually in engineering literature non - rigorous definitions are given for these concepts . in the following we introduce definitions , based on rigorous discussion in @xcite . * definition of hold - in range . * the largest interval @xmath187 of frequency deviations @xmath188 , such that the loop re - achieves locked state after small perturbations of the filters state , the phases and frequencies of vco , and the input signals , is called a hold - in range . this effect is also called steady - state stability . in addition , for a frequency deviation within the hold - in range , the loop in a locked state tracks small changes in input frequency , i.e. achieves a new locked state ( _ tracking process _ ) @xcite . assume that the loop power supply is initially switched off and then at @xmath189 the power is switched on , and assume that the initial frequency difference is sufficiently large . the loop may not lock within one beat note , but the vco frequency will be slowly tuned toward the reference frequency ( acquisition process ) . this effect is also called a transient stability . the pull - in range is used to name such frequency deviations that make the acquisition process possible . * definition of pull - in range . * the largest interval @xmath190 of frequency deviations @xmath188 , such that the loop achieves locked state for any initial states ( filters and initial phase of vco ) , is called a pull - in range @xcite . the largest frequency deviation @xmath191 is called a pull - in frequency@xcite . * definition of lock - in range . * lock - in range is a largest interval of frequency deviations @xmath192 inside pull - in range , such that after an abrupt change of @xmath0 within a lock - in range the pll re - acquires lock without cycle slipping , if it is not interrupted . here @xmath2 is called a lock - in frequency@xcite . finally , our definitions give @xmath193 @xmath194 which is in agreement with the classical consideration ( see , e.g. @xcite,@xcite,@xcite,@xcite,@xcite ) . recall linear model of costas loop in phase space ( see figure [ linear - pll ] ) . by , , and we can derive the open loop transfer function of the costas loop , which is defined by the ratio @xmath195 : @xmath196 ] figure [ f2 - 2 ] shows a bode plot of the magnitude of @xmath197 . the plot is characterized by the corner frequency @xmath99 , which is defined by @xmath198 , and gain parameters @xmath60 and @xmath49 . at lower frequencies the magnitude rolls off with a slope of 40 db / decade . at frequency @xmath99 the zero of the loop filter causes the magnitude to change its slope to 20 db / decade . to get a stable system , the magnitude curve should cut the 0 db line with a slope that is markedly less than 40 db / decade . setting the parameters such that the gain is just 0 db at frequency @xmath99 provides a phase margin of 45 degrees , which assures stability [ 2 ] . from the open loop transfer function we now can calculate the closed loop transfer function defined by @xmath199 after some mathematical manipulations we get @xmath200 it is customary to represent this transfer function in normalized form , i.e. @xmath201 with the substitutions @xmath202 where @xmath6 is called natural frequency and @xmath7 is called damping factor . the linear model enables us to derive simple approximations for lock - in range @xmath2 and lock time @xmath3 . for the following analysis we assume that the loop is initially out of lock . the frequency of the input signal ( figure [ costas_before_sync ] ) is @xmath0 , and the frequency of the vco is @xmath1 . the multiplier in the i branch therefore generates an output signal consisting of a sum frequency term @xmath203 and a difference frequency term @xmath204 . the sum frequency term is removed by the lowpass filter , and the frequency of the difference term is assumed to be much below the corner frequency @xmath76 of the lowpass filter , hence the action of this filter can be neglected for this case . under this condition the phase detector output signal @xmath41 will have the form ( cf . ( [ ud ] ) and ( [ 3 ] ) ) @xmath205 with @xmath206 . @xmath41 is plotted in figure [ f2 - 3 ] , left trace . this signal passes through the loop filter . in most cases the corner frequency @xmath207 is much lower than the lock - in range , hence we can approximate its transfer function by @xmath208 let us define the gain of this filter at higher frequencies by constant @xmath209 . now the output signal of the loop filter is a sine wave having amplitude @xmath60 @xmath210 as shown by the middle trace in figure [ f2 - 3 ] . consequently the frequency of the vco will be modulated as shown in the right trace . the modulation amplitude is given by @xmath211 . in this figure the reference frequency and the initial frequency @xmath30 of the vco are plotted as horizontal lines . when @xmath0 and @xmath30 are such that the top of the sine wave just touches the @xmath0 line , the loop acquires lock suddenly , i.e. the lock - in range @xmath2 is nothing more than the modulation amplitude @xmath60 @xmath49 @xmath210 . making use of the substitutions ( [ omega zeta ] ) we finally get @xmath212 now the lock process is a damped oscillation whose frequency is the natural frequency . because the loop is assumed to lock within at most one cycle of that frequency , the lock time can be approximated by the period of the natural frequency , i.e. we have @xmath213 we have seen that all signals found in this block diagram are sine functions , i.e. all of them seem to have zero average , hence do not show any @xmath214 component . this would lead to the ( erroneous ) conclusion that a pull - in process would not be possible . in reality it will be recognized that some of the signals become asymmetrical , i.e. the duration of the positive half wave is different from the duration of the negative one . this creates a non zero @xmath214 component , and under suitable conditions acquisition can be obtained . we are therefore going to analyze the characteristics of the signals in figure [ f2 - 4c ] . , scaledwidth=50.0% ] all considered signals are plotted in figure [ f2 - 5 ] . for signals @xmath80 and @xmath83 we obtain @xmath215 @xmath216 the sum frequency terms are discarded because they are removed by the lowpass filter . the signal @xmath41 is the product of @xmath80 and @xmath83 and is given by . for small arguments @xmath217 this can be written as @xmath218 where @xmath219 . because the phase detector gain is defined by @xmath220 we have @xmath221 . next the loop filter output signal @xmath42 is plotted . its amplitude is @xmath209 @xmath222 , and its phase is delayed by @xmath95 . this signal modulates the frequency of the vco as shown in the bottom trace of figure [ f2 - 5 ] . the modulation amplitude is given by @xmath223 . in order to get an estimate for the non zero @xmath214 component of @xmath41 we will have to analyze the asymmetry of the signal waveforms . it will be shown that @xmath98 ( the average of @xmath41 ) is a function of frequency difference @xmath87 and phase @xmath224 . the analysis becomes easier when we first calculate @xmath98 for some special values for @xmath224 , i.e. for @xmath225 ; and @xmath226 . let us start with @xmath227 , cf . figure [ f2 - 6a ] . for @xmath227 , scaledwidth=50.0% ] in figure [ f2 - 6a ] the waveforms for @xmath41 and @xmath228 are shown . the asymmetry of the signals is exaggerated in this plot . during the positive half cycle ( duration @xmath229 ) the average value of vco output frequency @xmath228 is increased , which means that the average difference frequency @xmath230 is lowered . consequently the duration of the positive half wave becomes larger than half of a full cycle . during the negative half cycle ( duration @xmath231 ) , however , the average value of vco output frequency @xmath228 is decreased , which means that the average difference frequency @xmath232 is increased . consequently the duration of the negative half wave becomes less than half of a full cycle . next we are going to calculate the average frequency difference in both half cycles . the average frequency difference during half cycle @xmath229 is denoted @xmath233 , the average frequency difference during half cycle @xmath231 is denoted @xmath234 . we get @xmath235 @xmath236 for the durations @xmath229 and @xmath231 we obtain after some manipulations @xmath237 @xmath238 now the average value @xmath98 can be calculated from @xmath239 the average signal @xmath98 is seen to be inversely proportional to the frequency difference @xmath87 . because @xmath98 is positive , the instantaneous frequency @xmath228 is pulled in positive direction , i.e. versus @xmath0 , which means that a pull - in process will take place . for @xmath240 , scaledwidth=50.0% ] next we are going to analyze the dependence of @xmath98 on phase @xmath224 . let us consider now the case for @xmath240 , cf . figure [ f2 - 6b ] . we observe that in interval @xmath229 the instantaneous frequency @xmath228 is pulled in negative direction , hence the average difference frequency @xmath233 becomes larger . consequently interval @xmath229 becomes shorter . in interval @xmath231 , however , the reverse is true . here the instantaneous frequency @xmath229 the pulled in positive direction , hence the average @xmath234 is reduced , and interval @xmath231 becomes longer . the average @xmath98 is now equal and opposite to the value of @xmath98 for @xmath227 . because it is negative under this condition , a pull - in process can not take place , because the frequency of the vco is `` pulled away '' in the wrong direction . last we consider the case @xmath241 , cf . figure [ f2 - 6c ] . in the first half of interval @xmath229 the instantaneous frequency @xmath228 is lowered , but in the second half it is increased . consequently the average difference frequency @xmath233 does not change its value during @xmath229 . the same happens in interval @xmath231 . @xmath234 does not change either , and @xmath98 remains 0 . for @xmath241 , scaledwidth=50.0% ] it is easy to demonstrate that @xmath98 varies with @xmath242 , hence we have @xmath243 eq . ( [ 20 ] ) tells us that the pull - in range is finite . the pull - in range can be found as the frequency difference for which phase @xmath241 . an equation for the pull - in range will be derived in section [ ss2.4 ] . we also will have to find an equation for the pull - in time . the model shown in figure [ f2 - 7 ] will enable us to obtain a differential equation for the average frequency difference @xmath244 as a function of time . recall equations of filter output @xmath245 and frequency deviation @xmath246 eqs . ( [ 20 ] ) , ( [ 21b ] ) , and ( [ 25 ] ) enable us to compute the three variables @xmath98 , @xmath247 , and @xmath87 as a function of time . this will be demonstrated in section [ ss2.4 ] . the pull - in range can be computed using eq . ( [ 20 ] ) . lock can only be obtained when the total phase shift @xmath224 is not more negative than @xmath248 . this leads to an equation of the form @xmath249 according to eqs . ( [ loop - filter - tf ] ) and ( [ 14 ] ) @xmath86 and @xmath93 are given by @xmath250 @xmath251 with @xmath252 . hence the pull - in range @xmath4 can be computed from the transcendental equation @xmath253 to solve this equation for @xmath4 we use the addition formula for the tangent function @xmath254 and can replace @xmath255 by @xmath256 . ( [ 27 ] ) can now be rewritten as @xmath257 . when the arctg expressions on both sides of the equation are equal , their arguments must also be identical , which leads to @xmath258 hence we get for the pull - in range @xmath259 last an equation for the pull - in time @xmath5 will be derived . ( [ 20 ] ) , ( [ 21b ] ) , and ( [ 25 ] ) describe the behavior of the three building blocks in figure [ f2 - 7 ] and enable us to compute the three variables @xmath98 , @xmath247 , and @xmath87 . we only need to know the instantaneous @xmath87 vs. time , hence we eliminate @xmath98 and @xmath247 from eqs . ( [ 21b ] ) and ( [ 25 ] ) and obtain the differential equation @xmath260 this differential equation is non linear , but the variables @xmath87 and @xmath261 can be separated , which leads to an explicit solution . putting all terms containing @xmath87 to the left side and performing an integration we get @xmath262 the limits of integration are @xmath263 and @xmath2 on the left side , because the pull - in process starts with an initial frequency offset @xmath264 and ends when @xmath87 reaches the value @xmath2 , which is the lock - in range . following that instant a lock - in process will start . the integration limits on the right side are @xmath265 and @xmath5 , respectively , which means that the pull - in process has duration @xmath5 , and after that interval ( fast ) lock - in process starts . performing the integration on the left imposes some considerable problems , when we remember that @xmath266 is given by @xmath267 finding an explicit solution for the integral seems difficult if not impossible , but the @xmath268 term can be drastically simplified . when we plot @xmath266 vs. @xmath87 we observe that within the range @xmath269 the term @xmath266 is an almost perfect straight line . hence we can replace @xmath266 by @xmath270 inserting that substitution into eq . yields a rational function of @xmath87 on the left side , which is easily integrated . after some mathematical procedures we obtain for the pull - in time @xmath5 @xmath271.\ ] ] making use of eqs . and we have @xmath272 using these substitutions eq . can be rewritten as @xmath273.\ ] ] this equation is valid for initial frequency offsets in the range @xmath274 . for lower frequency offsets , a fast pull - in process will occur , and eq . should be used . an analog costas loop for bpsk shall be designed in this section . it is assumed that a binary signal is modulated onto a carrier . the carrier frequency is set to 400 khz , i.e. the costas loop will operate at a center frequency @xmath275 400000 = 2512000 rad @xmath276 . the symbol rate is assumed to be @xmath277 symbols / s . now the parameters of the loop ( such as time constants @xmath278 and @xmath279 , corner frequencies @xmath99 and @xmath76 , and gain parameters such as @xmath280 ) must be determined . ( note that these parameters have been defined in eqs . ( [ 4 ] ) , ( [ loop - filter - tf ] ) , ( [ 6 ] ) and ( [ 13 ] ) ) . the modulation amplitude is set @xmath281 . according to eq . ( [ 3 ] ) the phase detector gain is then @xmath171 . it has proven advantageous to determine the remaining parameters by using the open loop transfer function @xmath282 of the loop [ 2 ] . this is given by @xmath283 the magnitude @xmath284 ( bode diagram ) is plotted in figure [ f2 - 9 ] . the magnitude curve crosses the 0 db line at the so called transit frequency @xmath285 . it is common practice to choose @xmath285 to be about @xmath286 . here we set @xmath287 , i.e. @xmath288 rad @xmath276 . furthermore we set corner frequency @xmath289 . when doing so , the slope of the asymptotic magnitude curve changes from 40 db / decade to 20 db / decade at @xmath290 . under this condition the phase of @xmath291 is -135@xmath292 at @xmath99 . consequently the phase margin of the loop becomes 45@xmath292 , which provides sufficient stability . according to eq . ( [ loop - filter - tf ] ) @xmath279 becomes @xmath293 . next corner frequency @xmath76 will be determined . the corner frequency of the lowpass filter must be chosen such that the demodulated data signal ( i. e. the output of the lowpass filter in the i branch ) is recovered with high fidelity . to fulfill this requirement , @xmath76 should be chosen as large as possible . on the other hand , the lowpass filter should suppress the double frequency component ( here at about 800 khz ) sufficiently , which means that @xmath76 should be markedly less than @xmath294 . it s a good compromise to set corner frequency to twice the symbol rate , i.e. @xmath295 . last the remaining parameters @xmath278 and @xmath49 must be chosen . they have to be specified such that the open loop gain becomes 1 at frequency @xmath290 . according to eq . ( [ 33 ] ) we can set @xmath296 because 2 parameters are still undetermined , one of those can be chosen arbitrarily , hence we set @xmath297 . finally from ( [ 34 ] ) we get @xmath298 . the design of the costas loop is completed now , and we can compute the most important loop parameters . for the natural frequency and damping factor we get from @xmath299 from the lock - in range becomes @xmath300 and from ( [ 13 ] ) the lock time becomes @xmath301 next we want to compute the pull - in range . ( [ 28 ] ) yields @xmath302 to convert the analog loop into a digital one , we first must define a suitable sampling frequency @xmath303 ( or sampling interval @xmath304 ) . to satisfy the nyquist theorem , the sampling frequency must be higher than twice the highest frequency that exists in the loop . in our case the highest frequency is found at the output of the multipliers in the @xmath74 and @xmath75 branches ( cf . figure [ costas_before_sync ] ) . the sum frequency term is about twice the center frequency , hence @xmath303 must be greater than 4 times the center frequency . a suitable choice would be @xmath305 @xmath306 mhz . next the transfer functions of the building block have to be converted into discrete transfer functions , i.e. @xmath307 . for best results it is preferable to use the bilinear @xmath308 transform . given an analog transfer function @xmath309 , this can be converted into a discrete transfer function @xmath310 by replacing @xmath311 by @xmath312 now the bilinear z transform has the property that the analog frequency range from @xmath313 is compressed to the digital frequency range from @xmath314 . to avoid undesired `` shrinking '' of the corner frequencies ( @xmath99 and @xmath76 ) , these must be `` prewarped '' accordingly , i.e. we must set @xmath315 @xmath316 where @xmath317 and @xmath318 are the prewarped corner frequencies . now we can apply the bilinear z transform to the transfer functions of the lowpass filters ( cf . ( [ 14 ] ) ) and of the loop filter ( cf . ( [ loop - filter - tf ] ) ) and get @xmath319+\left[1-\frac{2}{\omega_{3,p}t}\right]z^{-1}}{1+z^{-1}},\ ] ] @xmath320+\left[1-\frac{2}{\omega_{c , p}t}\right]z^{-1 } } { \frac{2\tau_1}{t}-\frac{2\tau_1}{t}z^{-1}}.\ ] ] because the vco is a simple integrator , we can apply the discrete @xmath308 transform of an integrator , i.e. @xmath321 the digital costas loop is ready now for implementation . a simulink model will be presented in section [ ss2.7 ] . a simulink model of a costas loop for bpsk is shown in figure [ f2 - 10 ] . a data signal is created by a random number generator at the left in the block diagram . the other blocks are self explanatory . the model is used now to check the validity of the approximations found for pull - in range and pull - in time . ( [ 28 ] ) predicts a pull - in range @xmath322 khz . the simulation revealed a pull - in range of @xmath323 khz . , which shows that the theoretical result is a rather crude approximation . a series of other simulation delivered results for the pull - in time @xmath324 . the results are listed in table 2 - 1 . [ cols="<,<,<,<",options="header " , ] table 5 - 1 . comparison of predicted and simulated results for the pull - in range the predictions come very close to the results obtained from the simulation . as demonstrated in figure [ f5 - 1 ] the phase error signal @xmath96 was obtained from the phase output of block `` complex @xmath169 mag , phase '' . the phase of the complex input signal to this block can be obtained from the @xmath325 function . this imposes no problem when a processor is available . this is the case in most digital implementations of the costas loop . as an alternative a phase error signal can also be obtained directly from the imaginary part of multiplier @xmath168 ; this is shown in figure [ f5 - 9 ] . , scaledwidth=40.0% ] it is easily seen that here @xmath96 is given by @xmath326 the blocks shown in figure [ f5 - 9 ] therefore represent a phase detector having gain @xmath327 . this must be taken in account when specifying the open loop transfer function , cf . section [ ss5.5 ] . for this design the pull - in time of the loop is given by @xmath328 this work was supported by russian science foundation ( project 14 - 21 - 00041 ) and saint - petersburg state university one needs to be cautious using model in figure [ f2 - 1b ] even for calculating hold - in range for bpsk costas . consider an example : costas loop with lead - lag loop filter @xmath329 and low - pass filters lpfs @xmath330 in locked state phase error @xmath32 satisfies @xmath331 therefore we get a bound for the hold - in range @xmath332 in order to find hold - in range we need to find poles of the closed - loop transfer function ( roots of the characteristic polynomial ) for the linearized model ( small - signal model ) of the system on figure [ costas_before_sync ] . open - loop transfer function is @xmath333 @xmath334 phase error @xmath335 corresponds to hold - in range ( see ) if all roots of the polynomial have negative real parts ( i.e. polynomial is stable ) . applying routh - hurwitz criterion to study stability of the polynomial , we get that for the following parameters @xmath336 polynomial is stable for all @xmath337 . however , if @xmath338 the following condition is necessary for stability @xmath339 then , taking into account static phase error equation , we get different hold - in ranges for different values of @xmath76 @xmath340 r. lazzari , c. parma , a. de marco , s. bittanti , enabling a flexible exchange of energy of a photovoltaic plant with the grid by means of a controlled storage system , international journal of control 88 ( 7 ) ( 2015 ) 13531365 . y. a. sidorkina , v. sizykh , b. shakhtarin , v. shevtsev , costas circuit under the action of additive harmonic interferences and wideband noise , journal of communications technology and electronics 61 ( 7 ) ( 2016 ) 807816 . f. bizzarri , a. brambilla , g. s. gajani , periodic small signal analysis of a wide class of type - ii phase locked loops through an exhaustive variational model , circuits and systems i : regular papers , ieee transactions on 59 ( 10 ) ( 2012 ) 22212231 . g. leonov , n. kuznetsov , m. yuldashev , r. yuldashev , nonlinear dynamical model of costas loop and an approach to the analysis of its stability in the large , signal processing 108 ( 2015 ) 124135 . http://dx.doi.org/10.1016/j.sigpro.2014.08.033 [ ] . r. best , n. kuznetsov , g. leonov , m. yuldashev , r. yuldashev , simulation of analog costas loop circuits , international journal of automation and computing 11 ( 6 ) ( 2014 ) 571579 , 10.1007/s11633 - 014 - 0846-x . n. kuznetsov , g. leonov , m. yuldashev , r. yuldashev , simulation of nonlinear models of modified bpsk costas loop for non sinusoidal waveforms in matlab simulink , in : 2014 6th international congress on ultra modern telecommunications and control systems and workshops ( icumt ) , vol . 2015-january , ieee , 2014 , pp . http://dx.doi.org/10.1109/icumt.2014.7002084 [ ] . n. kuznetsov , o. kuznetsova , g. leonov , p. neittaanmaki , m. yuldashev , r. yuldashev , simulation of nonlinear models of qpsk costas loop in matlab simulink , in : 2014 6th international congress on ultra modern telecommunications and control systems and workshops ( icumt ) , vol . 2015-january , ieee , 2014 , pp . n. kuznetsov , g. leonov , p. neittaanmki , s. seledzhi , m. yuldashev , r. yuldashev , nonlinear mathematical models of costas loop for general waveform of input signal , in : ieee 4th international conference on nonlinear science and complexity , nsc 2012 - proceedings , 2012 , pp . http://dx.doi.org/10.1109/nsc.2012.6304729 [ ] . g. leonov , n. kuznetsov , m. yuldahsev , r. yuldashev , analytical method for computation of phase - detector characteristic , ieee transactions on circuits and systems - ii : express briefs 59 ( 10 ) ( 2012 ) 633647 . http://dx.doi.org/10.1109/tcsii.2012.2213362 n. kuznetsov , g. leonov , m. yuldashev , r. yuldashev , analytical methods for computation of phase - detector characteristics and pll design , in : isscs 2011 - international symposium on signals , circuits and systems , proceedings , ieee , 2011 , pp . 710 . http://dx.doi.org/10.1109/isscs.2011.5978639 [ ] . n. kuznetsov , o. kuznetsova , g. leonov , p. neittaanmaki , m. yuldashev , r. yuldashev , limitations of the classical phase - locked loop analysis , proceedings - ieee international symposium on circuits and systems 2015-july ( 2015 ) 533536 . http://dx.doi.org/10.1109/iscas.2015.7168688 [ ] . r. best , n. kuznetsov , o. kuznetsova , g. leonov , m. yuldashev , r. yuldashev , a short survey on nonlinear models of the classic costas loop : rigorous derivation and limitations of the classic analysis , in : proceedings of the american control conference , ieee , 2015 , pp . 12961302 , art . 7170912 , http://arxiv.org/pdf/1505.04288v1.pdf . http://dx.doi.org/10.1109/acc.2015.7170912 [ ] . n. kuznetsov , g. leonov , m. yuldashev , r. yuldashev , rigorous mathematical definitions of the hold - in and pull - in ranges for phase - locked loops , ifac - papersonline 48 ( 11 ) ( 2015 ) 710713 . http://dx.doi.org/10.1016/j.ifacol.2015.09.272 [ ] . g. leonov , n. kuznetsov , m. yuldashev , r. yuldashev , hold - in , pull - in , and lock - in ranges of pll circuits : rigorous mathematical definitions and limitations of classical theory , ieee transactions on circuits and systems i : regular papers 62 ( 10 ) ( 2015 ) 24542464 . http://dx.doi.org/10.1109/tcsi.2015.2476295 [ ] . g. bianchi , n. kuznetsov , g. leonov , m. yuldashev , r. yuldashev , limitations of pll simulation : hidden oscillations in matlab and spice , international congress on ultra modern telecommunications and control systems and workshops ( icumt 2015 ) 2016-january ( 2016 ) 7984 . http://dx.doi.org/10.1109/icumt.2015.7382409 [ ] . n. kuznetsov , o. kuznetsova , g. leonov , s. seledzhi , m. yuldashev , r. yuldashev , bpsk costas loop : simulation of nonlinear models in matlab simulink , in : 2014 6th international congress on ultra modern telecommunications and control systems and workshops ( icumt ) , vol . 2015-january , ieee , 2014 , pp . http://dx.doi.org/10.1109/icumt.2014.7002083 [ ] . g. leonov , n. kuznetsov , hidden attractors in dynamical systems . from hidden oscillations in hilbert - kolmogorov , aizerman , and kalman problems to hidden chaotic attractors in chua circuits , international journal of bifurcation and chaos 23 ( 1 ) , art . http://dx.doi.org/10.1142/s0218127413300024 [ ] . n. kuznetsov , g. leonov , m. yuldashev , r. yuldashev , nonlinear analysis of classical phase - locked loops in signal s phase space , ifac - papersonline 47 ( 3 ) ( 2014 ) 82538258 . http://dx.doi.org/10.3182/20140824-6-za-1003.02772 [ ] . e. kudryashova , o. kuznetsova , n. kuznetsov , g. leonov , s. seledzhi , m. yuldashev , r. yuldashev , nonlinear models of bpsk costas loop , icinco 2014 - proceedings of the 11th international conference on informatics in control , automation and robotics 1 ( 2014 ) 704710 . http://dx.doi.org/10.5220/0005050707040710 [ ] . g. bianchi , n. kuznetsov , g. leonov , s. seledzhi , m. yuldashev , r. yuldashev , http://www.sciencedirect.com/science/article/pii/s2405896316312617[hidden oscillations in spice simulation of two - phase costas loop with non - linear vco ] , ifac - papersonline 49 ( 14 ) ( 2016 ) 45 50 . http://dx.doi.org/10.1016/j.ifacol.2016.07.973 [ ] . http://www.sciencedirect.com/science/article/pii/s2405896316312617

growing attention to cdse / cds core / thick shell nanocrystals ( ncs ) initially reported by two groups @xcite is stimulated by their superior optical properties superior to any other nc hetero - structures prepared up to now . the photoluminescence is never completely quenched in these ncs @xcite and the blinking is almost completely suppressed @xcite . at low temperatures these structures demonstrate suppression of non - radiative auger recombination @xcite and almost 100% photoluminescence quantum yield @xcite . as a result these nc structures demonstrate a very low threshold for optically pumped lasing @xcite . the origin of such outstanding optical properties of the cdse / cds core / thick shell nano - structures ( or giant cdse / cds ncs @xcite ) is not completely understood , but one thing is clear . the very thick shell always prevents carrier escape to the nc , no matter which mechanism is responsible for this escape : direct quantum mechanical tunneling , thermo - ionization of the ncs , or auger auto - ionization of the ncs , where two electron hole pairs were created simultaneously @xcite . the prevention of the nc ionization could explain the blinking suppression ; however , it does not explain the suppression of non - radiative auger recombination of charged excitons , biexcitons @xcite and multi - excitons @xcite . the alloying of the cdse / cds interface in giant cdse / cds ncs could be a reason for auger recombination suppression , as was suggested in several papers of the los alamos group @xcite . this explanation relies on a theoretical prediction that softening of the carrier confined potential could reduce the auger recombination rate by three orders of magnitude @xcite . indeed , growth of the thick shell in the giant cdse / cds core / shell ncs lasted for a week @xcite . the slow growth of the shell at temperatures above room temperature suggests that the strong interface diffusion takes place during the nc growth . the diffusion should lead to alloying adjacent to the surface area and softening of the confinement potential . this alloying assumption was confirmed recently by the demonstration of mixed phonon replicas in the fluorescence line narrowing ( fln ) of cdse / cds ncs with a thick shell @xcite . this argument , however , can not be used to explain the suppression of the auger recombination observed by dubertret group @xcite . the shell of their ncs is grown very fast and that practically excludes any inter - diffusion . in contrast , the fast growth of the lattice mismatched cdse / cds core / shell ncs should obviously results in some intrinsic strains . indeed , the samples exhibit a phonon spectrum of a pure cdse nc , the frequencies of which are shifted from bulk cdse frequencies due to the compressive strain within the nc @xcite . an understanding of the unusual and potentially useful properties of the giant cdse / cds core / shell ncs requires a theoretical description of the fine structure of the band edge excitons , and trions . up to now , the energy spectra of these nc quantum dots were studied only within simplified parabolic band models @xcite , which do not take into account the complex structure of the valence band and the complicated character of inter - particle coulomb interactions . the theoretical description of the energy spectra of the exciton and exciton complexes in these ncs also has some objective complications , because cdse / cds hetero - structures are not well characterized . we know that holes are always localized in cdse , and that the valence band offset is no less than 500mev . however , we do nt know the sign of the band - offset of the conduction band nor its magnitude . a small value of the cdse / cds conduction band offset suggests its complex behavior , because the band offset sign and magnitude should be sensitive to the strain and temperature @xcite . these qualitative data on the band offset suggest that an exciton created in cdse / cds ncs with a thick cds shell should look like a donor center , where a hole is strongly localized in the cdse core while an electron is attracted to this charged core by the coulomb potential . if cdse / cds is the type ii hetero - structure , the core potential in the nc pushes the electron into the cds shell . in the opposite case , if cdse / cds is the type i heterostucture , the core creates an additional well , which increases the electron localization around the hole . the temperature dependence of the cdse and cds energy gaps @xcite suggests that a decrease of temperature should favor type i. as a result , a decrease of the temperature increases the exciton binding energy and , consequently , shortens the radiative decay time of excitons , which is controlled by the electron - hole wave function overlap , and increases the dark - bright exciton splitting.@xcite due to the very strong confinement of holes , the positively charged trion which was observed in the cdse / cds ncs @xcite appears as a doubly - charged donor . the larger coulomb attraction of the electron to the core should significantly increase the electron binding energy and shorten the radiative decay time . the negatively charged exciton should look like a negatively charged donor center d@xmath0 . the binding energy of the second electron in the d@xmath0 center is usually quite small on the order of @xmath1 of the donor binding energy . this energy will be higher , however , in redthe type i cdse / cds core / shell structure , due to an additional attractive potential ( created by the cdse core ) for an electron . finally , a biexciton in these structures looks like the helium atom , with two holes localized in the core and two electrons with opposite spins occupying the same orbit . in this paper , we have calculated the binding energy and the fine structure of the exciton in cdse / cds ncs with a very thick shell as a function of the cdse / cds conduction band offset and the core radius . the calculations takes into account the complex structure of the cdse valence band and inter - particle coulomb and exchange interaction . we have also developed a general theory describing the fine structures and radiative decay time of positively and negatively charged trions . using this theory we have calculated optical spectra of trions in giant cdse / cds core / shell ncs as a function of conduction - band offset and core radius . this paper is organized as follows : in section ii and iii we describe the energy spectra and wave functions of one and two holes localized in the cdse core . in section iv we discuss the energy and the wave function of an electron in a central symmetrical potential created by localized holes and the conduction band offset . in section v we derive the wave functions and the energy of excitons and positively charged trions . in section vi we discuss the fine structure of these excitons and trions . in section vii we calculate the binding energy of the negatively - charged trions . in section viii we consider the radiative decay of all exciton complexes . finally , we discuss obtained results and compare them with available experimental data in section ix . in cdse / cds core shell nanocrystals , the valence band offset is substantially large and the holes are strongly confined in the cdse core . for the heavy ground state holes this potential can be considered as infinitely high in the first approximation . the first quantum - size level of holes in a spherical nc of a semiconductor with the degenerate @xmath2 valence subband is a @xmath3 state.@xcite this state has total angular momentum @xmath4 and it is fourfold degenerate with respect to its projection @xmath5 on the @xmath6 axis . @xcite . the wave functions of these four states can be written as@xcite @xmath7 where @xmath8 are wigner 3j - symbols , @xmath9 are spherical harmonics defined in ref . , and @xmath10 ( @xmath11 ) are the bloch functions of the fourfold degenerate valence band @xmath2 @xcite . the radial functions @xmath12 and @xmath13 corresponding to an impenetrable barrier at the cdse / cds interface with radius @xmath14 can be written in explicit form following refs . and : @xmath15 ~,\nonumber \\ r_0(r ) & = & { a\over a^{3/2}}\left [ j_0(\varphi r / a)- { j_0(\varphi ) \over j_0(\varphi\sqrt{\beta})}j_0(\varphi\sqrt{\beta}r / a ) \right]~ , \label{rfunction}\end{aligned}\ ] ] where @xmath16 and @xmath17 are spherical bessel functions , the constant a is determined by the normalization condition @xmath18=1 $ ] and @xmath19 is the ratio of light to heavy hole effective masses : @xmath20 and @xmath21 , respectively . the dimensionless parameter @xmath22 is a first root of the equation @xmath23 , and describes the energy of the lowest hole level @xcite : @xmath24 . the assymetry effects , which include the intrinsic asymmetry of the hexagonal lattice structure of the crystal s field , @xcite and the nonspherical shape of the nc @xcite , split the four - fold degenerate hole state into two two - fold degenerate states with @xmath25 and @xmath26 , respectively : e_1s_3/2 , j_z = e_1s_3/2 - ( j_z^2 - 5/4 ) , + [ splitting ] where @xmath27 where @xmath28 and @xmath29 are the intrinsic and size dependent shape contributions which can be found , for example , in ref . . before considering the positively charged excitons , which can be observed in ncs with an extra hole , let us consider the energy spectra of two holes localized in the cdse core . the application of the pauli exclusion principle to the two holes occupying the @xmath3 level is nontrivial . generally , the two holes with momentum @xmath30 according to the momentum summation rule @xmath31 could form four states with the total angular momentum @xmath32 . the wave functions of these states can generally be written as @xmath33 where @xmath34 is the projection of the total momentum @xmath35 on the @xmath6 axis . the pauli hole permutation requirement applied to the wave function described by eq . , however , allows only the two nontrivial solutions . as a result , the two holes occupying the @xmath3 level can only be in a 5-fold degenerate state with total momentum @xmath36 and a state with @xmath37.@xcite hole - hole exchange interaction splits the ground biexciton into the two states with total momentum @xmath36 and @xmath37 . the straightforward calculation of the @xmath38 coulomb interaction with the functions from eq . gives the energy of the corresponding levels : @xmath39 where @xmath40 is the band - edge energy of the two - hole state taking into account the direct coulomb repulsion only , @xmath41 is the exchange splitting and can be written in the following form @xcite @xmath42 where @xmath43 is the dielectric constant of the semiconductor and @xmath44 , @xmath45 . the direct coulomb interaction energy is given by @xmath46 in the spherical ncs , the ground two hole state has total momentum @xmath36.@xcite in the case of the cdse / cds structure with impenetrable interface , when the wave functions are described by eq . , the exchange splitting and coulomb energy can be written as @xmath47 in cdse ncs , where @xmath48 , @xmath49 @xcite and @xmath50 . the nc asymmetry , which lifts the degeneracy of the @xmath3 hole state in eq . , causes further splitting of the two hole states . the corresponding perturbation can be written as@xcite @xmath51 where @xmath52 and @xmath53 are the operators of hole momentum projection acting on the first and second hole , respectively . the resulting two - hole states are described in ref . . the ground two - hole state has angular momentum projection of @xmath54 and energy @xmath55 the first excited state has @xmath56 and energy @xmath57 , which is not affected by perturbations connected with nc asymmetry . finally , the upper state has @xmath54 and the energy @xmath58 two - hole wave functions @xmath59 ( for the energies @xmath60 ) and @xmath61 ( for the fourfold degenerate intermediate level @xmath57 ) are given by@xcite _ 0 ^ 2h , ( r_h1 , r_h2 ) & = & , + _ 1 ^ 2h ( r_h1 , r_h2)&= & ( _ 3/2^h(r_h1)_1/2^h(r_h2)- _ 1/2^h(r_h1)_3/2^h(r_h2 ) ) , + _ 2 ^ 2h ( r_h1 , r_h2)&= & ( _ 3/2^h(r_h1)_1/2^h(r_h2)- _ 1/2^h(r_h1)_3/2^h(r_h2 ) ) , [ biexfunction2 ] where b_0^ = , b_2^= , [ bs ] with @xmath62 . in section ii above we have considered holes in the cdse / cds core / shell ncs assuming that they are very strongly confined in the cdse core . in that case their wave functions are not modified by the coulomb interaction with the electron or by the hole - hole interaction , and are solely determined by the radius of the cdse core . on the contrary , the conduction band offset in such structures is known to be small and even the sign of the band offset is not confirmed . as a result , the band offset itself does not localize ( or only weakly localizes ) an electron in the cdse core . the attractive coulomb potential created by the holes plays an important role in electron localization at the center of the core / shell nc structures . we will show in the next section that the attractive coulomb potential created by holes localized in _ a spherically symmetrical _ cdse core does not have spherical symmetry . in the first approximation , however , the ground electron state is described by the wave function of @xmath63 symmetry , which can be written as _ 1/2^e(r)=r_e^z(r)y_00()u^c_1/2 , [ electronwf ] where @xmath64 are the bloch wave functions of the conduction band and @xmath65 is the electron spin projection . the radial wave function @xmath66 is described by the radial hamiltonian @xmath67 here @xmath68 is the conduction band offset ( cbo ) ( @xmath69 in type ii structures and @xmath70 in type i structures ) , @xmath71 and @xmath72 are the electron effective masses in cdse and cds , correspondingly , and @xmath73 is the effective coulomb potential of one ( @xmath74 ) or two ( @xmath75 ) holes acting on an electron . we use @xmath76 to describe the `` free '' electron problem in the case of the neutral core . the radial wave function @xmath66 is completely defined by eq . and the standard boundary conditions at the cdse / cds interface @xmath77 : @xmath78 later we will need the radial function of the `` free '' resident electron , which remains in ncs after negative trion recombination . its wave function @xmath79 can be written as @xmath80 where @xmath81 is the normalization constant , @xmath82 is the spherical hankel function,@xcite @xmath83 , @xmath84 and @xmath85 is the electron ground state energy level calculated from the bottom of the cds conduction band . this energy is described by the smallest solution of the following equation : @xmath86 where @xmath87 . it is important to note that even if cdse / cds core / thick shell ncs are type i structures , the resident electron is not localized in the cdse core with small radius @xmath14 and is spread over the entire cds shell . the electron localization occurs only if the power of a quantum well @xmath88 is larger than @xmath89 , which is a solution of the following equation : @xmath90 . for @xmath91 the electron delocalization occurs when @xmath92 , the condition which is identical to the standard expression for the critical depth of the 3d potential allowing the electron localization , @xmath93.@xcite however , the large electron effective mass in cds resulting in @xmath94 allows the localization in the cdse core with @xmath95 smaller than @xmath96 : @xmath97 . the electron wave function is strongly modified by a long range coulomb potential created by a hole or holes in the donor - like exciton or the positively charged trion in cdse / cds ncs with a small cbo . electrons are bound to the cdse core by a strong coulomb potential of a single hole or two holes , @xmath98 correspondingly , which are strongly confined by the core potential , similar to an electron in singly or doubly charged donors . this potential always leads to some binding of electrons to the cdse core , even in type ii cdse / cds core / shell structures . the attraction potential of holes is described by the distribution of the hole positive charge within the core of the radius @xmath14 . the straightforward use of the hole wave function definition in eq . gives the charge distribution of the single hole state with @xmath99 and @xmath100 correspondingly : _ 3/2(r)=|^h_3/2(r)|^2&=&(r_0(r)y_00 + 1r_2(r)y_20)^2 + 25r_2 ^ 2(r)(|y_21|^2+|y_22|^2 ) , + _ 1/2(r)=|^h_1/2(r)|^2&=&(r_0(r)y_00 - 1r_2(r)y_20)^2 + 25r_2 ^ 2(r)(|y_21|^2+|y_22|^2 ) , using the definition of spherical harmonics , @xmath101 , from @xcite we arrive to _ m(r)=14 . [ eq:13 ] the charge distribution of two _ isotropic _ holes with @xmath102 , localized in cdse core , @xmath103 can be obtained from the two - hole wave functions @xmath104 defined in eq . : _ j , j_z^2h(r)=d^3r_h1 d^3r_h2|_j , j_z^2h ( r_h1 , r_h2)|^2[(r - r_h1)+(r - r_h2 ) ] to obtain the hole charge distribution for @xmath105 one has to use the two hole wave functions described in eq . ( [ biexfunction2 ] ) . the calculations show that for the four - time degenerate hole state @xmath57 , which is characterized by the angular momentum projections @xmath106 , the charge distribution is spherical and is described as ^2h_1 , 2(r)=24[eq:17 ] for the upper , `` @xmath107 '' , and lower , `` @xmath108 '' , two hole states with @xmath109 similar calculations result ^2h,_0(r)=24[eq:18 ] in the limit @xmath110 eq . is simplified significantly : ^2h,_0(r)=24 . [ eq:19 ] comparison of eq . with eq . shows clearly that in the case @xmath111 , the lowest level is created from two heavy holes , while the uppermost level is created by two light holes . similar comparison of eq . with eq . shows that the middle level is always created by one light and one heavy hole . one can show that only the spherical part of the hole charge distribution contributes to the ground @xmath63 electron state in first order perturbation theory . for the single ( @xmath74 ) and double ( @xmath75 ) hole states this distribution can be written as ^z_0(r)=z4 the coulomb potential created by the hole spherical charge distribution can be written as v^z_coul(r)=-ze^2_0^adr^3^z_0(r)|r -r| [ eq:21 ] where @xmath43 is the dielectric constant , which is considered to be the same in both cdse and cds . equation can be rewritten as : v^z_coul(r)=-ze^2 4 . [ potentialv ] this potential outside the cdse core , @xmath112 , always has a standard coulomb orm , @xmath113 , due to the wave function normalization condition , @xmath114=1 $ ] . inside the core @xmath115 , and the potential depth @xmath116 depends only on the ratio @xmath117 of the light- to heavy - hole effective masses . in cdse where @xmath118 ; @xmath119 . we will consider only cdse / cds core / shell ncs with a very thick shell because these ncs show the most interesting and useful properties . from now on all calculations will be conducted for heterostructure with unlimited cds shell thickness ; later we will discuss the limitations of our approach . the ground state of a weakly bound electron in the donor - like exciton or positively charged trion created in cdse / cds is described by the hamiltonian of eq . with coulomb potential , @xmath120 , described by eq . . introducing dimensionless distance from the nc center , @xmath121 , and dimensionless energies @xmath122 , we obtain the dimensionless hamiltonian , @xmath123 for the radial component of the electron wave function , @xmath124 . in hamiltonian @xmath125 we use @xmath126 , @xmath127 . all energies are calculated from the bottom of the cds conduction band , and the confinement potential @xmath128 created by the conduction band offset is negative in type i , and positive in type ii , cdse / cds hetero - structures . we also introduce the unitless parameter @xmath129 that characterizes the strength of coulomb potential created by holes , where @xmath130 is the bohr radius . a central parabolic part of the hole potential described by eq . becomes important if @xmath131 is large . the wave function of the 1s ground state should be close to the wave function of the harmonic oscillator . in the opposite case of small @xmath131 , the core potential is weak and the electron moves mainly in the coulomb - like potential at large distances from the core . to describe both above - mentioned limits , we select a trial radial function in the following form : @xmath132 where @xmath133 and @xmath134 are dimensionless variation parameters , which will be found using the variational principle , and constants @xmath81,@xmath135 , and @xmath136 , which are completely determined by the interface boundary and normalization conditions . the standard boundary conditions at the cdse / cds interface ( @xmath137 ) described by eq . gives @xmath138 & x \le 1\\ b e^ { \left(1-x \right)/\zeta } \left.\displaystyle { 2 \alpha^2 \zeta \over 2 \left(\alpha ^2 -1\right ) \zeta + \alpha ^2 \mu}\right . & x > 1 ~ , \end{array } \right . \label{trial1abcd}\end{aligned}\ ] ] where @xmath135 can be found from the normalization condition @xmath139 ( see appendix ) . using eq . we can find the expectation value for the energy , @xmath140 , of one electron : |e^z_e = _ 0^ r_e^z(x)_r^z(x ) r_e^z(x ) x^2 dx [ trail1total ] to simplify the calculation we approximate the potential , @xmath141 which is inside the core ( @xmath142 ) by the polynomial @xmath143 , where @xmath144 and @xmath145 were selected to satisfy the exact potential properties at @xmath146 . outside the core ( @xmath147 ) the potential is given by @xmath148 according to eq.([potentialv ] ) . the potential and the potential derivatives should be continuous at @xmath149 , which allows us to determine @xmath150 and @xmath151 . next , we vary the coefficients @xmath152 and @xmath153 to fit the polynomial to the function @xmath154 , the form of which depends only on the parameter @xmath117 . the best fit of @xmath154 for @xmath155 was achieved using @xmath156 and @xmath157 . figure [ potential ] shows the dimensionless potential @xmath154 . the cob is introduced via @xmath158 one can see that when the band offset is positive , the electron is pushed away from the cdse core . the negative band offset added to the coulomb attractive potential stimulates electron localization . the second hole localized in the cdse core just increases the magnitude of the attractive potential by a factor of two . -0.1truecm -0.7truecm -0.1truecm using this approximation we were able to develop an analytical expression for the electron binding energy @xmath159 , via parameters @xmath133 and @xmath134 and find the absolute minimum of the total energy and a corresponding pair of @xmath134 and @xmath133 for given @xmath160 , @xmath161 , @xmath162 , and the band offset @xmath128 ( see online supplemental material ) . figure [ bindingenergy ] shows the binding energy of an electron for various core radii and band offsets , calculated for the cases when the cdse core is charged with one ( @xmath74 ) and two ( @xmath75 ) holes . the energy is calculated from the bottom of the of the cds conduction band . the size dependences are calculted for a cdse core with a 2.5 nm radius and shown for five band offsets : @xmath163 mev . one can see that in ncs with small negative or positive band - offset , this energy decreases with a radius increase , while in the case of -100 and -200 mev , it increases following the bottom of the conduction band . the decrease of the binding energy with radius increase in ncs with small negative or positive band - offset is connected with the fact that electron motion in such ncs is controlled by coulomb potential , which decreases with nc size . 0.1truecm -0.7truecm 0.1truecm figure [ bindingenergy ] shows the dependence of the electron binding energy on the band offset,@xmath158 , in ncs with core radius 1.5 , 2.5 , and 3.5 nm radius for the single and doubly charged cdse core . the band offset depends on temperature @xcite , and the shift of the exciton line can be observed in temperature dependence . the fine structure of the excitons and trions is strongly affected by the electron - hole exchange interaction , which can be written in spherical approximation as h_exch= -23a_0 ^ 3(r_e - r_h)_exch ( ) , [ eq:28 ] where,@xmath164 is the lattice constant , @xmath165 is the strength of the exchange interaction ( in energy units ) , @xmath166 are the pauli matrices , and @xmath167 are three matrices of the momentum 3/2 . the strength of the exchange interaction in cubic and hexagonal semiconductors can be determined from the bulk exciton fine structure . in cdse this constant extracted from the bulk exciton splitting @xmath168 mev . @xcite averaging eq . over electron and hole wave functions in the basis of the 8-fold degenerate exciton state @xcite , one can arrive at the following expression for the spin - spin exchange hamiltonian between an electron and a hole : h_exch^1= -_1 ( ) , where the exchange parameter @xmath169 depends strongly on the number of holes in cdse core : _ z=_exch_0^a dr r^2[r_e^z(r)]^2[r_0 ^ 2(r)+0.2r_2 ^ 2(r ) ] , [ exchange ] due to the strong effect of the core charge on the electron radial function , @xmath170 . the radial functions @xmath170 and @xmath171 are defined in eqs . and , respectively . integration in eq . goes only to the core radius @xmath14 because in our approximation the hole wave function @xmath171 vanishes at the cdse / cds interface . one can see from eq . , that if an electron is only weakly localized in the cdse core , then the parameter @xmath172 is small . this parameter is proportional to @xmath173 where @xmath174 is a localization radius of the electron . in fig . a we show the dependence of the exchange energy @xmath175 of the exciton as a function of the conduction band offset for several cdse core radii . one can see that increase of @xmath128 leads to an increase of electron localization in the core always increases @xmath175 . figure b we show the dependence of exciton exchange energy @xmath175 as a function of the cdse core radius for several ncs band offsets . for all reasonable set of radii , @xmath175 increases with the decrease of the nc radius . -0.1truecm -0.7truecm -0.1truecm the ground exciton state in spherical nc of zinc - blende semiconductors is characterized by the total exciton momentum @xmath176 . for the ground 5-fold degenerate dark exciton state @xmath177 , and for the excited 3-fold degenerate bright exciton state @xmath178 . in cubic semiconductor ncs with nonspherical shape and ncs with wurzite lattice structure these states are split into 5 excitons , each of them characterized by the projection of the total mometum on the hexagonal axis @xmath179.@xcite the respective wave functions are constructed from the hole functions of eq . and electron functions of eq . , with the radial functions @xmath180 for the electron moving in the coulomb potential @xmath181 created by one hole localized in the cdse core . the band edge exciton fine structure is completely characterized by two parameters @xmath175 and @xmath182 defined above . in spherical cubic ncs , the splitting between bright and dark excitons is given by @xmath183 . in the case corresponding to @xmath184 , the splitting between the ground dark exciton with @xmath185 and the first excited bright exciton exciton with @xmath186 is @xmath187 . the last case is most probably realized in presently grown cdse / cds ncs.@xcite in the more general case , corresponding to @xmath188 , the exciton fine structure is described by expressions which can be found in refs . , , and . the operator for the electron - hole exchange interaction in the positively charged trion can be written as h_exch(r_1)+h_exch(r_2)= -23a_0 ^ 3_exch [ eq:31 ] where @xmath189 , @xmath190 and @xmath191 , @xmath192 are coordinates and spin matrices of two holes , respectively . the averaging of eq . over radial components of the two hole and electron wave functions results in the effective electron - hole spin exchange hamiltonian h_exch= - 2_2 ( ) , j = j_1+j_2 , where @xmath193 is defined in eq . . the dependences of the trion exchange parameter @xmath193 on the core radius and the conduction band offset are shown in insets of fig . . one can see that due to stronger attraction of the electron to the cdse cores coursed by the second hole , the trion s @xmath193 are larger than those for the exciton . in the absence of the hole level splitting ( @xmath194 ) , the states of the positively charged trion are characterized by the total momentum @xmath195 and the trion wave functions can be generally found as @xmath196 where @xmath197 is the projection of the total momentum @xmath198 on the @xmath6 axis , @xmath199 is the projection of the electron spin on the @xmath6 axis , @xmath200 is the two - hole wave function described by eq . ( [ twoholes ] ) , and @xmath201 is the electron wave function described by eq . ( [ electronwf ] ) with the radial function @xmath202 for the electron moving in the coulomb potential created by two holes . the ground trion state has total momentum @xmath203 and the energy @xmath204 ; the next one has @xmath205 with the energy @xmath206 ; and , finally , the upper state has @xmath207 with the energy @xmath208 . the last trion state is not affected by electron hole exchange interaction because the electron interaction with two different holes exactly compensate each other . the splitting of the two lowest hole states @xmath57 coursed by the exchange interaction is @xmath209 . the order of two excited trion states depends on the relative strength of electron - hole or hole - hole exchange interactions . all these trion states are twofold degenerate with respect of the momentum projection @xmath197 , and their wave functions , @xmath210 , can be written ^tr_1/2,1/2&=&^2h_0,0^e_1/2 , ^tr_3/2,1/2=^2h_2,1^e_1/2 ^ 2h_2,0^e_1/2 , + ^tr_3/2,3/2&=&(2 ^ 2h_2,2^e_1/2-^2h_2,1^e_1/2 ) , ^tr_5/2,1/2=^2h_2,1^e_1/2+^2h_2,0^e_1/2 , + ^tr_5/2,3/2&= & ( ^2h_2,2^e_1/2 + 2 ^ 2h_2,1^e_1/2 ) , ^tr_5/2,5/2=^2h_2,2^e_1/2 , [ tr55 ] . when the ground state of holes is split by a crystal field or when the nc shape is asymmetric , the related perturbation can be written in the basis of the trion functions of eq . . the perturbation does not mix the the trion state with different signs of the spin projections . for the positive spin projections ( @xmath211 , and 1/2 ) the perturbation has the following form : @xmath212 a very similar matrix describes the perturbation created by nonzero @xmath213 on the states with negative angular momentum projections : @xmath214 , and -1/2 . that matrix can be obtained by replacing states @xmath215 in eq . by states with @xmath216 , where @xmath217 . the only difference is the sign of the matrix element taken between states @xmath218 and @xmath219 , which in the last case is @xmath220 . one can see from eq . , that nc asymmetry does not affect the three trion states @xmath221 , @xmath222 and @xmath223 . the energies of these states are e_5/2=e_2- 2_2 , e_3/2=e_2 + 3_2 , and their wave functions , @xmath224 for @xmath225 , and @xmath226 and @xmath227 for @xmath228 are the same as in the spherical case . the other three trion states , @xmath229 and @xmath230 can be found as solutions of the following third order equation : ( e_2-e-2_2)(e_2-e+3_2)(e_0-e)-^2(e_2-e+_2)=0 , [ trionlev ] in the case of small electron - hole and hole - hole exchange interactions @xmath231 , @xmath232 , the approximate solutions of eq . yield e_1/2^e_0^e_2 + _ exc/2 , e_1/2e_2+_2 . the wave functions of these trion states are @xmath233 and @xmath234 , respectively . in the limit of small exchange interaction , the order of the trion levels is independent of the relation between @xmath235 and @xmath193 . the fine structure of the positively charged trion is , ordered upward in energy , the ground twofold degenerate @xmath236 level , followed by the fourfold degenerate @xmath228 level , followed by the twofold degenerate @xmath230 and @xmath225 levels , and finally by the upper twofold degenerate @xmath237 level . the lowest and the uppermost trion states in the limit of large @xmath213 are formed from the lowest and uppermost two - hole states with a total zero momentum projection of two holes @xmath54 , while all intermediate states are formed from the medium fourfold degenerate two - hole level with @xmath238 . the medium level of the trion would be eightfold degenerate if one neglected the splitting caused by the electron - hole exchange interaction proportional to @xmath193 . the limit of small exchange interaction should describe quite well the cdse / cds ncs which have a core radius @xmath239 . in these ncs , the hole - hole exchange interaction , @xmath240mev and @xmath241 mev . the calculations of the electron - hole exchange interaction in fig . [ figexch ] show that @xmath193 is also much smaller than @xmath242 . let us consider now a negatively charged trion formed by the optically excited electron - hole pair and resident electron . the total energy of the negatively charged trion can be generally described as a sum of the two total energies of electrons interacting with the hole described in our case by eq . and the energy of the electron repulsion . in the case of the small conduction band offsets , the binding energy of the trion is mainly controlled by electron - electron and electron - hole coulomb interactions as is true for the negatively charge hydrogen atom . the resulting binding energy in such a case is significantly smaller than the binding energy of the exciton because the coulomb attraction between a second electron and the hole is almost exactly compensated by the coulomb repulsion between electrons . the studies of negatively charged hydrogen atoms ( see for example ref . ) demonstrate that the binding state of doubly charged atom exist only for a singlet state of two electrons occupying the two radial orbits of @xmath63 symmetry . this requirement decreases the coulomb repulsion energy between two electrons . that is why we select the two - electron wave functions of the negative trion @xmath243 as the product of the singlet spin wave function ( antysimmetrical ) and symmetrical radial function of two @xmath63-symmetry electrons in the following form : _ 2e(r_1,r_2)= ( u_1/2^c(e1)u_-1/2^c(e2)- u_-1/2^c(e1)u_1/2^c(e2 ) ) y_00(_e1)y_00(_e2)r_2e(r_1,r_2 ) , where @xmath244 is defined as r_2e(r_1,r_2)=a_2e[r_e^1(_1,_1;r_1/a)r_e^1(_2,_2;r_2/a)+r_e^1(_1,_1;r_2/a)r_e^1(_2,_2;r_1/a ) ] [ wf2e ] with @xmath245 defined in eq . and @xmath246 defined by the normalization condition @xmath247 where the repulsion energy between two electrons @xmath248 can be found as e_coul^2e = = the two - electron trial wave function defined in eq . is described by four fitting parameters , and it generally describes the electron motion on two radial orbits characterized by two different sets @xmath249 , @xmath250 and @xmath251 , @xmath252 . it is important to note that if @xmath253 and @xmath254 , the energy of two electrons in eq . is not reduced to the kinetic energy and repulsion of uncorrelated electrons occupying two different orbits . the total energy of electrons has a contribution from the interference of both orbits and depends on all four parameters . one can find the analytical expression of the expectation energy @xmath255 as a function of four parameters in the online sm . this expression developed within _ mathematica _ were used to find the binding energy of two electrons and a hole . the trion binding energy , which is the difference between the total energy of the trion and the binding energy of the exciton ( see fig . [ bindingenergy ] ) , is shown in fig . [ trionbindingenergy ] as a function of of the nc core radius and band offset . -0.1truecm -0.7truecm -0.1truecm one can see that for large negative values of @xmath128 , the cdse / cds ncs are well - defined type - i structures . the strong confinement allows the two electrons to occupy a ground state level with two opposite spins leading to a large trion binding energy ( see fig . [ figovlp ] ) . decrease of @xmath256 decreases the confinement , and the trion binding energy is mainly controlled by the electron - electron and the electron - hole coulomb interactions . the coulomb repulsion between two electrons forces them to occupy different orbits , which reduces the repulsion and leads to a small , but positive binding energy for the trion . the orbits are characterized by the two radial wave functions , @xmath257 and @xmath258 , the spatial extension of which is characterized by parameters @xmath259 and @xmath260 . the radius of these two orbits is shown in fig . [ trionbindingenergy]c and [ trionbindingenergy]d as a function of core radius and band offset . one can see that one of the orbits is mainly localized in the cdse core while the other explores the cds shell . the radius of the large orbit obviously increases with the decrease of @xmath256 for any core radius , as one can see in fig . [ trionbindingenergy]c . the cbo dependence shown in fig . [ trionbindingenergy]d is more complex , however . the radius of the external orbit is smaller in the nc with a large core radius at large cbos , but it becomes larger at small cbos . this unusual behavior is connected with the increased role that the hole coulomb attraction potential plays in electron localization in core / shell ncs with a small cbo . the optical selection rule and the relative transition probabilities of the excitons in cdse / cds core shell ncs are described by eqs . ( 25)(28 ) of ref . , where the square of the overlap integral , @xmath261 , between the electron and hole wave functions should be written in more general form @xmath262 where @xmath263 is the electron radial function described in eq . , with parameters @xmath133 and @xmath264 that maximized the exciton binding energy . in the cdse / cds core shell ncs , where the electron is not localized in the cdse core , @xmath265 is a strong function of the band offset of the conduction band and of the core radius . in figs . [ figovlp]a and [ figovlp]b we show the dependence of the exciton @xmath265 on the core radius and the band offset , respectively . -0.1truecm -0.7truecm -0.1truecm the experimentally observed increase of the the radiative decay time by a factor of two in the temperature interval from 0 to 150 k @xcite could be connected with a significant decrease of the conduction band offset between cdse and cds shell layers . the decrease of the absolute value of the negative band offset decreases @xmath265 , as one can see in fig . [ figovlp ] c , d and f. the radiative decay time of the optically active excitons is inversely proportional to @xmath265 and is directly proportional to the characteristic exciton decay time , as can be described by @xmath266 where @xmath267 is the refractive index of the media , @xmath268 is the speed of light , @xmath269 is the kane transition matrix element , @xmath270 is the light frequency , and @xmath271 is the depolarization factor , which takes into account the decrease of the magnitude of the light electric field when it penetrates into the nc . in cdse structures @xmath272 should be on the order of 1.7ns.@xcite this value was estimated using @xmath273 , @xmath274mev and @xmath275ev . the depolarization factor @xmath271 generally depends on the nc shape and structure . in giant cdse / cds ncs the depolarization factor can be estimated using @xmath276 ^ 2 $ ] for spherical cds ncs , where @xmath277 is the high frequency dielectric constant of cds and @xmath278 is the dielectric constant of the media surrounding the nc . this results in @xmath279 . the radiative decay times of the bright band - edge excitons has a complicated dependence on the exciton exchange energy @xmath175 and the hole level splitting @xmath213.@xcite the expression for the decay time is significantly simplified when @xmath280 . in this limit the radiative decay time , @xmath281 , of the first optically active exciton can be written as = k_ex2_0 [ exdecay ] this case is realized in giant cdse / cds ncs , due to a small exchange energy . from fig.[figovlp ] one can see that the overlap integral squared decreases by a factor of 23 when cbo decreases from 50 mev to 300 mev . this decrease of @xmath265 is consistent with observed changes in the radiative decay time . one can see also from fig.[figovlp]a that an increase of the core radius leads to a strong localization of an electron in the core and increases @xmath265 significantly . this in turn should shorten the radiative decay time . the fine structure of the band - edge positive charge trions in ncs made of semiconductors with four - fold degenerate or quasi - degenerate valence bands ( like cdse , cds , inas , cdte and etc . ) generally consists of four doubly degenerate states @xmath229 , @xmath225 and @xmath230 and one four - fold degenerated state @xmath228 , whose energies we described above . all of these states are optically active and can decay radiatively with different decay times . this in principle could lead to a very complex dependence of the decay time on the temperature as a result of the thermal population of the different levels within the trion fine structure . the resonant optical excitations of the positively charged trion in ncs with an extra hole and its radiative decay are determined by the probabilities of the optical transitions between one of a hole state , @xmath3 , with @xmath282 and the one of the trion states , @xmath283 [ @xmath284 ; @xmath285 ; @xmath286 ; @xmath287 , and @xmath288 . these probabilities are proportional to the following sum of the matrix element squared : t(m , tr)&= & _ c i , j=1,2 ; i j & = & 2|_m^h(r_h1 ) ( r_e - r_h2 ) | e | _ g , g^tr ( r_e , r_h1 , r_h2 ) |^2 . [ ttr ] the sum takes into account the probability of an electron recombination with two holes ( @xmath289 ) . the sum can be replaced by a factor of two because the matrix elements of the transitions with different hole states are identical , due the symmetrization of the two - hole wave function . here @xmath290 is the polarization of the emitted or absorbed light , and the momentum operator @xmath291 acts only on the valence band bloch functions . the calculation of @xmath292 with the wave functions is straightforward but cumbersome , because the hexagonal axes of a nc ensemble are randomly oriented relative to the light propagation of emission directions . first , let us consider transition probabilities @xmath292 s responsible for the radiative decay of the ground and first excited trion state . we will be considering here only the effect of linearly polarized light , which is sufficient for the description of the radiative decay . for linearly polarized light , @xmath293 in eq . can be written as = e_z_z+[e_-_++e_+_- ] [ pol ] where @xmath6 is the direction of the hexagonal axis of the nc , @xmath294 , @xmath295 , and @xmath296 and @xmath297 are the components of the polarization vector and the momentum operator , respectively , that are perpendicular to the nc hexagonal axis . in studied cdse / cds core / shell ncs , the crystal field and shape asymmetry parameter @xmath213 is much larger than the trion exchange parameter @xmath172 . as a result the trion ground state is @xmath298 . let us calculate the transition probabilities @xmath299 and @xmath300 between the trion ground state and the hole ground state @xmath301 stimulated by the linear polarized light . substituting eq . into eq . and using the trion and hole wave - functions , after some straightforward but cumbersome calculations gives the following : @xmath302 where @xmath303 is the angle between the vector polarization of light @xmath304 and the hexagonal axis of the nc , and where the trion overlap integral squared @xmath305 is determined now by the electron wave function @xmath306 of the trion , which is calculated for the coulomb potential created by two holes . this radial function is described by eq . , with parameters @xmath133 and @xmath264 that maximize the trion binding energy . the dependence of the overlap integral squared calculated for the positively charged trion is shown as functions of the cdse core radius and cbo in figs . 5 c and d. equation describes the probability of the trion ground state radiative decay into the ground hole state . there is nothing , however , that prevents the trion decay into the light hole excited states with @xmath307 . the photon emitted during such decay will be smaller ( by an energy of about @xmath213 ) than the photon emitted during the ground - to - ground transition . this energy will be absorbed by the hole . the probability of observing such a red `` shake up '' line is determined by the four independent transition probabilities : t(-1/2,-1/2 ^ -;)&=&t(1/2,1/2 ^ - ; ) = ( b_2 ^ --b_0 ^ -)^2 ^2 , + t(1/2,-1/2 ^ -;)&= & t(-1/2,1/2 ^ - ; ) = ( b_2 ^ --b_0 ^ -)^2 ^2 , which are much smaller than the probability of the observation of the main line , because @xmath308 for large @xmath213 . summing up the probabilities of the trion radiative decay into the ground and excited hole states over all light polarizations , we obtain the radiative recombination rate of the trion ground state as @xmath309 one can show that the uppermost trion excited state @xmath310 has the same radiative decay rate as the ground one ( @xmath311 ) and they are different from the decay time of the first optically active exciton state only by the overlap integrals ( see eq . ) . comparison of the overlap integral squared for excitons and positively charged trions in fig . [ figovlp ] shows that in ncs with small conduction band offsets they became much larger for the trions due to stronger attraction of electrons to the cdse core . henceforth in this paper we will use notations for the decay times , which are related to the notation of corresponding levels of the positively charged trion : @xmath312 , @xmath313 and @xmath314 and @xmath315 . we find the probability of the transitions from the excited trion states @xmath316 and @xmath317 to the hole @xmath99 state to be t(3/2,(5/2,5/2 ) ; ) = ^2 , t(3/2,(5/2,3/2 ) ; ) = ^2 . [ eq:50 ] the transition to the hole ground state has an energy of about @xmath213 larger then the main transition from the ground state . the total transition probability from the trion @xmath318 states to the hole ground @xmath319 state is described by the sum of the two probabilities : t(1/2,(5/2,3/2);)=t(-1/2,(5/2,-3/2 ) ; ) ) = ^2 . [ eq:51 ] in the same limit of large @xmath213 , transitions from the trion excited state to the hole excited state have nearly the same energy as the main transition from the ground to ground states . when we average all possible crystal orientation and sum up eqs . and , we can find the radiative recombination rate of the trion excited state @xmath228 : @xmath320 similar calculations give us the radiative rates for @xmath230 and @xmath225 excited trion states : @xmath321 the situation is significantly simplified in studied here cdse / cds core / shell ncs . due to small value of @xmath322 , the second trion level is eightfold degenerate with an average decay rate of = 14(++)= k_tr^+2_0 , which coincides with the decay rate of the uppermost and lowest trion states @xmath323 ( see eq . ) . as a result , the positive trion radiative decay time in cdse / cds core / shell ncs does not depend on thermo - population of the fine level structure and is temperature independent . the radiative decay time of the negatively charged exciton leaves a resident electron in ncs . in cdse / cds core / shell ncs , the wave function of the resident electron , @xmath324 , is very different from the electron wave functions of the negatively charged trion , which is mainly controlled by the coulomb interactions . the different size of the electron and hole wave functions significantly decreases the overlap integral , which controls the rate of trion radiative decay . the radiative decay and the resonant optical excitation of the negatively charged trion are determined by the probabilities of the optical transitions between the resident `` free '' electron state , with spin projection @xmath325 and radial wave function @xmath324 , and one of the trion states characterized by the hole momentum projection @xmath326 . these probabilities are proportional to t(s_z , m)= 2|^e_s_z(r_1 ) ( r_1 - r_e2 ) | e | _ m^h(r_h ) _ 2e(r_e1 , r_e2 ) |^2 , [ tr ] where the factor 2 is due to the sum over electron coordinates . the ground negative trion state is formed by the hole with @xmath327 . the probability from the ground state is @xmath328 here , k_tr^-=|r_1 ^ 2dr_1 r_2 ^ 2 dr_2 r_e^0(r_1)r_2e(r_1,r_2)r_0(r_2 ) |^2 , [ eq:61 ] where @xmath244 is the symmetrized two - electron wave function in eq.([wf2e ] ) , @xmath329 is the wave functions of a single hole in eq.([rfunction ] ) , and @xmath330 is the wave function of a single resident electron in eq.([freewf ] ) . figures 5e and 5f show the dependence of the overlap integral squared , @xmath331 , for the negatively charged trion on the cdse core radius , @xmath14 , and the cbo , @xmath256 . one can see from these figures that @xmath331 always increases with @xmath14 and @xmath256 . at the same time the figures show that for each @xmath14 there is critical @xmath256 and vise versa where @xmath331 vanishes . these critical points are connected with the delocalization of the resident electron into the cds shell . as we show above , the delocalization condition is described as @xmath332 . at this condition the spread of the electron wave function into the cds shell vanishes the overlap integral . taking a sum over all light polarizations in eq . , we obtain the radiative decay rate from the negative trion ground state as = . as in the case of the positively charged trion , the radiative decay time of the negatively charged trion is different from the radiative decay time of the first optically active exciton only by the overlap integral . for the negative trion excited states with @xmath307 , we have following transition probabilities : @xmath333 taking a sum of the transition probabilities over all states and light polarizations , we obtain the radiative decay rate of the negative trion with @xmath307 : = . the radiative decay time of the trion excited state , @xmath334 , is the same as one for the ground state : @xmath335 . this paper has developed a theoretical background for qualitative and quantitative analyses of multiple experimental measurements conducted in cds / cds core / thick shell ncs and cdse / cds dots - in - rods nanostructures ( ns ) . the major unusual properties of these ns are connected with the relatively small , and temperature and strain dependent conduction band offset , which provides only weak confinement of electrons in the cdse core , while the large valence band offset in these ns leads to the strong confinement of holes . as a result , the coulomb potential created by strongly confined holes plays an important role in electron confinement . the optical properties of such ns should be considered to be in the _ intermediate _ confinement regime.@xcite the calculation procedure consists of several steps : ( i ) finding the energy spectra and wave function of strongly confined holes ; ( ii ) calculation of the adiabatic coulomb potential creating by the hole charge distribution ; and , finally , ( iii ) calculation the energy spectra of the electron moving into the total potential created by the conduction band offset and by the coulomb potential . in this paper , we realized this program assuming that the thickness of cds shell is much larger than the radius of electron localization around the cdse core . the limit is satisfied in `` giant '' cdse / cds core / shell ncs grown by the los alamos @xcite and deburter@xcite groups . the theory could be extended , however , to quantitative consideration of cdse / cds dots - in - rods ns and/or to take into account the finite thickness of the cds shell . some effects of the finite cds shell can be taken into account quantitatively on the basis of the current calculations if the radius of electron localization around the cdse core , @xmath336 , is smaller than the cds layer thickness . for example , the negative trion ionization energy should be increased on the energy of the first confined level in cds ncs with the same shape and size as original cdse / cds core / shell ncs . the radiative decay of the negative charged trion also does not vanish in nc with finite shell thickness when a delocalization condition is reached . the time can be calculated by replacing @xmath324 in eq . with the wave function of the the first confined level in cds ncs . in the paper we have also derived general expressions for the radiative decay time of the band - edge trions in various ncs . the calculations show that these times are proportional to @xmath337 . the major difference in these times is connected with the square of the overlap integral between electron and hole wave functions in positively charged @xmath338 and negatively charged @xmath331 trions . in cdse / cds core / thick shell ncs where the electron exchange energy becomes smaller than the temperature , the radiative decay time of excitons is two times longer than @xmath339 given by eq . due to the equal population of the dark and bright exciton states . a detailed comparison of our calculations with experimental data on the optical properties of the cdse / cds core / shell ncs requires a knowledge of the temperature dependence of the conduction band offset . this temperature dependence suggested first in ref . is not universal , because it has a strain contribution @xcite and should depend on the sample preparation technique . nevertheless , the decrease of the radiative decay time by a factor of two observed in refs . and is consistent with an increase of the overlap integral squared with the cbo increased from 100 to 300 mev . the small splitting of the dark / bright exciton states measured in the giant cdse / cds ncs @xcite is also in qualitative agreement with our calculations . in summary , we have developed a theory of the optical properties of cdse / cds core / thick shell ncs . the calculations takes into account the complex structure of the cdse valence band and inter - particle coulomb and exchange interaction . the theory describes the fine structures and radiative decay time of excitons and positively and negatively charged trions . the results of our calculations are in qualitative agreement with experimental data available for these ncs . we believe that calculated energy spectra of negatively and positively - charged trions will explain suppression of nonradiative auger processes observed in these structures . the authors thank j. feldman for the critical reading of the manuscript . a.s . acknowledges support of the center for advanced solar photophysics ( casp ) an energy frontier research center founded by obes , os , u.s . is grateful for the financial support received from the swiss national science foundation . l. e. acknowledges support of the office of naval research and alexander - von - humboldt foundation . from the normalization condition for an electron bound to the singly or doubly charged core we obtain the following analytical expression for the normalization constant squared : @xmath340 ^ 2 \over \alpha^3 \left[q\left ( \mu , \alpha , \zeta\right ) + \sqrt{2 \pi } \exp\left(2/\alpha^2\right ) \mathrm{erf}\left(\sqrt{2}/\alpha \right ) p\left ( \mu , \alpha , \zeta\right)\right ] } \end{aligned}\ ] ] 55 b. mahler , p. spinicelli , s. buil , x. qulin , j .- p . hermier , and b. dubertret , nature materials * 7 * , 659 ( 2008 ) . y. chen , j. vela , h. htoon , j. l. casson , d. j. werder , d. a. bussian , v. i. klimov and j. a. hollingsworth , j. am . soc . * 130 * , 5026 ( 2008 ) . p. spinicelli , s. buil , x. qulin , b. mahler , b. dubertret , and j .- hermier , phys . . lett . * 102 * , 136801 ( 2009 ) . c. galland , y. ghosh , a. steinbrck , m. sykora , j. a. hollingsworth , v. i. klimov , and h. htoon , nature * 479 * 203 ( 2011 ) . h. htoon , a. v. malko , d. bussian , j. vela , y. chen , j. a. hollingsworth , and v. i. klimov , nano letters * 10 * , 2401 ( 2010 ) . y .- s . park , a. v. malko , j. vela , y. chen , y. ghosh , f. garca - santamara , j. a. hollingsworth , v. i. klimov , and h. htoon , phys . . lett . * 106 * , 187401 ( 2011 ) . f. garca - santamara , y. chen , j. vela , r. d. schaller , j. a. hollingsworth , and v. i. klimov , nano letters * 9 * , 3482 ( 2009 ) . l. efros , nature materials * 7 * , 612 ( 2008 ) . c. galland , y. ghosh , a. steinbrck , j. a. hollingsworth , h. htoon , and v. i. klimov , nature comm . * 3 * , 908 ( 2012 ) . f. garca - santamara , s. brovelli , r. viswanatha , j. a. hollingsworth , h. htoon , s. a. crooker , and v. i. klimov , nano letters * 11 * , 687 ( 2011 ) . x. y. wang , x. f. ren , k. kahen , m. a. hahn , m. rajeswaran , s. maccagnano - zacher , j. silcox , g. e. cragg , al . l. efros , and t. d. krauss , _ nature _ * 459 * , 686 ( 2009 ) . g. e. cragg and al . l. efros , nano lett . * 10 * , 3131 ( 2010 ) . v. dzhagan , b. dubertret , e. cassette , c. javaux , r. d. rodriguez , and d. r.t . zahn , to be published . g. rain , t. stferle , i. moreels , r. gomes , j. s. kamal , z. hens , and r. f. mahrt , acs nano * 5 * , 4031 ( 2011 ) . s. brovelli , r. d. schaller , s. a. crooker , f. garca - santamara , y. chen , r. viswanatha , j. a. hollingsworth , h. htoon , and v.i . klimov , nature comm . * 2 * , 280 ( 2011 ) . the landolt brnstein database , springer . i. ekimov , f. hache , m. c. schanne - klein , d. ricard , c. flytzanis , i. a. kudryavtsev , t. v. yazeva , a. v. rodina , and al . l. efros , j. opt . soc . am . b * 10 * , 100 ( 1993 ) b. l. gelmont and m. i. dyakonov , sov . phys . semiconduct . * 5 * , 2191 ( 1973 ) . a. r. edmonds , _ angular momentum in quantum mechanics _ princeton university press , princenton , new jersey , 1957 . g. l. bir and g. e. pikus , _ symmetry and strain - induced effects in semiconductors _ ( wiley , new york 1975 ) . l. efros , phys . rev . b 46 , 7448 ( 1992 ) . l. efros , m. rosen , m. kuno , m. nirmal , d. j. norris , and m. bawendi , phys . rev . b * 54 * , 4843 ( 1996 ) . l. efros and a. v. rodina , phys . b * 47 * , 10005 ( 1993 ) . a.v . rodina and al . l. efros , phys . b * 82 * , 125324 ( 2010 ) . l. efros and a. v. rodina , solid state commun . * 72 * , 645 ( 1989 ) . , eds . m.abramowitz and i. a. stegun , national bureau of standards , 1998 . l. d. landau and e. m. lifshitz , quantum mechanics ( non - relativistic theory ) pergamon press , oxford , 1977 . l. efros , _ `` fine structure and polarization properties of the band edge excitons in semiconductor nanocrystals , '' _ ch . 3 in the book _ `` semiconductor and metal nanocrystals : synthesis , electronic and optical properties , '' _ ed . v. klimov , pergamon press ( 2003 ) . h. a. bethe and e. e. salpeter , _ quantum mechanics of one- and two- electron atoms _ , springer - verlag , berlin - gttingen - heidelberg , ( 1957 ) . c. javaux , b. mahler , b. dubertret , a. shabaev , a. v. rodina , al . l. efros , d. r. yakovlev , f. liu , m. bayer , g. camps , l. biadala , s. buil , x. qulin , and j .- hermier , to be published . l. efros and a. l. efros , sov . . semicond . * 16 * , 772 ( 1982 ) .

molecular clouds at high galactic latitude represent an abundant source of gamma - rays due to interactions with cosmic rays @xcite . high latitude clouds are primarily low mass and harbor little to no active star formation ( see @xcite for a review ) . they are also often relatively isolated from localized cosmic ray acceleration sites such as supernova remnants or ob associations . thus , their gamma - ray emission should reflect the steady - state cosmic ray density and spectrum of the surrounding region in the galaxy . these clouds are all nearby with most having a distance @xmath1 pc for @xmath2 given a scale height of 150 pc @xcite . high latitude clouds therefore represent potentially pristine probes of the cosmic ray spectrum in the solar neighborhood . diffuse gamma - ray emission comes from a combination of cosmic ray interactions with the interstellar medium ( ism ) , inverse compton scattering of ambient radiation by cosmic ray electrons and positrons , and extragalactic diffuse emission . cosmic ray interactions with the ism produce primarily gev gamma - rays via proton - proton collisions which lead to the production of neutral and charged pions . neutral pions decay directly into gamma - rays while the charged pion species ultimately decay into electrons and positrons . cosmic ray leptons interact with gas to create gamma - rays mainly via bremsstrahlung emission . the ism is effectively transparent to cosmic rays , so the gamma - ray emission is sensitive to the total gas column density regardless of dust properties or gas state ( hi , hii , ) . the necessary likelihood modeling of _ fermi _ large area telescope ( lat ) data incorporates contributions to the gamma - ray emission from gas traced by 21 cm hi emission and the integrated 2.6 mm co ( ) line emission ( ) @xcite . is assumed to be directly proportional to the column density of molecular hydrogen : @xmath3 . in fact , gamma - rays have been used to calibrate @xmath4 ( e.g. , * ? ? ? * ) . however , likelihood models of gamma - ray emission based solely on the distribution of these two species exhibit significant residual gamma - ray emission @xcite . this excess gamma - ray emission is assumed to trace unseen molecular or atomic gas , which has been called dark gas . in cold molecular clouds , the dark gas mass can be a significant fraction of the total gas mass@xcite . dark gas is expected in photodissociation regions and diffuse clouds exposed to the interstellar radiation field . because co self - shields less efficiently and has a lower dissociation energy than , co dissociation occurs deeper into a cloud ( i.e. , to higher ) . therefore , between @xmath5 , co fails to trace linearly @xcite . clouds in this range of are classified as translucent " @xcite , and represent most molecular clouds at high latitude @xcite . translucent clouds tend to be smaller and less dense than giant molecular clouds , but should be much more numerous @xcite . a lingering problem with modeling gamma - rays from high latitude clouds had been that existing co maps were unavailable , sparse , or incomplete . the _ planck _ co map is the first all - sky map of the co ( ) emission line @xcite . we use this co map in a complete , flux - limited survey of high latitude molecular clouds with the _ fermi _ lat . in this paper , we describe the methods and systematic uncertainties of the survey and report on gamma - ray observations of 9 high latitude , translucent molecular clouds . every cloud has detectable gamma - ray emission consistent with maps of either co , dark gas , or both . we also report the photon index of each cloud , as it reflects the incident cosmic ray spectrum . this work presents the initial results from the full survey . for this pilot survey , we identify well - known molecular clouds with @xmath6 from previous surveys to analyze in gamma - rays . clouds at such latitudes are all nearby , which ensures the gamma - ray flux is high enough to study . large clouds at lower galactic latitude , such as orion , have already been extensively studied with the _ fermi _ lat @xcite . @xcite presented a study on the possible gamma - ray emission based on a number of co surveys @xcite . we find candidate clouds from these surveys and identify them in the _ planck _ co map . we choose several of the brightest high latitude clouds from these surveys . three other , fainter clouds are chosen to explore the low limits for gamma - ray detection . two bright clouds , g313.1 - 28.6 and g315.1 - 29.0 , are identified via visual inspection of the _ planck _ co map . these were observed in an earlier co catalog towards dark clouds @xcite and named chamaeleon - east ii @xcite , but were not part of the surveys described in @xcite , nor mapped by @xcite . in addition , we choose a region devoid of large - scale gas and dust emission by visual inspection of the _ planck _ co map and the color excess map of @xcite . this region , centered on galactic coordinates @xmath7 , is used to test the false detection rate of gamma - ray emission from co or dark gas . some relevant properties of the chosen clouds are listed in table [ tab : prop ] . the areas of the clouds are calculated from the co extent in the _ planck _ map . distances to the clouds are taken from the literature and were derived from color excess @xcite . many , but not all , of the clouds have masses estimated in prior surveys , ( e.g. , * ? ? ? the locations of the clouds relative to the solar system can be seen in figures [ fig : dist_xy ] , [ fig : dist_xz ] , and [ fig : dist_yz ] . most of the clouds lie below the galactic plane , as shown in figures [ fig : dist_xz ] and [ fig : dist_yz ] , but represent a large range of galactic longitudes seen in figure [ fig : dist_xy ] . the lat on the _ fermi gamma - ray space telescope _ is a pair - tracking telescope , sensitive to gamma - rays between 20 mev and 300 gev . the tracker is surrounded by anti - coincidence detectors to distinguish between cosmic ray and gamma - ray events . the photon localization strongly depends on the photon energy ; at 1 gev , the 68% containment radius is 08 , decreasing with energy to roughly @xmath8 at around 10 gev @xcite . we use data of the entire sky from the _ fermi _ lat between august 4 , 2008 and june 19 , 2013 . we use the _ fermi _ science tools ( v9r27p1 ) available from the _ fermi _ science support center , utilizing the p7_v6 instrument response function . when selecting the data , we consider both front and back converted photons in the source " class . we select data from a @xmath9 radius around the chosen coordinates and between 250 mev and 10 gev . these energies are chosen to maximize both source localization and photon statistics . including photons between 10 and 100 gev does not improve the significance of the detection , as will be explained in section [ sec : results ] . we exclude photons with incidence angle @xmath10 from the zenith and any time the spacecraft rocking angle exceeds 52 . these constraints remove most gamma - ray contamination coming from the earth s limb . we perform a binned likehihood analysis , selecting the data which lie inside a @xmath11 square centered on the region of interest ( roi ) center . the basic procedure for the likelihood analysis of gamma - ray data is described in @xcite . to evaluate source detection and model significance , we consider the test statistic ( @xmath12 ) , which is proportional to the difference of the log likelihoods of two different models : @xmath13 where @xmath14 and @xmath15 are the likelihoods for two models we are comparing . the @xmath12 represents the significance of model @xmath16 over model @xmath17 . to create the model , point sources are taken from the fermi 2 year catalog @xcite version 6 ( 2fgl ) . for those clouds with point sources from the 2fgl coincident with the co emission , we remove the point source from the model . sources outside of 7 from the roi center have all free parameters fixed for the fitting procedure , while closer sources only have their spectral indices fixed . after an initial fitting , weak point sources ( @xmath18 ) are removed . new point sources are identified by subtracting the best fit model from the counts map , then smoothing this gamma - ray residual map and identifying regions exceeding 3 standard deviations above the average residual . two additional components common to any model include the isotropic and galactic inverse compton emission . the isotropic emission , originating from extragalactic diffuse gamma - ray emission and misclassification of cosmic rays in the lat , is modeled by the iso_p7v6source.txt " provided by the fermi science support center . the inverse compton component uses galprop @xcite , and is added in as a data cube from the galprop input galdef file 54_77xvarh7s " @xcite . finally , we model diffuse emission arising from interstellar gas and dust . we use six different models for this analysis to determine the significance of the gamma - ray emission from each gas component . the baseline model , against which we compare all others , contains all the neutral gas elements : hi , co , and dark gas ( _ codg _ ) . to check the significance of gamma - ray emission from the molecular cloud , we compare the baseline model to one containing only hi ( _ hi _ ) . two additional models are used to check the significance of gamma - ray emission from co and dark gas individually : one model with hiand dark gas ( _ dg _ ) and another model with hiand co ( _ co _ ) . to check whether the gamma - ray emission comes from an extended source or a point source , the co and dark gas templates in the model are replaced with a point source located at the peak of the co emission ( _ ps _ ) . finally , we test for any background point sources by adding one at the peak co emission in addition to hi , co , and dark gas ( _ codgps _ ) . in the cases where the roi has confused 2fgl sources ( see table [ tab : ptsrces ] ) , we add the specific 2fgl source into the model instead of adding a separate source . for neutral atomic gas , we separate the hi data from the lab survey @xcite into two templates . in the first , we integrate over the velocity axis between @xmath19 to account for nearby , ambient gas associated with the cloud . the second template uses the rest of the hi data to account for background hi gas . we calculate the hi column density , @xmath20(hi ) , using a spin temperature @xmath21 k , consistent with previous gamma - ray studies @xcite : @xmath22 where @xmath23 @xmath24 k@xmath25 , and @xmath26 is the background temperature at 1.4 ghz . the @xmath20(hi ) map is in a local coordinate system centered on the clouds and is linearly interpolated to @xmath27 per pixel from the all sky map . for the template , we use maps taken from _ planck _ with a @xmath28 cut . we use the type 2 " map , which combines data from different frequency channels to improve overall signal - to - noise by removing contaminating signals from cmb , dust , and free - free emission , though @xmath29co contamination remains . the @xmath29co contribution is compensated for by dividing by a factor of 1.16 @xcite . in order to make a dark gas template , we remove a linear combination of @xmath20(hi ) and from the visual extinction ( ) adapted from the map of color excess , , from the @xcite dust model , assuming = @xmath30 , where we use the standard @xmath31 here @xcite . this residual map , hereafter referred to as , is estimated as : @xmath32 the parameters @xmath33 and @xmath34 are varied until the best fit combination of @xmath20(hi ) and is found @xcite . both the and maps are use the same coordinates as the @xmath20(hi ) map and are interpolated from the healpix maps provided by _ planck_. the four gas templates used for mbm 12 are shown in figure [ fig : mbm12temp ] . these templates are all made 3 larger on each side than the region selected to account for possible gamma - ray photons leaking into the selected region due to the large point - spread function . finally , the _ fermi _ bubbles @xcite represent an additional background source of gamma - ray emission for one cloud in this study , mbm 36 . we add a constant component with a power law spectrum across the mbm 36 roi @xcite to account for this emission . every model component adds linearly to the total gamma - ray intensity @xcite : @xmath35 where @xmath36 is the inverse compton contribution to the gamma - ray intensity , @xmath37 is the isotropic contribution , and @xmath38 is the contribution due to point sources . the fit parameters are @xmath39 and @xmath40 , which are normalization coefficients , and @xmath41 , and @xmath42 which are the emissivities of the respective gas templates in units of photon flux per h - atom , per k , and per magnitude of residual extinction , respectively . to perform the fitting , we used the python likelihood analysis tools provided by the fssc . using gtlike , we find an approximate solution with the ` drmnfb ' optimizer , and then refine the solution by fixing point sources with @xmath43 , removing any source with @xmath44 , and refitting with the ` newminuit ' optimizer . we find diffuse gamma - ray emission from molecular gas to high significance in all 9 regions studied . the control region has no diffuse gamma - ray emission from co or dark gas . all clouds show extended gamma - ray emission , and all four 2fgl associations listed in table [ tab : ptsrces ] are identified as diffuse gamma - ray emission from the clouds themselves . only mbm 12 shows evidence of a background agn . mbm 12 , given its provisional classification as a blazar in the 2fgl and potential association with a radio source , goes through additional analyses discussed in a later subsection . table [ tab : tsvals ] lists the @xmath12 values for each model compared to the baseline model _ codg_. each model was fit separately to the data . the significance of gamma - ray emission from both co and dark gas is given as @xmath45 by comparing model _ codg _ to _ hi_. the significance of gamma - ray emission from the dark gas template is given as @xmath46 by comparing model _ codg _ to _ co_. the significance of gamma - rays from co - emitting gas is given as @xmath47 by comparing model _ codg _ to model _ dg_. determining whether any gamma - ray emission comes from an extended source is given as @xmath48 by comparing model _ codg _ to model _ ps_. finally , we test for contributions due to a background point source , presumably an agn , given as @xmath49 , by comparing model _ codg _ with model _ codgps_. a @xmath50 indicates a significant difference between the two models tested . in the control region , @xmath51 , indicating that we do not detect gamma - rays from gas traced by co or dark gas . for the rois around the clouds mbm 04 and dir 071 - 43 there is no evidence for significant gamma - ray emission from gas traced by co , as shown by a low @xmath52 . these rois show evidence for gamma - ray emission dominated by dark gas . conversely , the rois for mbm 55 and mbm 32 show significant gamma - ray emission from gas traced by co but not by dark gas ; the @xmath52 is high while the @xmath53 is low . the remainings rois show evidence for emission from both gas traced by co and dark gas to different degrees . for example , the roi around mbm 12 show significantly more gamma - ray emission from gas traced by co as opposed to dark gas , while the roi around mbm 36 shows more gamma - ray emission from dark gas . since this sample contains clouds dominated by gamma - ray emission from co and some dominated by emission from dark gas , the translucent clouds may represent the transition to fully molecular clouds . for every roi , @xmath48 is large enough to conclude that the gamma - ray emission is not coming from a single point source . a @xmath54 indicates the presence of a background agn . only mbm 12 , discussed in the following section , shows significant evidence for a background source . thus , out of the four point sources given in table [ tab : ptsrces ] , we confirm the existence of one background gamma - ray point source and attribute the other three to diffuse gamma - ray emission from molecular clouds . we also compare the baseline model to one utilizing the _ fermi _ standard galactic diffuse model for pass 7 . however , only one cloud in our survey had co data from @xcite included in the diffuse model ( mbm 12 ) . therefore our models , which include co , should more accurately reflect the gas origin of the gamma - ray emission . further , as will be discussed in section [ sec : uncertain ] , the dust alone results in a poor fit to the data . over the entire energy range , when our gas templates all assume a single power law spectrum , the standard diffuse model has a higher likelihood than any of the models we fit ( @xmath55 ) . the higher likelihood of the _ fermi _ galactic diffuse model is a result of it being fit in many small energy bins , rather than assuming a single power law over the entire energy range , as we do . a single power law does not capture the curvature in the spectrum . therefore , we tested a log - parabola and a broken power law spectrum for the hi template and find the fits strongly favored the broken power law spectrum with a @xmath12 of at least 200 over the single power law . the _ fermi _ galactic diffuse model still fit better with a @xmath56 . however , for mbm 12 , we have enough data to fit several energy bins , more closely replicating the _ fermi _ galactic diffuse model . in every energy bin , our model fits the data as well or better than the standard galactic diffuse model . thus , we validate our model based on the _ planck _ co , lab hi , and _ planck _ dust data . for the broken power law spectrum , the average energy break occurs at @xmath57 gev and the power law index changes from @xmath58 to @xmath59 . the break at around 1 gev is expected from cosmic ray proton interactions with the cloud and the power law index above the break energy agrees with predictions for gamma - ray emission from a cloud located far from any cosmic ray acceleration site @xcite . changing the spectral shape from a single power law to broken power law results in a 30% increase in the hi emissivity but does not affect the other gas emissivities . we leave the co and spectra as single power laws for simplicity because hi is the dominant source of gamma - rays across the roi . changing the spectrum of the hi co , and templates to a broken power law resulted in no improvement ( @xmath60 for six additional free parameters ) over the model where only the hi had a broken power law spectrum . mbm 12 is the best studied high latitude , translucent molecular cloud ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? it has the highest peak @xmath61 of the sample and was tentatively identified as a blazar with a confused " designation in the fermi 2-year catalog ( 2fgl j0257.9 + 2025c ) . it is also coincident with a radio source from the 3@xmath62 mit - green bank radio survey ( mg3 j025805 + 2029 ) . mbm 12 , therefore , merits further analysis . figure [ fig : mbm12temp ] shows the distribution of the gas in the region around mbm 12 . figure [ fig : mbm12cmap ] shows the total number of gamma - rays detected in the region around mbm 12 over the time period analyzed . a 2fgl source , as well as the coincident radio source , lie near the centroid of the molecular cloud . to establish the molecular cloud origin of the gamma - rays over an agn source , we perform both a spatial and a variability test . the results of the spatial tests are given as @xmath63 and @xmath64 . mbm 12 , after removing 2fgl j0257.9 + 2025c , has @xmath65 which verifies the presence of the 2fgl point source . all other @xmath12 values are calculated with this point source included in the model . a @xmath66 implies that the gamma - ray emission is extended . the results from these two spatial tests argue in favor of the translucent cloud origin of the gamma - ray emission . for the variability test , we compare the gamma - ray flux in 15-week time bins to the flux over the entire 58 month period of the survey . as mbm 12 lies far from possible cosmic ray acceleration sites , the cosmic ray source is dominated by the mean steady - state diffusion of cosmic rays through the galaxy , and should be constant . as a result , significant variability in the gamma - ray flux from mbm 12 would indicate a background agn , which are the most common gamma - ray sources at high latitude @xcite and typically exhibit variability @xcite . the lightcurve of mbm 12 is given in figure [ fig : mbm12lc ] and is the sum of the gamma - ray fluxes from hi , co , and dark gas . by eye , they all appear to have constant emission over the observed period . time bins 4 and 5 as well as 10 and 11 were combined to minimize convergence errors . the dashed line corresponds to the integrated flux found over the entire time period , and the shaded region is the associated uncertainty . with 13 time bins , we find @xmath67 when comparing the flux to a constant , implying that the gamma - ray flux from mbm 12 is consistent with zero variability . the tests show that we both detect diffuse gamma - ray emission from mbm 12 and confirm the existence of 2fgl j0257.9 + 2025c , associated with mg3 j025805 + 2029 . because we can recover point sources that have been intentionally removed from the model , we gain confidence that we can detect other new sources not included in the 2fgl and disentangle them from diffuse cloud emission . as the brightest source in the survey , we also use mbm 12 to test the validity of our chosen energy range . data analyzed in this region between 10 to 100 gev shows no significant emission from molecular gas , with a @xmath68 . additionally , the hi emissivity between 250 mev and 100 gev differs by only 5% from the hi emissivity between 250 mev and 10 gev . this difference is well within the systematic uncertainty . extending the analysis to 100 gev , therefore does not significantly change the emissivity . we are ultimately trying to probe for small effects , such as possible gradients in the cosmic ray flux in the solar neighborhood , so it is important to identify and characterize as many sources of uncertainty as possible . systematic errors introduced by the lat instrument are estimated to be around 10% @xcite . they arise primarily from uncertainties in the instrument response function , the energy determination of the photons , and the effective area of the lat . the remaining uncertainties are derived from the results of the likelihood analysis . non - local hi , co , and dark gas each have two free parameters associated with them and local hi has four free parameters . in addition , the inverse compton and isotropic emission templates include a normalization term , bringing the total number of template free parameters to 12 . additional , compounded , free parameters are used to create the maps and are discussed below . point sources add additional free parameters and models with a lot of free parameters tend to converge poorly or force a parameter to one of its limits , which skews the error calculation . therefore , our analysis procedure removes weak point sources . we test models with different initial values and find that when the models converge , the resulting gamma - ray fluxes and photon indices typically vary by less than 10% . we take the combined uncertainty due to the lat systematics and the likelihood variations to be @xmath6915% . the uncertainties in the emissivities depend on the detailed model inputs . the lab survey measured the radiation roughly @xmath70 to @xmath71 around the 21 cm line with a sensitivity of 0.09 k @xcite . the high velocity gas is all very far away while the low velocity gas is much closer , so we perform a velocity cut of @xmath72 on the data to separate nearby gas from far . most of the hi emission in our clouds lies inside of this cutoff . the velocity cutoff is a fairly small source of error . generally , the higher velocity gas contributes at most 10% more column density . high velocity gas , given its much farther distance , is expected to contribute little to the total gamma - ray flux . when the higher velocity hi is significantly detected , the effect on the hi flux of the cloud is @xmath73 . in addition , the hi template suffers from uncertainty due to the assumed spin temperature . spin temperatures likely change across an roi and even across an individual cloud @xcite . we tested four spin temperatures : @xmath74 k , 125 k , 400 k , and @xmath75 . the gamma - ray flux from hi changed less than 7% while the emissivity decreased by 15% with increasing @xmath76 over the entire range of @xmath76 values . surrounding the molecular cloud may be a shell of optically thick hi with a very low spin temperature , that also contributes to the dark gas phenomenon @xcite . any unaccounted for hi due to a lower spin temperature than the 125 k used should be captured in the map . this should not affect the hi emissivity or gamma - ray flux by more than 5% , as the hi in the cloud accounts for less than 5% of the total hi emission across the roi . the main uncertainties from the co template comes from the 2@xmath77 cutoff used to remove noise , and the @xmath29co contamination of the _ planck _ data @xcite . we tested a 1@xmath77 cutoff as well and saw emissivities and fluxes systematically lower by around 5% . this is expected since the co covers a larger area in the 1@xmath77 template versus the 2@xmath77 template . yet the distribution of co does not change much ; most of the difference between the 1@xmath77 and 2@xmath77 templates is noise and distributed roughly uniformly around the map . nearly the same flux is being emitted from a larger amount of co , therefore the emissivity decreases accordingly . the map is a linear combination of the lab hi map , the _ planck _ co map , and the @xcite color excess map , where the uncertainties for the color excess are generally less than 8% , with an average across each roi of less than 4% . the uncertainty in the color excess dominates the errors over those of the hi and co maps . the fluxes and emissivities for co and are insensitive to the change in spin temperature . even between the two extremes , @xmath74 k and @xmath78 , the template changes by less than 2% on average , which is much smaller than the uncertainties of the color excess map . additional uncertainties in the map arise in @xmath79 . while overall variations in @xmath30 will not affect the hi emissivity , co and dark gas emissivities may be affected if @xmath30 varies across the roi . @xmath31 in diffuse regions @xcite but increases to @xmath80 in some molecular clouds @xcite . a constant @xmath30 may underpredict the extinction values in molecular clouds by as much as 30% . the uncertaintiy in the emissivity is not clear due to the relationship between and . to study this effect , other dust templates may be more approprate . for example , a properly scaled @xmath81 map is recommended for regions of higher @xcite instead of the map . this test was performed in the chamaeleon region in @xcite . changing the dust map necessarily affects the map , and therefore also the fitted and subsequently calculated quantities . we test both dust maps in mbm 12 . comparing both models yields a @xmath82 . therefore neither nor @xmath81 are preferred over the other . the emissivity changed by up to 30% while the hi and co emissivities changed by less than 5% . we quantify the effect of the dust template more thoroughly in the full survey and defer that discussion to future work . we argue that with only 9 clouds in our current sample , a 30% change in dark gas emissivity does not significantly affect the results discussed in section [ sec : disc ] . finally , we also tested a model using the dust map as the sole gas tracer . in principle , the extinction should trace all gas species and thus should provide a model with significantly fewer free parameters . the dust model , however , does not reproduce the gamma - ray data . the @xmath12 between model _ codg _ and the model with dust alone is 175 ; the dust model fits significantly worse than our baseline model , and exhibits large , structured residuals in regions with co and dark gas . this may be due to varying dust properties @xcite , a lack of sensitivity in dense regions , a limited range of applicablility , or a nonlinear response . this validates our combination of gas templates . in sum , there a is 15% systematic uncertainty in the gamma - ray flux from both the lat instrument and the likelihood analysis along with an additional uncertainties in the emissivities due to the choice of model of @xmath83 , @xmath84 . for this study , we adopt @xmath85 . from these observations , we may derive the gamma - ray emissivity , the gamma - ray photon flux per h - atom , for the brightest clouds within 270 pc of the sun between 250 mev and 10 gev . the fit treats the energy range as a single bin . table [ tab : fit ] lists the emissivities for hi , co , and the dark gas , while table [ tab : fit_iciso ] lists the remaining model parameters excluding point source normalizations . the hi emissivities are all generally consistent with each other , and close to the average emissivity of @xmath86 photons sr@xmath25 h - atom@xmath25 . this result is consistent with the emissivities , all between 250 mev and 10 gev , found in the cygnus region , @xmath87 @xcite , and that found in the region around the cepheus / polaris flare , @xmath88 @xcite . the chamaeleon region has been analyzed twice and the hi emissivity above 250 mev was found to be @xmath89@xcite and @xmath90@xcite and the hi emissivity around r coronae australis was found to be @xmath91 @xcite . all values are less than 2@xmath77 from our results . the emissivities for the molecular gas tracers , the co and dark gas , vary more significantly . the co component has an average emissivity of @xmath92 mev 10 gev@xmath93 photons s@xmath25 sr@xmath25 ( k ) @xmath25 and the dark gas component has an average emissivity of @xmath94 mev 10 gev@xmath95 photons sr@xmath25 mag@xmath25 , taking into account statistical errors . the co emissivity was found to be @xmath96 in the cygnus region while the clouds chamaeleon , r coronae australis , and cepheus / polaris flare have @xmath97 , @xmath98 , and @xmath99 photons sr@xmath25 ( k ) @xmath25 , respectively @xcite . the dark gas emissivity for all four regions was @xmath100 @xcite and @xmath101 , @xmath102 , and @xmath103 photons sr@xmath25 mag@xmath25 @xcite , respectively . our results are entirely consistent with the previously found values . a change in the gamma - ray emissivity indicates a change in the incident cosmic ray flux . on the right of figure [ fig : rvqhi ] , we plot the hi emissivity and see no variation . mbm 12 has a high value , though systematic uncertainties place it within the 2@xmath77 range of the average hi emissivity . a constant hi emissivity is in agreement with the conclusions of @xcite and the predictions of galprop models @xcite . @xcite looked for gradients in @xmath104 and compared it to predictions from galprop . they measure a 10% decrease in @xmath104 from the gould belt to the local arm , a distance of about 1 kpc . the maximum galactocentric distance between any of the clouds in this survey covers about 0.25 kpc , so we might expect to see a 2.5% change in @xmath104 across our sample , which is well within our quoted uncertainties . additionally , we report the gamma - ray photon index from the co to isolate the molecular cloud . the indices from co and dark gas components are identical within the uncertainties . this index is similar to the average of the two power law indices in the broken power law spectrum of hi . the indices of the clouds are listed in table [ tab : fit ] and are taken from the co fit . where co is not detected , the index is taken from the dark gas fit . the left side of figure [ fig : rvqhi ] shows the power law index for every cloud . they all lie close to the best fit value @xmath105 with no evidence of a variation . the lack of detected variations implies a uniform cosmic ray flux incident on each cloud verifying that there are no cosmic ray sources near the clouds studied . the farthest a cloud can be from a supernova remnant and still receive a cosmic ray excess is 100 - 200 pc for a @xmath106 year old supernova remnant @xcite . we trace molecular gas with two components , and , and so we require two conversion factors to estimate the column density of . is not entirely molecular gas , however ; some fraction is atomic hydrogen @xcite . the molecular gas column density can therefore be written as : @xmath107 where @xmath108 is the molecular fraction of the dark gas . this expression is proportional to the traditional x - factor , @xmath4 , where co is assumed to trace all the . analysis of gamma - rays alone does not determine what fraction of dark gas is , but we can put upper and lower limits on @xmath109 : @xmath110 assuming a constant cosmic ray flux and that cosmic rays penetrate the entire cloud , every proton should be subject to the same number of cosmic ray interactions . this assumption was verified in the chamaeleon clouds in @xcite , and implies @xmath111 . the gamma - ray emission is proportional to the number of molecules , but we use in our analysis : @xmath112 . this leads to the relationship @xmath113 . a similar argument leads to the relationship @xmath114q_\mathrm{{h\,{\small i}\,}})$ ] , where the factor @xmath115 $ ] arises because a fraction of the gas represented by molecular hydrogen which has two protons . from the emissivities of mbm 12 in different energy bins given in table [ tab : fit_energy ] , we plot @xmath116 versus @xmath117 in each energy bin in figure [ fig : mbm12covhi ] . we find a linear relationship between the two emissivities and plot the best fit line . similarly , we plot @xmath118 versus @xmath119 in figure [ fig : mbm12dgvhi ] . again , we see a linear relationship , which gives us confidence that the cosmic ray flux at mbm 12 is constant and penetrates through the entire cloud . for mbm 12 , the slope of the best fit line in figure [ fig : mbm12covhi ] is @xmath120 ( k ) @xmath25 , which agrees with the value in table [ tab : calc ] obtained as a ratio of the emissivities from table [ tab : fit ] . the slope of the best fit line in figure [ fig : mbm12dgvhi ] is @xmath121 mag@xmath25 , roughly @xmath122 from the value in table [ tab : calc ] . most clouds are not detected significantly enough to make a spectrum , so we can not verify the linear relationship in each case . mbm 12 has the highest of any of the clouds in this study , so we conclude cosmic rays penetrate through every cloud in this study . that our clouds are far from cosmic ray acceleration sources and the hi emissivity is constant across them gives us some confidence that the cosmic ray flux is constant across an entire molecular cloud . we therefore assume @xmath113 and @xmath114q_\mathrm{{h\,{\small i}\,}})$ ] for every cloud in this study . the average value for among the clouds is ( @xmath123 ) ( k ) @xmath25 . this is consistent with previous gamma - ray studies of nearby molecular clouds @xcite , where @xmath124 ( k ) @xmath25 . this value is lower than that found for high latitude clouds of @xmath125 ( k ) @xmath25 @xcite . our for cham - east ii is consistent with a recent analysis of the chamaeleon cloud complex @xcite which finds @xmath126 k . however , is always higher than as seen in table [ tab : calc ] , with an average of ( @xmath127 ) . as long as any fraction of represents , the combination of the two will increase @xmath4 . assuming @xmath128 , we report the average @xmath4 of each cloud in table [ tab : calc ] . the average among all the clouds is @xmath129 ( k ) @xmath25 with large cloud - to - cloud variations . this result agrees with previous studies of @xmath4 at high latitudes @xcite , suggesting the dark gas in these previous studies is primarily molecular . our average @xmath4 is higher than the average value found in perseus @xcite , which may suggest that the molecular fraction of the dark gas in perseus is significantly lower than 1 or cosmic rays do not penetrate deeply into the co - bright regions of the cloud . our @xmath4 for every cloud is consistent with that found in the high latitude cloud mbm 40 which found an average @xmath130 ( k ) @xmath25 @xcite . figure [ fig : xco_dist ] shows @xmath4 as a function of galactocentric distance . the solid line in the figure shows one of the more extreme variations of @xmath4 considered near the solar neighborhood @xcite . as with the gamma - ray spectrum , we find no evidence for overall variation in @xmath4 over this small extent of 270 pc . figure [ fig : xco_comp ] shows @xmath4 across the co - bright part of mbm 12 , assuming @xmath128 uniformly across the entire cloud . this places an upper limit on the value of @xmath4 . @xmath109 appears to increase towards the edge of the co - emitting region where the total extinction drops by as much as an order of magnitude . the middle of the cloud shows a low @xmath109 , where the total extinction increases due to increased gas density . a higher density supports the transition to fully molecular gas . this transition reduces the fraction of dark gas in the cloud . thus , the departure of @xmath4 from in the co - bright part of the cloud may reflect the transition of the cloud from atomic to molecular gas or it may reflect a level of small - scale clumpiness in the cloud , supporting some interior dissociation along the line of sight . if @xmath108 decreases to zero toward the edge of the co - bright part of the cloud , we can potentially recover a constant @xmath4 across the cloud , though not expected to occur @xcite . as can be seen by visual inspection of the co and dark gas maps in figure [ fig : mbm12temp ] , dark gas extends beyond the co boundaries as expected from numerical studies @xcite . in these regions , we can only place limits on @xmath4 . therefore , we do not attempt to estimate @xmath4 beyond the co - bright boundary of the molecular cloud . finally , as an initial evaluation of the dust model , we replace the _ map in mbm 12 with the _ planck _ @xmath81 dust opacity map , scaled by the given ratio : @xmath131 @xcite . in mbm 12 , changing the dust model greatly affects . however @xmath109 changes by @xmath132 ( k ) @xmath25 , which is smaller than the variations of @xmath109 . @xmath109 is thus moderately insensitive to large variations in due to suppression by the ratio @xmath133 , which averages between 0.01 0.10 mag ( k ) @xmath25 . it is worth noting that , while the distribution of @xmath4 seen in figure [ fig : xco_comp ] depends on the dust tracer used , the magnitude of the variation of @xmath109 is larger than any uncertainty of @xmath109 . we will quantify the uncertainty due to changing the dust tracer in future work . we study the gamma - ray emission from nine high latitude , translucent molecular clouds and find the gamma - ray spectrum and emissivity from the gas in the clouds . all nine _ planck_-selected co clouds were significantly detected , showing extended emission associated with molecular gas . we estimate the systematic uncertainties associated with modeling the gamma - ray emission with hi , co , and dark gas templates . the hi emissivity does not vary across the regions in the sample . for some clouds , the gamma - ray emission is dominated by the co - emitting gas , while for other clouds the co - dark gas dominates the gamma - ray emission . ranges from 0.2 ( k ) @xmath25 to 1.6 ( k ) @xmath25 with large uncertainties , and spans a similarly large range around @xmath134 mag@xmath25 . in order to compare these x - factors to the traditional conversion factor between and and are added together assuming the molecular fraction of the dark gas is 1 . the results are consistent with previous studies and suggest no change in cosmic ray flux across @xmath135 pc region around the solar system . the combination of and may also explain the low values of @xmath4 found in @xcite compared with estimates using other methods @xcite . however , the choice of dark gas tracer should be made carefully . the authors thank the _ fermi _ lat team for their support , particularly at the 2013 _ fermi _ summer school and through the _ fermi _ science support center website , as well as the anonymous referee for insightful comments . this work was supported in part by the nasa new york space grant consortium based at cornell university ( # nnx10ai94h ) and by grant # 63388 - 00 41 from the professional staff congress of the city university of new york . tadp acknowledges support from nsf grant ast-1153335 . ade , p. a. r. and the _ planck _ collaboration . 2014 , 571 , 13 ade , p. a. r. and the _ fermi _ and _ planck _ collaborations . 2014 , submitted , arxiv:1409.3268 aharonian , f. a. 2004 , very high energy cosmic gamma radiation : a crucial window on the extreme universe ( river edge , nj : world scientific publishing ) gaussian , with co contours at @xmath136 and @xmath137 k . the cyan circles mark the positions of significant 2fgl point sources , and the black @xmath138 marks the position of a radio source behind mbm 12 . the counts map covers a smaller area than the gas templates in figure [ fig : mbm12temp ] . ] lcccccc mbm 36 & 3.9 & 35.7 & @xmath140 & @xmath141 & @xmath142 dir 071 - 43 & 71.6 & -42.8 & @xmath143 & @xmath144 & 1.35@xmath145 mbm 55 & 88.5 & -41.2 & @xmath146 & @xmath147 & 21.23@xmath148 mbm 02 & 108.3 & -51.9 & @xmath149 & @xmath150 & 1.77@xmath151 mbm 04 & 132.9 & -45.6 & @xmath152 & @xmath153 & 1.16@xmath154 mbm 32 & 147.6 & 40.7 & @xmath155 & @xmath156 & 2.36@xmath157 mbm 12 & 159.1 & -34.3 & @xmath158 & @xmath159 & 10.24@xmath160 mbm 20 & 210.9 & -36.5 & @xmath161 & @xmath162 & 2.48@xmath163cham - east ii & 312.9 & -28.6 & @xmath164 & @xmath165 & 12.45@xmath166control & 250.0 & 30.0 & @xmath167 & & @xmath168 [ tab : prop ] lcccccc mbm 12 & 2fgl j0257.9 + 2025c & -2.19 & 25.55 & blazar & 6 mbm 36 & 2fgl j1553.5 - 0324@xmath169 & -2.22 & 15.85 & & g313.1 - 28.6 & 2fgl j1942.7 - 8049c & -2.40 & 32.5 & & 2 , 6 g315.1 - 29.0 & 2fgl j1925.7 - 7836c & -2.31 & 14.1 & & 1 , 6 [ tab : ptsrces ] l c c c c c c mbm 36 & 595 & 140 & 92 & 314 & 9dir 071 - 43 & 76 & 47 & 2 & 76 & 7mbm 55 & 341 & 0 & 273 & 318 & -2mbm 02 & 62 & 10 & 17 & 41 & 1mbm 04 & 80 & 47 & 18 & 80 & 18mbm 32 & 166 & 9 & 76 & 122 & 6mbm 12 & 691 & 37 & 346 & 308 & 96mbm 20 & 457 & 43 & 63 & 124 & 0cham - east ii & 832 & 93 & 453 & 624 & 16control & 1 & 0 & -1 & & [ tab : tsvals ] lcccc mbm 36 & @xmath170 & @xmath171 & @xmath172 & @xmath173dir 071 - 43 & 5.9@xmath174 & @xmath175 & @xmath176 & @xmath177mbm 55 & @xmath178 & @xmath179 & @xmath180 & @xmath181 mbm 02 & @xmath182 & @xmath183 & @xmath184 & @xmath185mbm 04 & @xmath186 & & @xmath187 & @xmath188mbm 32 & @xmath189 & @xmath190 & @xmath191 & @xmath192 mbm 12 & @xmath193 & @xmath179 & @xmath194 & @xmath195 mbm 20 & @xmath196 & @xmath197 & @xmath198 & @xmath199 cham - east ii & @xmath200 & @xmath201 & @xmath202 & @xmath203 control & @xmath204 & & & @xmath205 [ tab : fit ] lccc mbm 36 & & @xmath206 & @xmath207 dir 071 - 43 & & @xmath208 & @xmath209 mbm 55 & @xmath210 & & @xmath211 mbm 02 & & @xmath212 & @xmath213 mbm 04 & & @xmath214 & @xmath215 mbm 32 & @xmath216 & @xmath217 & @xmath218 mbm 12 & & & @xmath207 mbm 20 & & @xmath219 & @xmath215 cham - east ii & & & @xmath211control & @xmath220 & & @xmath215 [ tab : fit_iciso ] lcccc 250 400 & @xmath221 & @xmath222 & @xmath223 & @xmath224400 630 & @xmath225 & @xmath226 & @xmath227 & @xmath228630 1000 & @xmath229 & @xmath230 & @xmath231 & @xmath2321000 1580 & @xmath233 & @xmath234 & @xmath235 & @xmath2151580 2510 & @xmath236 & @xmath237 & @xmath238 & 2510 3980 & @xmath239 & @xmath240 & @xmath241 & @xmath2423980 10000 & @xmath243 & @xmath244 & @xmath244 & @xmath245 [ tab : fit_energy ] lccccc mbm 36 & @xmath246 & @xmath247 & @xmath248 dir 071 - 43 & @xmath249 & @xmath250 & @xmath251 mbm 55 & @xmath252 & @xmath253 & @xmath254 mbm 02 & @xmath255 & @xmath256 & @xmath257 mbm 04 & & @xmath258 & mbm 32 & @xmath259 & @xmath260 & @xmath261 mbm 12 & @xmath262 & @xmath263 & @xmath264 mbm 20 & @xmath265 & @xmath266 & @xmath267 cham - east ii & @xmath268 & @xmath269 & @xmath270 [ tab : calc ]

virtual knots and virtual links are a generalisation of classical knots and classical links . they were introduced by l. kauffman . they can be seen as a combinatorial interpretation of copies of @xmath0 embedded in thickened surfaces with genus @xmath1 instead of copies of @xmath0 embedded in @xmath2 . virtual link diagrams are four valent graph embedded into @xmath3 together with a decoration of every vertex . the decorations are either _ overcrossings _ @xmath4 , _ undercrossings _ @xmath5 or _ virtual crossings _ , latter are marked with a circle . a virtual link is an equivalence class of virtual link diagrams modulo the _ generalised reidemeister moves _ in figure [ figure0 - 1 ] and isotopies . the notion of _ oriented _ virtual link diagrams and an _ oriented _ virtual links are equivalent , but latter modulo _ oriented _ generalised reidemeister moves and isotopies . we call the moves rm1 , rm2 and rm3 the _ classical moves _ , the moves vrm1 , vrm2 and vrm3 the _ virtual moves _ and the move mrm the _ mixed move_. in this paper we are also interested in _ virtual tangle diagrams _ and _ virtual tangles_. the first ones are graphs embedded in a disk @xmath6 such that every vertex is either one valent or four valent . the four valent vertices are , like before , labeled with an _ overcrossing _ @xmath4 , an _ undercrossing _ @xmath5 or a _ virtual crossing_. the one valent vertices are part of the boundary of @xmath6 and are called _ boundary points_. we call a virtual tangle diagram with @xmath7 one valent vertices a virtual tangle diagram with _ @xmath7-boundary points_. a virtual tangle with @xmath7-boundary points is an equivalence class of virtual tangle diagrams with @xmath7-boundary points modulo the generalised reidemeister moves and boundary preserving isotopies . we note that every of the moves of figure [ figure0 - 1 ] can be seen as a virtual tangle diagram . moreover , examples are pictured in section [ v - tangles ] , e.g. in figure [ figure1 - 1 ] . the notions of _ oriented _ virtual tangle diagrams and an _ oriented _ virtual tangles can be defined analogous , but latter modulo _ oriented _ generalised reidemeister moves and boundary preserving isotopies . we use the short hand notation c- and v- for everything that starts with classical or virtual , e.g. c - knot means classical knot and v - crossing means virtual crossing . if the reader is unfamiliar with the notion v - link or v - tangle , we refer to some introduction papers of l. kauffman and v. manturov , e.g. @xcite and @xcite and the references therein . a well - known invariant of v - links is the _ virtual jones polynomial _ , i.e. an extension of the jones polynomial from c - links to v - links . moreover , m. khovanov published in @xcite a _ categorification _ of the jones polynomial , i.e. a @xmath8-graded chain complex whose homotopy type is a c - link invariant and whose graded euler characteristic is the jones polynomial . this categorification is called _ khovanov ( co-)homology_. the khovanov complex is a highly studied invariant of c - links . d. bar - natan gave a cobordism based exposition of this categorification in @xcite . d. bar - natans picture has many nice properties , e.g. it can be done _ local _ , i.e. it works for c - tangles as well . it is a natural question if the khovanov complex extents from c - links to v - links . the answer is positive , i.e. there is more then one _ categorification of the virtual jones polynomial _ , one over @xmath9 given by v. manturov in @xcite , one over @xmath10 given by v. manturov in @xcite and one over @xmath10 using cobordisms in the spirit of d. bar - natan in @xcite . the case of v - links is more difficult ( combinatorial ) then the classical case . that is a reason why in @xcite the d. bar - natan approach was not extended to v - tangles . this is the main goal of the paper . in this paper we will answer the question if the categorification extends from c - tangles to v - tangles positive , i.e. it is trivial to extend the notions over @xmath9 or in a trivial way ( by setting open saddles to be zero ) . latter has an obvious disadvantage , i.e. it is neither a `` good '' invariant of v - tangles nor can it be used to calculate bigger complexes by `` tensoring '' smaller pieces . but we give a local notion that is a strong invariant of v - tangles ( see section [ v - tangles ] ) and allows `` tensoring '' of smaller pieces ( see section [ circuit ] ) . as an application of this construction we prove in section [ leedeg ] that the degenerations of lee s variant ( see her paper @xcite ) , i.e. that the homology in her case just `` counts '' the number of components , also extents to v - links . note that latter result is quite surprising since the complex will contain lots of @xmath11-morphisms ( our main observation is that it will also contain many isomorphisms ) . we note that , given the semi - local properties of our construction , it is possible to write a `` fast computation '' algorithm in the sense of d. bar - natan in @xcite before we give a brief summary we recall the main ideas of the cobordism based construction of virtual khovanov homology in @xcite . we will assume that the reader is familiar with the notion of the classical khovanov complex , e.g. the construction of the khovanov cube based on so - called _ resolutions of the crossings_. there are many good introduction to classical khovanov homology , e.g. a nice exposition of the classical khovanov homology can be found in d. bar - natan s paper @xcite . moreover we assume that the reader is familiar with the cobordism based construction of the classical khovanonv complex of d. bar - natan in @xcite . recall that latter involves the _ local relation _ in figure [ figure0 - 2 ] . should be zero , a cobordism that contains a torus @xmath12 should be two times the cobordism without the torus and the four tubes relation . ] to maintain readability , we repress in this short summary the more technical points about the construction , e.g. the placement of the signs . for more details see @xcite . note that this technical points are a main reason for the length of the paper . the following table shows the connection between the classical and the virtual case . [ cols="^,^,^",options="header " , ] hence , a main point in the construction of the virtual khovanov complex is to say which saddles , i.e. morphisms , are orientable and which are non - orientable , how to place the decorations and how to place the signs . this is done in the following way . * every saddle either split one circle in a resolution ( orientable , called _ comultiplication _ , denoted @xmath13 . see figure [ figure1 - 2 ] - upper row , fourth morphism ) , glues two circles in a resolution ( orientable , called _ multiplication _ , denoted @xmath14 . see figure [ figure1 - 2 ] - upper row , fifth morphism ) or does not change the number of circles at all ( non - orientable , called _ mbius cobordism _ , denoted @xmath15 . see figure [ figure1 - 2 ] - upper row , last morphism ) . * every saddle can be local denoted ( up to a rotation ) by a formal symbol @xmath16 ( both smoothings are neighbourhoods of the crossing ) . the _ glueing numbers _ , i.e. the decorations , are now spread by _ choosing a formal orientation _ for the resolution . we note that the construction will not depend on the choice . * after all resolutions have an orientation , a saddle could for example be of the form @xmath17@xmath18 . this is the _ standard form _ , i.e. in this case every glueing number will be @xmath19 . * now spread the decorations as follows . every boundary component gets a @xmath19 iff the orientation is like in the standard case and a @xmath20 otherwise . the _ degenerated _ cases ( everything non - alternating ) , e.g. @xmath21@xmath18 , are the non - orientable surfaces and do not get any decorations . * there is a relation how the saddles behave if one changes the decorations . see for example figure [ figure1 - 3 ] bottom row and figure [ figure1 - 4 ] right part . because of this relations we say that @xmath14 is of type @xmath22 , @xmath13 of type @xmath23 and @xmath15 of type @xmath11 . * the signs are spread based on numbering of the circles in the resolutions and on a special x - marker for the crossings ( rather technical , see @xcite for details or definition [ defin - saddlestructure ] for a short introduction ) . a x - marker is for example pictured in figure [ figure0-main ] . to construct the virtual khovanov complex for v - tangles we need to extend the notions above in such a way that they still work for open cobordisms . a first generalisation is easy , i.e. we will still use immersed possible non - orientable surfaces with boundary decorations , but we allow vertical boundary components , e.g. the three v - reidemeister cobordisms vrm1 , vrm2 and vrm3 cobordisms in figure [ figure0 - 3 ] . one main point is the question what to do with the _ open saddles _ , i.e. saddles with no closed boundary . see bottom row in figure [ figure1 - 2 ] . a possible solution is to define them to zero . but this has two major problems . first the loss of information is big and second we would not have local properties like in the classical case ( `` tensoring '' of smaller parts ) since an open saddle can , after closing some of his boundary circles , become either @xmath14 , @xmath13 or @xmath15 . see figure [ figure0 - 4 ] . hence , an information mod 3 is missing . we therefore consider morphisms with an _ indicator _ , i.e. an element of the set @xmath24 , e.g. see figure [ figure1 - 2 ] . then , after taking care of some technical difficulties , the concept extens from c - tangles to v - tangles . the outline of the paper is as follows . * in section [ v - tangles ] we define the notions of a _ v - tangle ( diagram ) _ with @xmath7-boundary points . moreover , we define the geometric category we are working in and define _ two different _ versions of the virtual khovanov complex . we show in theorem [ theo - geoinvarianz ] that both are invariants of v - tangles and we show in theorem [ theo - marker ] that both versions are equal for c - tangles and v - links . * in section [ circuit ] we discuss the local properties of the construction . we define the notion of a _ decorated circuit diagram _ , i.e. the virtual analogon of a planar diagram , and we show in theorem [ theo - semiloc ] that the construction has semi - local properties , i.e. the classical notion extents to a certain class of v - tangle diagrams . a main tool is the _ dot - calculus _ which is sketched in figure [ figure0-main ] , i.e. change the decorations iff the orientations of the input diagram and the local piece mismatch ( red dot ) , do not change them iff they match ( green dot ) and add in a bolt symbol ( non - orientable ) iff at the top and the bottom one gets a green and a red dot . * in section [ leedeg ] we show that the extension of lee s variant of khovanov homology still has the same `` degeneration '' as in the classical case , i.e. the homology just `` counts '' the number of v - link components . see theorem [ theo - leedeg ] . note that this is rather surprising since the map @xmath15 will be a @xmath11-morphism if @xmath25 is invertible . our main observation why the degeneration still holds is explained in example [ beisp - leedeg ] . in order to prove theorem [ theo - leedeg ] we use a special trick , i.e. we define two idempotents called _ down _ and _ up _ , go to the _ karoubi envelope _ and prove that there , using the semi - local construction of section [ circuit ] , the question simplifies to the case of counting _ non - alternating resolutions_. see figure [ figure4 - 2 ] . for a v - tangle diagram @xmath26 we call @xmath27 the _ @xmath11- _ and @xmath28 the _ @xmath29-resolution _ of the crossing @xmath4 . for an oriented v - tangle diagram @xmath26 we call @xmath30 a _ positive _ and @xmath31 a _ negative _ crossing . the _ number of positive crossings _ is called @xmath32 and the _ number of negative crossings _ is called @xmath33 . for a given v - tangle diagram @xmath26 with @xmath34-numbered crossings we define a collection of closed curves and open strings @xmath35 in the following way : let @xmath36 be a word of length @xmath34 in the alphabet @xmath37 . then @xmath35 is the collection of closed curves and open strings which arise when one performs a @xmath38-resolution at the @xmath39-th crossing of @xmath26 for all @xmath40 . we call such a collection @xmath35 the _ @xmath36-th resolution _ of @xmath26 . we can choose an orientation for the different components of @xmath35 . we call such a @xmath35 an _ orientated resolution_. if we ignore orientations , then there are @xmath41 different resolutions @xmath35 of @xmath26 . we say a resolution has length @xmath14 if it contains exactly @xmath14 1-letters . that is @xmath42 . for two resolutions @xmath35 and @xmath43 with @xmath44 and @xmath45 for one fixed @xmath46 and @xmath47 for @xmath48 we define a _ saddle between the resolutions @xmath49_. this means : choose a small ( no other crossing , classical and virtual , should be involved ) neighbourhood @xmath50 of the @xmath46-th crossing and define a cobordism between @xmath35 and @xmath43 to be the identity outside of @xmath50 and a saddle inside of @xmath50 . sometimes we need a so - called _ spanning tree argument _ , i.e. choose a spanning tree of a cube ( like in figure [ figure0 - 5 ] ) and change e.g. orientations such that the edges of the tree change in a suitable way , starting at leafs and remove them and repeat . note that any v - tangle diagram should be oriented in the usual sense . but with a slight abuse of notation , we will repress this orientation throughout the whole paper . recall that these orientations are needed for the shifts in homology gradings , see for example @xcite . i wish to thank v. manturov for corrections and many helpful comments . moreover i wish to thank a. kaestner and l. kauffman for suggestions and helpful comments helping me to write a computer program for calculations in the future . in this section we describe how the virtual khovanov complex can be extended in such a way that it also works for v - tangles and not just for v - links . our extension follows the ideas of d. bar - natan in @xcite . first we define the notion of a _ virtual tangle ( diagram ) _ , called _ v - tangle ( diagram)_. [ defin - vtangle ] a _ virtual tangle diagram with @xmath51 boundary points _ @xmath26 is a planar graph embedded in a disk @xmath6 . this planar graph is a collection of _ usual vertices _ and @xmath7-_boundary vertices_. we also allow circles , i.e. closed edges without any vertices . the usual vertices are all of valency four . any of these vertices is either an overcrossing @xmath4 or an undercrossing @xmath5 or a virtual crossing . latter is marked with a circle . the boundary vertices are of valency one and are part of the boundary of @xmath6 . we call the crossings @xmath4 and @xmath5 _ classical crossings _ or just _ crossings _ and a virtual tangle diagram without virtual crossings a _ classical tangle diagram_. a _ virtual tangle with @xmath51 boundary points _ @xmath52 is an equivalence class of virtual tangle diagrams @xmath26 module boundary preserving isotopies and _ generalised reidemeister moves _ ( see figure [ figure0 - 1 ] ) . we call a virtual tangle @xmath52 _ classical _ if the set @xmath52 contains a classical tangle diagram . a v - circle is a circle without classical crossings and a v - string is a string starting and ending at the boundary without classical crossings . we call a v - circle / v - string without virtual crossings a _ c - circle / c - string_. the _ closure of a v - tangle diagram with * -marker _ @xmath53 is a v - link diagram which is constructed by capping of neighbouring boundary points ( starting from a fixed point marked with the * -marker and going counter - clockwise ) without creating new virtual crossings . for an example see figure [ figure1 - 1 ] . there are exactly two , maybe non equivalent , closures of any v - tangle diagram . in the figure below the two closures are pictured using green edges . the notions of an _ oriented _ virtual tangle diagram and of an _ oriented _ virtual tangle are defined analogue . the latter modulo _ oriented _ generalised reidemeister moves and boundary preserving isotopies . from now on every v - tangle ( diagram ) should be oriented . but we repress this notion to avoid confusion with the ( more important ) notions of oriented resolutions and orientations for circuit diagrams ( see section [ circuit ] ) . moreover , a _ decorated _ v - tangle diagram is a v - tangle diagram with orientations for all of its circles / strings . we discuss the category of _ open cobordisms with boundary decorations_. it is almost the same as in @xcite , but the corresponding cobordisms could be open , i.e. they could have vertical boundary components , and are decorated with an extra information , i.e. a number in the set @xmath24 . we picture the number @xmath11 as a bolt . if the reader is unfamiliar with the notions and relations of cobordisms we suggest to check for example the book of j. kock @xcite . [ defin - category ] let @xmath51 and let @xmath54 be a commutative and unital ring . our category should be @xmath55pre - additive . the symbol @xmath56 should denote the disjoint union ( the coproduct in our category ) . * the objects : * the _ objects _ _ @xmath57 _ ) are numbered v - tangles diagrams with @xmath7 boundary points without classical crossings . we denote the objects as @xmath58 . here @xmath59 are the v - circles or v - strings and @xmath60 is a finite , ordered index set . the objects of the category should be unique up to _ boundary preserving planar isotopies _ of planar graphs . * the generators : * the _ generators _ of @xmath61 are the cobordisms from figure [ figure1 - 2 ] . the cobordisms pictured are all between c - circles or c - strings . every generator has a decoration from the set @xmath24 . we call this decoration the _ indicator _ of the cobordisms . if no indicator is pictures it should be @xmath22 . every generator with a decoration @xmath62 has extra decorations from the set @xmath63 at every horizontal boundary component . we call these decorations the _ gluing numbers _ of the cobordism . we consider these cobordisms up to boundary preserving homeomorphisms ( as abstract surfaces ) . hence , between circles or strings with v - crossings the generators are the same up to boundary preserving homeomorphisms immersed into @xmath64 $ ] . the vertical boundary components are pictured in red . we denote the different generators ( from left to right ; top row first ) @xmath65 and @xmath66 , @xmath67 and @xmath68 , @xmath69 , @xmath70 and @xmath15 , @xmath71 and @xmath72 , @xmath73 and @xmath74 , @xmath75 , @xmath15 and @xmath76 . * the morphisms : * the _ morphisms _ _ @xmath61 _ are cobordisms between the objects in the following way . first we identify the collection of numbered v - circles or v - strings with circles or strings immersed into @xmath6 . given two objects @xmath77 with @xmath78 numbered v - circles or v - strings , then a morphism @xmath79 is a surface immersed in @xmath64 $ ] whose non vertical boundary lies only in @xmath80 and is the disjoint union of the @xmath81 numbered v - circles or v - strings from @xmath82 in @xmath83 and the disjoint union of the @xmath84 numbered v - circles or v - strings from @xmath85 in @xmath86 . the morphisms should be generated ( as abstract surfaces ) by the generators from above ( see figure [ figure1 - 2 ] ) . * the decorations : * every morphisms has an _ indicator _ from the set @xmath24 . moreover , every morphism @xmath79 in @xmath61 is a cobordism between the numbered v - circles or v - strings of @xmath82 and @xmath85 . let us say that the v - circles or v - strings of @xmath82 are numbered @xmath87 and the v - circles or v - strings from @xmath85 are numbered for @xmath88 . every cobordism with @xmath89 as an indicator should have a decoration on the @xmath39-th boundary circle . this decoration is an element of the set @xmath63 . we call this decoration of the @xmath39-th boundary component the _ @xmath39-th gluing number _ of the cobordism . hence , the morphisms of the category are pairs @xmath90 . here @xmath79 is a cobordism from @xmath82 to @xmath85 immersed in @xmath64 $ ] and @xmath91 is a string of length @xmath92 in such a way that the @xmath39-th letter of @xmath91 is the @xmath39-th gluing number of the cobordism and the last latter is the indicator or @xmath93 if the cobordism has @xmath11 as indicator . * short hand notation : * we denote a morphism @xmath94 which is a connected surfaces as @xmath95 . here @xmath96 are words in the alphabet @xmath63 in such a way that the @xmath97th character of @xmath98 ( of @xmath99 ) is the gluing number of the @xmath97th circle of the upper ( of the lower ) boundary . the number @xmath100 should be the indicator . the construction above ensures that this notation is always possible . therefore we denote an arbitrary morphism @xmath90 as @xmath101 here @xmath102 are its connected components and @xmath103 are words in @xmath63 . for a morphism with @xmath11 as indicator we do not need any boundary decorations . with a slight abuse of notation , we denote all these cobordisms like the non orientable cobordisms as @xmath15 . * the relations : * and just the product otherwise . ] and just the product otherwise . ] and just the product otherwise . ] and just the product otherwise . ] and just the product otherwise . ] and just the product otherwise . ] and just the product otherwise . ] there are different relations for the cobordisms , namely _ topological relations _ and _ combinatorial relations_. the latter relations are describe by the gluing numbers and indicators of the cobordisms and the gluing of the cobordisms . the topological relations are not pictured but it should be clear how they should work . see for example j. kock in @xcite and our paper in @xcite . the monoidal structure should be induced by the disjoint union @xmath56 . the category is a semi - strict monoidal category . moreover the category is symmetric . hence , by mac lane s coherence theorem ( see @xcite ) we can assume that the category is a strict , symmetric , monoidal category . we define the full category @xmath104 to be the category whose objects are @xmath105 and whose morphisms are @xmath106 . note the following collection of formulas that follow from the relations . recall that @xmath68 and @xmath72 change the decorations and that @xmath15 and @xmath76 change the indicators . with a slight abuse of notation we repress to write tensors if it is not neccessary , i.e. for the indicator changes . moreover , since @xmath68 and @xmath72 satiesfy similar formulas , we only write down the equations for @xmath68 . [ lemm - calc ] let @xmath107 be two objects in @xmath108 . let @xmath109 be a morphism that is connected , has @xmath110 as an indicator and @xmath98 and @xmath99 as decorated boundary strings . then we have the following identities . we write @xmath111 as a short hand notation if the indicators and gluing numbers do not matter . * @xmath112 ( indicator changes commute ) . * @xmath113 ( @xmath15 commutes ) . * @xmath114 ( first decoration commutation relation ) . * let @xmath115 denote the decoration change at the corresponding positions of the words @xmath96 . then we have @xmath116(second decoration commutation relation ) . everything follows by a straightforeward usage of the relations in definition [ defin - category ] above . next we define some short hand notations for the different categories we use . for any pre - additive category @xmath117 we define @xmath118 as the _ category of formal matrices _ and @xmath119 as the _ category of formal chain complexes _ over @xmath117 . if the reader is unfamiliar with the notations we refer to @xcite . [ defin - category2 ] we call @xmath120 the category @xmath121 . here our objects are formal chain complexes of formal direct sums in the category of ( possible non orientable ) open cobordisms with boundary decorations . then we define @xmath122 as the category @xmath120 modulo formal chain homotopy . moreover , we define @xmath123 , which has the same objects as the category @xmath108 , but the morphisms are all modulo the local relations from figure [ figure0 - 2 ] . furthermore we define @xmath124 and @xmath125 in the obvious sense . the notations @xmath126 or @xmath127 mean that we consider all possible cases , namely with or without a @xmath128 and with or without a @xmath99 . moreover , we use the same notions if @xmath129 . we are now able to define the _ geometric complex of a v - tangle diagram _ @xmath26 . for this construction we use our notations for the _ saddle decorations _ and _ saddle signs _ of v - link diagrams @xmath130 from section two in @xcite . the reader unfamiliar with the notations may check the corresponding definitions . we only repeat these two important constructions in a very brief summary . the only new notion in the definition below is the spreading of the _ indicator of the saddle_. [ defin - saddlestructure ] let @xmath130 be a v - link diagram . moreover , let @xmath131 be a saddle between the numbered , orientated resolutions @xmath132 . every saddle can be viewed as an formal symbol @xmath16 together with a formal local orientation of these strings which is equal to the orientations of @xmath132 . we say the saddle of the form @xmath17@xmath18 is of _ standard form_. moreover , choose a position for the _ x - marker _ at any crossing and a numbering of the v - circles in the resolutions . recall that the x - marker is just a choice in which direction one rotates a saddle @xmath133 until one can compare it to the saddle in standard form . the complex will be well - defined for any choice and different choices give chain isomorphic complexes ( section two in @xcite ) . for every orientable saddle the _ saddle sign _ is the product of two numbers , both either @xmath29 or @xmath23 . if the x - marker is on the lower numbered saddle , then the first number should be @xmath29 and @xmath23 otherwise . moreover , if the number of v - circles with a higher number then the x - marked v - circle is odd , then the second number should be @xmath29 and @xmath23 otherwise . the reader should check that this definition corresponds to the one in section two of @xcite . now we spread the _ saddle decorations _ as follows : * if the saddle is non orientable , then we do not need any decorations . * if the saddle is orientable , then we compose the standard saddles @xmath70 or @xmath69 with @xmath68 for every not standard oriented string . the saddle should have a formal minus sign iff the saddle sign is negative . every orientable surface should carry an indicator @xmath22 iff the number of upper boundary components of the saddle is two and a @xmath23 iff the number is one . every non orientable saddle gets a zero as indicator . everything together , i.e. boundary decorations , saddle sign and indicator , is denoted as the _ saddle decoration _ of @xmath49 . beware that many choices are involved . but the do not change the complex up to chain isotopies ( lemma [ lemma - everythingfine ] ) . now let @xmath26 be a tangle diagram with a * -marker on the boundary and let @xmath53 be the closure of the diagram . the _ saddle decoration _ of the saddles of @xmath26 should be the ones induced by the saddle decorations of the closure . [ defin - geocomplex ] for a v - tangle diagram @xmath26 with a * -marker on the boundary and with @xmath34 ordered crossings we define _ the geometric complex _ @xmath134 as follows : * for @xmath135 the @xmath136 _ chain module _ is the formal direct sum of all oriented resolutions @xmath35 of length @xmath39 . * there are only morphisms between the chain modules of length @xmath39 and @xmath137 . * if two words @xmath138 differ only in exactly one letter and @xmath44 and @xmath45 , then there is a morphism between @xmath35 and @xmath43 . otherwise all morphisms between components of length @xmath39 and @xmath137 are zero . * this morphism is a _ saddle _ between @xmath35 and @xmath43 . * the saddles should carry the _ saddle decorations _ from definition [ defin - saddlestructure ] . note that it is not clear at this point why we can choose the numbering of the crossings , the numbering of the v - circles and the orientation of the resolutions of the closure . furthermore it is not clear why this complex is a well - defined chain complex . we show in lemma [ lemma - everythingfine ] that the complex is independent of these choices , i.e. if @xmath139 and @xmath140 are well - defined chain complexes with different choices , then they are equal up to chain isomorphisms . the same lemma ensures that the complex is a well - defined chain complex . beware that the position of the * -marker is important for v - tangle diagrams . but theorem [ theo - marker ] ensures that the position is not important for classical v - tangles and v - links . if it does not matter which of the possible two different chain complexes is which , i.e. it is just important that they could be different , then we denote them as @xmath141 and @xmath142 for a v - tangle diagram @xmath26 without a chosen marker position . for an example see figure [ figure1 - 10 ] . this figure shows the virtual khovanov complex of a v - tangle diagram with two different * -marker positions . the vertical arrow between them indicates that they are chain isomorphic . [ lemma - everythingfine ] let @xmath26 be a v - tangle diagram with a * -marker and let @xmath143 be its geometric complex from definition [ defin - geocomplex ] with arbitrary orientations for the resolutions of the closure . let @xmath144 be the complex with the same orientations for the resolutions except for one circle @xmath145 in one resolution @xmath35 . if a face @xmath146 from @xmath143 is anti - commutative , then the corresponding face @xmath147 from @xmath144 is also anti - commutative . moreover , if @xmath143 is a well - defined chain complex , then it is isomorphic to @xmath144 , which is also a well - defined chain complex . the same statement is true if the difference between the two complexes is the numbering of the crossings , the choice of the rotation for the calculation of the saddle signs or the fixed numbering of the v - circles of the closure . moreover , the same is true for any rotations / isotopies of the v - tangle diagram . hence , we can speak of _ the _ geometric complex @xmath134 of the v - tangle diagram with a * -marker . the complex is a well - defined chain complex . for v - tangle diagrams @xmath26 with @xmath148 the statement is the same as the corresponding statement in section two in @xcite . recall that the trick is to reduce all faces through a finite sequence of vrm1 , vrm2 , vrm3 and mrm moves ( see figure [ figure0 - 1 ] ) and virtualisations ( see figure [ figure1 - 11 ] ) to a finite number of different possible faces . then one does a case - by - case check . because the saddles in the two chain complexes are topological the same , we only have to worry about the decorations . but the decorations are spread based on the closure of the v - tangle diagram and the relations from definition [ defin - category ] are build in such a way that the open cases behave like the closed ones . hence , we can use the statement for @xmath148 to finish the proof . the next theorem is very important but the proof itself is almost equal to the proof of d. bar - natan in @xcite and our version in @xcite . therefore we skip the details . [ theo - geoinvarianz ] let @xmath149 be two v - tangle diagrams with the same * -marker position which differs only through a finite sequence of isotopies and generalised reidemeister moves . then the complexes @xmath134 and @xmath150 are equal in @xmath151 . because the proof of d. bar - natan in @xcite and therefore the corresponding proof in @xcite is local we could copy the arguments from there . the lemma [ lemma - everythingfine ] guarantees that we can choose the numbering and orientations without changing anything up to chain isomorphisms . furthermore we use the statements of the theorem [ theo - semiloc ] and the corollary [ koro - reidemeister ] to finish the proof . beware that the chain homotopies in @xcite should all carry @xmath22 as indicator . we need some notions now . let @xmath26 denote a v - tangle diagram . we call a part of @xmath26 a _ connected part _ if it is connected as the four - valent graph on a surface with boundary which is created by replacing every c - crossing with a four - valent vertex and eliminating v - crossings by adding handles to a sphere with boundary . the genus of the surfaces should be high enough such that the graph has no v - crossings any more . we call a connected part of a v - tangle diagram _ fully internal _ if it is not adjacent to the boundary . see figure [ figure1 - 12 ] . the left v - tangle diagram has one connected part , which is not fully internal , and the right v - tangle diagram has two connected parts , one fully internal and one not fully internal . a virtual crossing is called _ negligible _ if it is part of a fully internal part , e.g. all v - crossings of the right v - tangle diagram in figure [ figure1 - 12 ] are negligible . we call a v - tangle diagram @xmath26 _ nice _ if there is a finite sequence of of vrm1 , vrm2 , vrm3 and mrm moves and virtualisations such that every v - circle is negligible , e.g. every v - link diagram is nice and every c - tangle diagram is nice . an example of a not nice v - tangle diagram is pictured in figure [ figure2 - 4 ] . we note that for a v - tangle diagram @xmath26 the chain complexes @xmath141 and @xmath142 are almost the same , i.e. they have the same vertices but possible different edges ( which are still in the same position ) . the next lemma concretise the observation . the numbering in the lemma should be the same for the two complexes . [ lemma - factorbyisos ] let @xmath26 be a v - tangle diagram . let @xmath152 and @xmath153 denote the numbered saddles of @xmath141 and of @xmath142 . then we have a factorisation of the form @xmath154 for invertible cobordisms @xmath155 for every @xmath39 if @xmath26 is a nice v - tangle diagram . that the saddles are topological equivalent is clear . we only have take a look at the decorations . the main point is the following observation : of the four outer ( two on both sides ) cobordisms in the bottom row of figure [ figure1 - 2 ] , i.e. @xmath71 , @xmath72 , @xmath156 and @xmath76 , only the third is not invertible . the first is the identity , the second and fourth are their own inverses . the third is not invertible because the @xmath11-indicator can not be changed to an @xmath157-indicator . note that none of the vrm1 , vrm2 , vrm3 and mrm moves nor a virtualisations changes the indicator of a saddle cobordism . hence , it is sufficient to show the statement for a v - tangle diagram with only negligible v - crossings . from the observation above it is enough to show that every saddle gets an @xmath11-indicator in one closure iff it gets an @xmath11-indicator in the other closure . the only possible way that a saddle gets an indicator from @xmath62 for one closure and an @xmath11-indicator for the other closure is the rightmost case in figure [ figure0 - 4 ] . but for this case the existence of a non negligible v - crossing is necessary . hence , we get the statement . [ prop - nicechainiso ] let @xmath26 be a v - tangle diagram . then @xmath141 and @xmath142 are chain isomorphic if @xmath26 is nice . let @xmath26 be a nice v - tangle diagram . then lemma [ lemma - factorbyisos ] ensures that every saddle is the same , up to isomorphisms , in @xmath141 and @xmath142 . furthermore lemma [ lemma - everythingfine ] ensures that both are well - defined chain complexes . hence , the number of signs of every face is odd ( also counting the ones from the decorations ) . hence , we can use a spanning tree argument to construct the chain isomorphism explicit , i.e. start at a leaf of a spanning tree of the khovanov cube and change the orientations of the vertex such that the unique outgoing edge of the tree has the same sign in both cases ( lemma [ lemma - everythingfine ] ensures that nothing changes modulo chain isomorphisms ) . continue along the vertices of the spanning tree remove already visited leafs . this construction generates a chain isomorphism . next repeat the whole process , but change the indicators now . the lemma [ lemma - factorbyisos ] ensures that this will also generate a chain isomorphism . the chain isomorphism that we need is the composition . see for example figure [ figure1 - 10 ] . the following theorem ensures that the choice of the position of the marker is not important for classical v - tangles or v - links . hence , we can say that our notions extents both , the classical version and the version for v - tangle diagrams . note that this is also an invariant for the question if a v - tangle diagram is classical or not . [ theo - marker ] let @xmath26 be a v - tangle diagram with two different * -marker positions . let @xmath141 and @xmath142 be the geometric complex from definition [ defin - geocomplex ] for the two positions . then the two complexes are equal in @xmath151 if the v - tangle has @xmath148 or is a c - tangle . for a v - tangle with @xmath148 we can use the proposition [ prop - nicechainiso ] above . for a classical v - tangle we can choose a diagram without virtual crossing without changing anything up to chain homotopies , because of theorem [ theo - geoinvarianz ] . then we can use the proposition [ prop - nicechainiso ] again . note that the whole construction can be done with an arbitrary _ non virtual closure _ of a v - tangle diagram , i.e. capping it of without creating new v - crossings . one can repeat the arguments above to prove analogous statements like above for this construction , too . [ rem - gradings ] we could use the euler characteristic to introduce the structure of a graded category ( see @xcite ) on @xmath108 ( and hence on @xmath158 ) . the differentials in the geometric complex from definition [ defin - geocomplex ] have all @xmath159 ( after a grade shift ) , because their euler - characteristic is @xmath23 . then it is easy to proof that the geometric complex is a v - tangle invariant under graded homotopy . the proof of this statement is analogue to the one of d. bar - natan from @xcite . in the present section we describe the notion of a _ circuit algebra_. a circuit algebra is almost the same as a planar algebra , but we allow virtual crossings . planar algebras were introduced by v. jones ( see @xcite ) and were for example studied by d. bar - natan in the case of classical khovanov homology ( see @xcite ) . hence , we can use most of his constructions in our case , too . a crucial difference is that we need to _ decorate _ our _ circuit diagrams_. this is necessary because our cobordisms are also decorated . we start the section with the definition of a circuit diagram . recall that we call v - tangle diagram _ decorated _ if all of its strings / circles are oriented . we call a cobordisms _ decorated _ if it has gluing numbers at the boundary and indicators , see definitions [ defin - vtangle ] and [ defin - category ] . moreover , in the whole section every v - tangle diagram should carry a * -marker . [ defin - circuit diagram ] let @xmath160 denote a disk embedded into @xmath161 , the so - called _ outside disk_. let @xmath162 denote @xmath163 disks @xmath6 embedded into @xmath161 such that for all @xmath163 the disk @xmath162 is also embedded into @xmath160 without touching the boundary of @xmath160 , i.e. @xmath164 and @xmath165 . we denote @xmath166 . a _ circuit diagram with @xmath14 input disks _ @xmath167 is a planar graph embedded into @xmath168 with only vertices of valency one and four in such a way that every vertex of valency one is in @xmath169 and every vertex of valency four is in @xmath170 . all vertices of valency four are marked with a virtual crossing . again we allow circles , i.e. closed edges without any vertices . see for example figure [ figure2 - 1 ] . * -marked circuit diagram _ is the same with @xmath171 extra * -marker for every boundary of @xmath168 . moreover , we call the vertices at @xmath172 the _ outer boundary points_. a _ closure _ of a * -marked circuit diagram with @xmath14 input disks @xmath173 is a circuit diagram with @xmath14 input disks and without any outer boundary points which is constructed from @xmath167 by capping of neighbouring strings starting from the * -marker and proceeding counter - clockwise . note that we only cap of the outside disk and not the small inside disks . a _ decoration _ for a * -marked circuit diagram is a tuple of a numbering and an orientation of the strings of the diagram in such a way that its also an numbering and orientation of the closure . we call a circuit diagram together with a decoration a _ decorated circuit diagram_. see for example figure [ figure2 - 1 ] . the decoration of the circuit diagram in this figure is also a decoration for the closure ( the diagram together with the green lines ) . with this setting we can implement the definition of a _ ( decorated ) circuit algebra_. we follow the definitions of d. bar - natan and v. jones ( see @xcite , @xcite ) . recall that our v - tangle diagrams should always be oriented . [ defin - circuit algebra ] let @xmath174 be the set of ( decorated ) v - tangle diagrams with @xmath7 boundary points and a * -marker and let @xmath175 denote the quotient by boundary preserving isotopies and generalised reidemeister moves . furthermore let @xmath167 denote a ( decorated ) circuit diagram with @xmath14 input disks and @xmath176 outer boundary points in such a way that the @xmath177-th input disk has @xmath178 numbered boundary points . because @xmath167 has no classical crossings , this induces operations @xmath179 by placing the @xmath39-th v - tangle diagram from @xmath180 in the @xmath39-th boundary component of @xmath167 , i.e. glue the v - tangle inside in such a way that the * -markers match . see the right side of figure [ figure2 - 2 ] . there is an identity operation on @xmath181 ( it is of the form @xmath182{figs / circuitid.pdf}}\end{array}\hspace{-3pt}}}$ ] ) and the operations are compatible in a natural way ( associative ) . we call a set of sets @xmath183 along with operations @xmath167 like above a _ circuit algebra _ , provided that the identity and associativity from above holds . if the operators and elements are decorated , first with numbers and orientations and latter only with orientations , then we call a set of sets @xmath183 like above a _ decorated circuit algebra_. note that in this case we have to define how the decorations change after gueing , since we can run into ambiguities . a necessary definition is the notion of a @xmath184-morphism . here @xmath184 is a subset of a circuit algebra @xmath183 . we need this notion because , in contrast to the classical case , the bracket is not a morphism of circuit algebras . see ( c ) of theorem [ theo - semiloc ] below . [ defin - cirmor ] a _ morphism @xmath185 of ( decorated ) circuit algebras _ @xmath183 and @xmath186 is a map @xmath187 which satisfies @xmath188 for every ( decorated ) operation @xmath167 . let @xmath184 be a subset of @xmath183 . a _ @xmath184-morphism @xmath185 of ( decorated ) circuit algebras _ @xmath183 and @xmath186 is a map @xmath187 which satisfies @xmath189 for every ( decorated ) operation @xmath167 here are some examples . the reader may also check the corresponding section in the paper of d. bar - natan from @xcite . [ beisp - circuit ] the first example is the set @xmath190 from definition [ defin - category ] , i.e. v - tangles diagrams with @xmath51 boundary points without classical crossings and an extra * -marker . this is a sub - circuit algebra of the circuit algebra that allows classical crossings . this example gets more exciting if we want to view it as a decorated circuit algebra . we have to define the operations , i.e. the gluing inside the input disks , in more detail now , since we can run into ambiguities , see top row of figure [ figure2 - 3 ] . we can run into ambiguities if the decorations of the operator get glued together do not match up . in this case we define the new decoration based on the rule _ `` lower first '' _ , i.e. the new number should be the lower and the new orientation should be the one from the lower numbered string . see lower row of figure [ figure2 - 3 ] . not all four cases are pictured but it should be clear how the other two work . furthermore if we glue a decorated v - tangle diagram in an input disk , then we run into ambiguities if the shared decorations , i.e. the orientations , does not match up . in this case we change the decorations of the v - tangle diagram . we add in a _ red dot _ if we have to change the orientation and a _ green dot _ otherwise . this is pictured in the top row of figure [ figure2 - 3 ] . the reader should check that this gives rise to a decorated circuit algebra . another important example is the whole collection @xmath191 from definition [ defin - category ] , i.e. decorated cobordisms with @xmath51 vertical boundary lines and an extra * -marker . we want to view this example as a decorated circuit algebra again . hence , we have to define the operations . the most interesting point are the decorations again , because it should be clear how to glue a cobordism with @xmath14 vertical boundary lines into @xmath192 $ ] . this time we have to define the behaviour of three decorations , i.e. the gluing numbers and the signs and indicators . the gluing numbers should be treated like the orientations before , i.e. if they do not match , then we use the gluing number of the lower numbered string . the indicators ( recall that they are just numbers in @xmath24 ) should be multiplied . recall that a cobordism with an @xmath11-indicator does not get any gluing numbers . we simply remove them in this case . we get the following interaction of the operation of @xmath167 on @xmath190 and @xmath191 ( compare to figure [ figure2 - 3 ] ) : * a saddle is composed with @xmath68 iff the v - tangle gets a red dot at the corresponding position . * a saddle is composed with an @xmath11-indicator surfaces iff the v - tangle gets a red and a green dot in both resolutions at the corresponding position . * a saddle is composed with an @xmath23-indicator surfaces iff the v - tangle numbers get identified at the top resolution at the corresponding position . * a saddle gets an extra minus sign iff the v - tangle numbers change in such a way that the sign of @xmath193 is changed at the corresponding position . these rules define a new decoration for the new cobordism . the reader should check again that this gives rise to a decorated circuit algebra . we summarise the notions in a definition . recall that v - tangle diagrams are decorated with orientations and a * -marker and cobordisms are decorated with orientations on the boundary , an indicator and a * -marker . [ defin - dotcalculus ] let @xmath167 denote a decorated circuit diagram with @xmath14 input disks and @xmath176 outer boundary points in such a way that the @xmath177-th input disk has @xmath178 numbered boundary points . then @xmath167 induces an associative and unital ( in the above sense ) operation on decorated v - tangles ( with a corresponding number of boundary points ) by the `` lower first''-rule , i.e. if the orientation does not match , then the lower number induces the new orientation . put a red dot on every string that has its orientation changed and a green dot otherwise . we call this the _ tangle dot - calculus_. moreover , @xmath167 induces an associative and unital ( in the above sense ) operation on decorated cobordisms ( with a corresponding number of boundary lines ) by the `` lower first''-rule , i.e. if the orientation does not match , then the lower number induces the new orientation . put a red dot on every string that has its orientation changed and a green dot otherwise and compose the corresponding boundary with @xmath68 iff the string has a red dot , multiply indicators via identity surfaces with corresponding indicators @xmath89 iff the v - tangle numbers get identified at the bottom / top resolution at the corresponding position , multiply with an @xmath11-identity iff in both resolutions the strings are identified and multiply with a formal sign iff the numbers change in such a way that the sign is changed at the corresponding position . we call this the _ dot - calculus_. if the crossings carries x - markers , then they should be in the same position as before after the operation . the reader should compare the notions above with figure [ figure0-main ] . we follow the notions of d. bar - natan in the next theorem , i.e. we use the notions @xmath120 and @xmath122 from definition [ defin - category2 ] , but we also allow @xmath194 , i.e. * -marked v - tangles with a finite number of boundary points . we denote the collection of * -marked nice v - tangles as @xmath195 . recall that every v - link diagram and every c - tangle diagram is nice . [ theo - semiloc ] * the whole collection @xmath196 has a natural structure of a decorated circuit algebra . * the operations on @xmath196 preserves homotopy equivalence . thus we obtain that @xmath197 has a natural structure of a decorated circuit algebra . * the bracket @xmath198 is not a morphism of decorated circuit algebras . * the bracket @xmath198 is a @xmath199-morphism of decorated circuit algebras . we follow the proof of the classical version in @xcite . thus we start by extending the structure of a decorated circuit algebra of the set @xmath190 and the one of @xmath191 from example [ beisp - circuit ] to decorated circuit structures @xmath200 and @xmath201 multilinear . let @xmath167 denote a decorated circuit diagram with @xmath14 input disks and @xmath202 outer boundary points in such a way that the @xmath177-th input disk has @xmath178 numbered boundary points . \(a ) for complexes @xmath203 with @xmath204 @xmath205^{d^{l-1}_{j } } & \mathrm{co}^l_j\ar[r]^{d^l_j } & \mathrm{co}^{l+1}_{j}\ar[r]^{\phantom{a}d^{l+1}_{j } } & \cdots } \end{xy}\ ] ] of v - tangle diagrams with @xmath178 boundary points we define the new `` tensored '' complex @xmath206 as follows . let @xmath207 denote the composition with the morphisms that we compose after applying the circuit diagram on cobordisms ( see example [ beisp - circuit ] and definition [ defin - dotcalculus ] above ) , i.e. the red dots induce a composition with @xmath68 ( or a composition with a @xmath11-identity surface in the degenerated case ) and a change in the numbering induces a composition with a cobordisms that changes signs and indicators . therefore we denote the operation of @xmath167 on cobordisms , i.e. the dot - calculus , as @xmath208 to illustrate the difference to the classical case . we skip this notion for the objects . the @xmath39-th chain module should be @xmath209 and the chain maps should be @xmath210 beware that the needed minus signs are not missing . they are already implied in the definition of the decorated circuit algebra structure of @xmath201 , i.e. the dot - calculus in particular adds or removes signs . see example [ beisp - circuit ] and definition [ defin - dotcalculus ] . we use lemma [ lemma - everythingfine ] to see that we still get a chain complex . moreover , we need to define the operation of @xmath167 on the morphisms of the category @xmath196 , i.e. on the chain maps . so let @xmath211 be a chain map . every chain map itself is a sequence of matrices of decorated cobordisms @xmath212 . the operation on the chain map should just insert the cobordisms topological , i.e. without changing gluing numbers , indicators or signs . note that we do not run into ambiguities since the decorations are defined for the closure for the fixed * -marker . we denote an entry of the matrix @xmath213 with abuse of notation as @xmath213 and similar for @xmath214 . hence , we have to check that the commutativity of the faces @xmath215 & \mathrm{co}_1^{i}\ar[r]^{d_1^{i}}\ar[d]^{f_i } & \mathrm{co}_1^{i+1}\ar[r]\ar[d]^{f_{i+1 } } & \cdots\\ \cdots\ar[r ] & \mathrm{co}_2^{i}\ar[r]^{d_2^{i } } & \mathrm{co}_2^{i+1}\ar[r ] & \cdots . } \ ] ] is preserved . the operation on the @xmath216 is defined as @xmath217 where @xmath218 are the decoration , indicator and signs changing morphisms . the rest is a case - by - case check using the relations in lemma [ lemm - calc ] and the equation @xmath219 . * if one of the two sides of the equation @xmath219 has a @xmath11-indicator , then the other has also a @xmath11-indicator . hence , in this case everything works out by the commutation relations in lemma [ lemm - calc ] . * since everything works out topological and the indicator changes are spread if strings get glued together , we see that if @xmath218 are only indicator changes , then , by commutativity of the indicator changes , the equation is preserved . * an analogous argument works if @xmath218 are only sign changes . * hence , lets assume that @xmath218 are only decoration changes , i.e. tensors of @xmath68 and @xmath67 , and we do not have any @xmath11-indicators . then one has different cases to check , i.e. based on the indicators for @xmath220 and @xmath221 . they all work analogous , so we only do one case , i.e. only @xmath222 and @xmath213 should have negative indicators . then we have by the commutation relations @xmath223 where @xmath224 should denote a short hand notation for the one from lemma [ lemm - calc ] . note that the equation @xmath219 also holds for indicators , i.e. not all combinations of indicators for @xmath220 and @xmath221 are possible . \(b ) the proof of ( b ) is an analogous argument like in ( a ) , i.e. we define the operation on the homotopies @xmath225 like before and we get the statement with an argument like before and the corresponding statement for tensor products . \(c ) this is true , because a surfaces with an @xmath11-indicator can not be changed to a surfaces with an @xmath157-indicator . for an concrete example see figure [ figure2 - 4 ] . the upper complex is not homotopy equivalent to the lower complex because we can not change the @xmath11-indicator . \(d ) this follows because of lemma [ lemma - factorbyisos ] , i.e. if a surface with an @xmath11-indicator appears in the complex before the operation @xmath167 , then it will appear for both closures . otherwise the , i.e. if the indicators are @xmath89 , then we can use the factorisation of lemma [ lemma - factorbyisos ] . note that the proof of lemma [ lemma - factorbyisos ] does not depend on theorem [ theo - geoinvarianz ] . then the equality is a consequence of the definition of the operation on the cobordisms and of an argument like in @xcite . note that the next corollary finish the proof of theorem [ theo - geoinvarianz ] . [ koro - reidemeister ] let @xmath226 be a v - tangle diagram that looks like a generalised reidemeister move ( see figure [ figure0 - 1 ] ) . then @xmath227 for any operation @xmath167 if @xmath228 is either a classical tangle or a reidemeister - tangle @xmath226 for @xmath229 . this is a direct consequence of point ( d ) of theorem [ theo - semiloc ] and the observation that the reidemeister - tangles , except the mrm - tangle , are all part of the set @xmath199 . that the statement also works if one of the @xmath228 is a mrm - tangle is a case - by - case check for the two different closures of a mrm - tangle ( both sides ) , i.e. in all cases the appearing cobordisms have @xmath89 indicators . [ rem - gradings2 ] like before and like in the paper of d. bar - natan @xcite , the whole construction works with gradings , too . but since this is a straightforward generalisations of the notions in @xcite , we skip the details here . in the whole section let @xmath54 denote a commutative and unital ring such that @xmath25 is invertible , e.g. @xmath230 $ ] . moreover , through the whole section we denote the geometric complex as @xmath231 and a corresponding algebraic version , i.e. an utqft ( see @xcite ) applied to the geometric complex , as @xmath232 or short as @xmath233 , e.g. we denote lee s version as @xmath234 note that , in order for the signs to work out correct , we have to fix a marker position . in the whole section we , by convention , say that the marker for @xmath4 is at the left side and for @xmath5 is at the top . recall ( see @xcite for the classical and @xcite for the virtual case ) that _ lee s variant _ , i.e. the algebraic version , for v - links is given by the following maps and the filtered module @xmath235/(x^2=1)$ ] . @xmath236 and @xmath237 moreover , recall that the geometric picture of lee s variant is given by the _ dot - relations _ in figure [ figure3 - 1 ] with @xmath238 , while the graded case of the khovanov complex is @xmath239 . note that @xmath240 allows us to use the _ dot - relation _ in figure [ figure3 - 1 ] instead of the local relations of figure [ figure0 - 2 ] . we give an example of the lee complex of a v - knot in example [ beisp - leedeg ] . lee s variant has a remarkable property in the classical case , i.e. e.s . lee showed that her variant just `` counts '' the number of components of the c - link , i.e. she showed that ( for @xmath241 ) the homology of an @xmath34-component link @xmath242 is @xmath243 so on the first hand this seems to be a boring invariant . but j. rasmussen ( see @xcite ) used this degeneration in a masterfully way to define the _ rasmussen invariant _ of a c - knot . note that this is a unexpected result since @xmath244 for @xmath245 ( see the relations in definition [ defin - category ] ) . hence , there are `` tons '' of @xmath11-morphisms in the complex . but these @xmath11-morphisms also come with isomorphism `` in a lot '' cases . the following example for the lee complex of a v - knot is a blueprint of this effect . [ beisp - leedeg ] consider the diagram of the virtual trefoil @xmath130 given in figure [ figure4 - 1 ] . in this example the number of negative crossings is zero , i.e. the leftmost object is the @xmath11-degree of the chain complex . let us consider @xmath241 . then @xmath244 and therefore the first two maps are @xmath11-morphisms . but note that the two right morphisms are not the same , i.e. one is @xmath246 and the other is @xmath247 . so on the algebraic level we get , using the maps from before , the following complex if we fix @xmath248 as a basis for @xmath249 and @xmath250 for @xmath251 . @xmath252^{\begin{pmatrix } 0 & 0 \\ 0 & 0 \\ 0 & 0 \\ 0 & 0 \end{pmatrix } } & & & a\oplus a\ar[rrr]^{\begin{pmatrix } 0 & 1 & 0 & -1\\ -1 & 0 & -1 & 0\\ -1 & 0 & 1 & 0\\ 0 & 1 & 0 & 1 \end{pmatrix } } & & & a\otimes a. } \end{xy}\ ] ] an easy calculations shows that the second matrix is an isomorphism . hence , the homology of the virtual trefoil is only non - trivial for @xmath148 , i.e. @xmath253 another example is the v - knot in figure [ figure4 - 2 ] , e.g. with the pictured orientation and numbering of the circles from left to right , the three outgoing morphisms from resolution @xmath254 to @xmath255 , @xmath256 and @xmath257 are ( up to , in this case , not important signs ) the morphisms @xmath258 , @xmath259 and @xmath260 , i.e. one alternating and two non - alternating . hence , the kernel is trivial . the reader should check that the rest also works out in the same fashion . the approach ( we follow d. bar - natan and s. morrison @xcite ) to show that the degeneration is still true is the following . first we define two orthogonal idempotents in our category , which we call _ down and up_. we can look at the _ karoubi envelope _ of our category , denoted @xmath261 . the idea of the karoubi envelope is to find a `` completion '' of a category such that every idempotent splits . it is named after the french mathematician m. karoubi , but it already appears in an earlier work by p. freyd in @xcite . then we show that the geometric complex of a simple crossing ( as a v - tangle ) , if considered in @xmath261 , is homotopy equivalent to a very simple complex with only @xmath11-morphisms . after that we use the local construction from section [ circuit ] to finish the proof . moreover , we proof the following interesting result of the number of decorations of v - link resolution with the `` colours '' down and up . note that we call an oriented resolution @xmath262 of a v - link diagram _ non - alternating _ if it is of the form @xmath263or @xmath21 at the corresponding positions of the saddles . recall that all the v - link diagrams should be oriented and that such a diagram with @xmath264 components has @xmath41 different orientations @xmath265 . we note that one can also colour the resolutions with honest colours , say red and green , in such a way that the colour changes at every v - crossing . we call this a _ colouring of a v - link resolution _ if at the corresponding saddle - position the colours are different , i.e. ( red , green ) or ( green , red ) . the reader should compare this with the coloured dots in figure [ figure0-main ] . [ theo - nonalternating ] let @xmath130 denote a v - link diagram with @xmath264 components . there are bijections of sets @xmath266 if @xmath130 is a v - knot , i.e. @xmath267 , then the two non - alternating resolutions are in homology degree @xmath11 . given an orientation @xmath268 of the v - link diagram @xmath130 , the map @xmath221 should assign the resolution @xmath262 which is obtained by replacing every oriented crossing of the form @xmath30 and @xmath31 to @xmath263 . this is clearly an injection . now , given a non - alternating resolution @xmath262 , we assign to it an orientation of @xmath130 in the following way . at any non - alternating part of the form @xmath263and @xmath21replace the non - alternating part with the corresponding oriented crossing @xmath30 and @xmath31 ( or a rotation in the @xmath21 case ) . these two maps are clearly inverses and for a v - knot the corresponding non - alternating resolutions are in homology degree @xmath11 since all @xmath32-crossings are resolved @xmath11 and all @xmath33-crossings are resolved @xmath29 in this procedure . to see the second bijection use a checker - board colouring of the v - link diagram . then start at any point of the non - alternating resolution and use the right - hand rule , i.e. the index finger follows the orientation and the string should get the colour of the face on the side of the thumb . [ beisp - nonalt ] let @xmath130 be the v - knot diagram in figure [ figure4 - 2 ] . then only the @xmath270 resolution of the v - knot diagram allows a non - alternating resolution . moreover , the orientation of the diagram induces this non - alternating resolution by replacing the three crossings with @xmath21,@xmath263and @xmath263 . the other orientation induces the non - alternating resolution @xmath263,@xmath21and @xmath21 . note that , by construction , these resolution are in homology degree @xmath11 . a computation like in example [ beisp - leedeg ] shows that these two non - alternating resolutions give the only two generators of the homology , i.e. @xmath253 for any category the notion of an idempotent morphisms , i.e. an arrow with @xmath272 , makes sense . moreover , in a pre - additive category the notion @xmath273 also makes sense . a classical trick in modern algebra is to use an idempotent , e.g. in an algebra @xmath249 , to split the algebra into @xmath274 hence , it is a natural question to ask if on can `` split '' , given an idempotent @xmath275 , an object of a category @xmath276 in the same way , i.e. @xmath277 the main problem is that the notion of an image of an arrow could not exist in an arbitrary category . the karoubi envelope is an extension of a category such that for a given idempotent @xmath275 the notions @xmath278 makes sense . therefore one can `` split '' a given object in the karoubi envelope that could be indecomposable in the category itself . * objects are ordered pairs @xmath281 of an object @xmath276 and an idempotent @xmath275 of @xmath94 . * morphisms @xmath282 are all arrows @xmath283 of @xmath94 such that the equation @xmath284 holds . * compositions are defined in the obvious sense . the identity of an object is @xmath275 itself . it is straightforward to check that this is indeed a category . we denote an object @xmath281 as @xmath278 , the _ image _ of the idempotent @xmath275 . moreover , we identify the objects of @xmath94 with their image via the embedding functor @xmath285 note that if @xmath94 is pre - additive , then @xmath286 is also an projection and , under the identification above , we can finally write @xmath277 the following proposition is well - known ( see e.g. @xcite ) . the propositions allows us to shift the problem if two chain complexes are homotopy equivalent to the karoubi envelope . recall that @xmath287 denotes the category of formal chain complexes . [ prop - karoubi ] let @xmath288 be two objects , i.e. formal chain complexes , of @xmath287 . if the two objects are homotopy equivalent in @xmath289 , then the two objects are also homotopy equivalent in @xmath287 . we note that we call the idempotents _ `` down and up''_. the reader should be careful not to confuse them with the orientations on the resolutions or the colourings of theorem [ theo - nonalternating ] , i.e. latter colours change at v - crossings but `` down and up '' do not change . note that ( e ) is very important in the following . 1 . @xmath294 and @xmath295 ( idempotent ) . 2 . @xmath296 ( orthogonal ) . 3 . @xmath297 ( complete ) . 4 . @xmath298 and @xmath299 ( eigenvalues ) . @xmath300 and @xmath301 ( change of orientations ) @xmath302=\operatorname{id}(1)_{\mathrm{dot}}=-[\mathrm{u},\phi^-_+]$ ] ( commutator relation ) . now we take a look at the karoubi envelope @xmath304 . the discussion above shows that there is an isomorphism@xmath305{figs / none}}}}\simeq\text{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / down}}}}\oplus\text{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / up}}}}.\ ] ] with this notation we get @xmath305{figs / smoothing}}}}\simeq\text{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / dd}}}}\oplus\text{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / du}}}}\oplus\text{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / ud}}}}\oplus\text{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / uu}}}}\ ] ] and analogous @xmath306{figs / hsmoothing}}}}\simeq\text{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / dd1}}}}\oplus\text{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / du1}}}}\oplus\text{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / ud1}}}}\oplus\text{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / uu1}}}}.\ ] ] recall that the standard orientation for the complex @xmath307 is ( see e.g. figure [ figure0-main ] ) @xmath308[r]{\raisebox{0.2cm}{$\lcurvearrowleft$}}}}}}.\ ] ] to avoid mixing the notions of the down and up - colours and the orientations we denote this complex simply as @xmath309 , i.e. standard orientations for all strings . moreover , under the convention left = first top subscript , right = second top subscript , bottom = first bottom subscript and top = second bottom subscript , a notation like @xmath310 makes sense . [ theo - karoubi ] let @xmath311 denote two homotopy equivalent complexes . then , in @xmath120 , there are sixteen chain homotopies ( only four are illustrated , but it should be clear how the rest look like ) @xmath312{figs / du}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / ud}}}}}\xrightarrow{0}\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / du1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / ud1 } } } } } , & { \llbracket \overcrossing \rrbracket}^{+-}_{-+}\simeq \text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / dd}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / uu}}}}}\xrightarrow{0}\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / dd1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / uu1}}}}},\\ { \llbracket \overcrossing \rrbracket}^{++}_{+-}\simeq \text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / du}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / ud}}}}}\xrightarrow{0}\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / dd1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / uu1 } } } } } , & { \llbracket \overcrossing \rrbracket}^{+-}_{++}\simeq \text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / dd}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / uu}}}}}\xrightarrow{0}\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / du1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / ud1}}}}}. } \end{xy}\ ] ] moreover , similar formulas hold for @xmath313 . we use the observations from above , i.e. in @xmath304 the differential of @xmath309 is a @xmath314-matrix of saddles . hence , for @xmath309 we get ( for simplicity write @xmath315 and @xmath316 and @xmath317 for the saddle under the action of down and up ) @xmath318{figs / dd}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / ud}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / du}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / uu}}}}}\ar[rrr]^{\begin{pmatrix } s_{\mathrm{d } } & 0 & 0 & 0\\ 0 & 0 & 0 & 0\\ 0 & 0 & 0 & 0\\ 0 & 0 & 0 & s_{\mathrm{u } } \end{pmatrix } } & & & \text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / dd1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / ud1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / du1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / uu1}}}}}. } \end{xy}\ ] ] this is true , because all other saddles are killed by the orthogonality relations of the colours down and up . note that both non zero saddles are invertible , i.e. their inverses are the saddles @xmath319 with only @xmath19 as boundary decorations . to see this one uses lemma [ lemm - idem ] and the neck cutting relation . thus we get @xmath320{figs / du}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / ud}}}}}\xrightarrow{0}\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / du1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / ud1}}}}}.\ ] ] to prove the further statements one has to use the relation ( e ) of lemma [ lemm - idem ] , i.e. the only surviving saddles change according to the action of @xmath68 . we note that this is a very important observation . [ bem - importantsaddles ] we observe a very important fact . the four complexes similar to @xmath322[r]{\raisebox{0.2cm}{$\rcurvearrowright$}}}}}},\ ] ] i.e. the ones for which we choose a non - alternating orientation of the resolutions , are the only ones that , due to the orthogonality relation for down and up , survive a closing with an even number of v - crossings , i.e. like @xmath323 in figure [ figure0 - 4 ] , in their source and target . so if we only consider v - link diagrams , then any such saddle is a multiplication , i.e. one bottom component , or a comultiplication , i.e. one top component . therefore such morphisms will always be killed at the bottom ( multiplication ) or at the top ( comultiplication ) . and the only saddle that survives the closure of type @xmath11 in figure [ figure0 - 4 ] is the mbius cobordism . latter one is a @xmath11-morphism if @xmath25 is invertible ( compare to the relations in definition [ defin - category ] ) . hence , the dot - calculus of definition [ defin - dotcalculus ] ensures that all components will survive and the complex will be of the form @xmath324{figs / dd}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.1cm}{\includegraphics[scale=0.5]{figs / uu}}}}}\xrightarrow{0}\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / dd1}}}}}\oplus\text{\raisebox{-0.1cm}{{\raisebox{-0.215cm}{\includegraphics[scale=0.5]{figs / uu1}}}}}.\ ] ] [ theo - leedeg ] let @xmath130 denote a @xmath34-component v - link diagram . then @xmath325 is homotopy equivalent ( in @xmath326 ) to a chain complex with only zero differentials and @xmath41 generators given by the @xmath41 non - alternating resolutions . we will repress the notion of the formal signs of the morphisms to maintain readability . moreover , we will choose a specific orientation for the resolutions . we can do both because of lemma [ lemma - everythingfine ] . so the main part of the proof will be to choose the orientations in a good way and use theorem [ theo - semiloc ] ( the reader may think of tensors ) . recall that the number of crossings is finite . hence , we can choose an orientation of any resolution such that the number @xmath14 of alternating crossings is minimal . we observe that theorem [ theo - semiloc ] can be used because every crossing itself is a nice v - tangle diagram . moreover , with theorem [ theo - karoubi ] , we see that the complex will be homotopy equivalent to a complex with only @xmath11-differentials . hence , the only remaining thing is to show that the number of generators will work out as claimed . note that , if a resolution contains a lower part of a multiplication or a upper part of a comultiplication , then by remark [ bem - importantsaddles ] these resolution is killed . moreover , we can ignore top and bottom parts of @xmath15 , since they will always be non - alternating . now we define the _ dual graph of an resolution _ , denoted @xmath327 , as follows . recall that a resolution is a four valent graph without any c - crossings . any edges of this graph is a vertex of @xmath327 . two vertices are connected with a labeled edge iff they are connected by a v - crossing or a top part of a resolution of a multiplication or a bottom part of a comultiplication . first edges should be labeled @xmath328 , the second type of edges should get an labeling that s corresponds to the given orientation of the resolution . we will work with the simple graph of that type , i.e. remove circles or parallel edges of the same type . see figure [ figure4 - 6 ] , i.e. the figure shows two resolutions from figure [ figure4 - 2 ] and their dual graphs . the advantage of this notation is that the question of surviving resolutions simplifies to the question of a colouring of the dual graph , i.e. a colouring of the dual graph is a colouring with two colours , say red and green , such that every @xmath328-labeled edge has two equally coloured adjacent vertices , every alternating crossing has also two different coloured adjacent vertices , but every non - alternating crossing has also two equal colours at adjacent vertices . then , because of theorem [ theo - karoubi ] , a resolution will have surviving generators iff it does not contain lower parts of multiplication or upper parts of comultiplications and , given an orientation of the resolution , it allows such a colouring . * the dual graph of the resolution is a tree , i.e. no circles . * all circles in the dual graph have an even number of v - labeled edges . * there is one circle in the dual graph with an odd number of v - labeled edges . if @xmath329 , i.e. the resolution is non - alternating , we get exactly the claimed number of generators , since there are , by construction , no lower parts of multiplication or upper parts of comultiplications and the dual graph is of type ( i ) or ( ii ) and in both cases the graph can be coloured . so let @xmath330 and let @xmath331 be an alternating crossing in a resolution @xmath54 . as in remark [ bem - importantsaddles ] the whole resolution is killed if the @xmath331 is a lower part of a multiplication or an upper part of a comultiplication . hence , we can assume that all alternating crossings of @xmath54 are either top components of multiplications or bottom components of comultiplications . so we only have to check the three cases from above . if the resolution is one of type ( i ) , then it is possible to choose the orientations in such a way that all crossings are non - alternating , i.e. this would be a contradiction to the minimality of @xmath14 . if the resolution is of type ( ii ) , then the resolution only survives , i.e. the dual graph allows a colouring , iff the number of other alternating crossings in every circle is even . but in this case one can also choose an orientation with only non - alternating crossings . hence , we would get a contradiction to the minimality of @xmath14 again . hence , only non - alternating resolutions generate non vanishing objects . moreover , any non - alternating resolution will create exactly two of these generators . thus , with theorem [ theo - nonalternating ] the statement follows . * if @xmath334 , then their is an isomorphism @xmath335 where @xmath336 is all torsion . * if @xmath241 or @xmath230 $ ] , then their is an isomorphism @xmath337 * the only possible torsion is @xmath25-torsion . for ( a)+(c ) recall the universal coefficients theorem . i.e. their is a short exact sequence @xmath338 therefore ( a ) follows from ( b ) with @xmath241 , since the tor - functor will vanish in this case . and ( c ) follows from ( b ) with @xmath230 $ ] , since all non 2-torsion elements will exits for @xmath10 and @xmath339 $ ] . d. bar - natan , khovanov s homology for tangles and cobordisms , geom . topol . 9 ( 2005 ) , 14431499 . arxiv : math/0410495 d. bar - natan , on khovanov s categorification of the jones polynomial , algebraic and geometric topology 2 ( 2002 ) , 337370 . arxiv : math/0201043 d. bar - natan , fast khovanov homology computations , journal of knot theory and its ramifications 16 - 3 ( 2007 ) , 243255 . arxiv : math/0606318 d. bar - natan and s. morrison , the karoubi envelope and lee s degeneration of khovanov homology , algebraic and geometric topology 6 ( 2006 ) , 14591469 arxiv : math/0606542v3 p. freyd , _ abelian categories . an introduction to the theory of functors _ , new york : harper and row publishers , 1964 ( harpers series in modern mathematics ) . v. jones , planar algebras i , eprint ( 1999 ) . arxiv : math/9909027 l. kauffman , introduction to virtual knot theory , j. knot theory ramifications 21,13 ( 2012 ) . l. kauffman and v. manturov , virtual knots and links , proc . steklov inst . math . 1 ( 2006 ) , 104121 . arxiv : math/0502014 m. khovanov , a categorification of the jones polynomial , duke math . j. 101 ( 2000 ) , 359426 . arxiv : math/9908171 j. kock , _ frobenius algebras and 2d topological quantum field theories _ , cambridge : cambridge university press , 2003 . lee , an endomorphism of the khovanov invariant , adv . 197(2 ) ( 2005 ) , 554586 . arxiv : math/0210213 m. mackaay , p. turner and p. vaz , a remark on rasmussen s invariant of knots , knot theory and its ramifications 16(3 ) ( 2007 ) , 333344 . arxiv : math/0509692 v. manturov , the khovanov complex for virtual links , fundam . prikl . mat . 11:4 ( 2005 ) , 127152 . arxiv : math/0501317 v. manturov , khovanov homology for virtual knots with arbitrary coefficients , izv math 71 ( 2007 ) , 967999 . arxiv : math/0601152 j. rasmussen , khovanov homology and the slice genus , invent . 182(3 ) ( 2010 ) , 419447 . arxiv : math/0509692 s. mac lane , _ categories for the working mathematician _ , new york : springer - verlag , 1998 . d. tubbenhauer , khovanov homology for virtual links using cobordisms , preprint ( 2011 ) . arxiv:1111.0609v2 d.t . : _ courant research center `` higher order structures '' , university of gttingen , gttingen , germany ; mathematisches institut , georg - august - universitt gttingen , gttingen , germany _ email : dtubben@uni-math.gwdg.de

there may be around @xmath1 stellar - mass black holes ( bhs ) in the galaxy ( e.g. , * ? ? ? * ; * ? ? ? * ) , but only two dozen of them have been dynamically confirmed in x - ray binary ( xrb ) systems , in which a bh is accreting from its companion star ( * ? ? ? * ; * ? ? ? * for recent reviews ) . a large fraction of the bh xrbs are close binaries with orbital periods @xmath2 day . the spectral types of the companion stars range from a2v to m1v , indicating that most of them are less massive than @xmath3 , so these xrbs are correspondingly classified as low - mass x - ray binaries ( lmxbs ) . all known bh lmxbs are transient systems , spending most of their time in quiescence with x - ray luminosities below @xmath4 , and occasionally exhibiting outbursts when the x - ray luminosities can rise up to @xmath5 . galactic bh lmxbs are believed to evolve from primordial binaries consisting of a massive star and a low- or intermediate - mass companion ( * ? ? ? * for a review ) . the standard formation scenario involves a common envelope ( ce ) phase , during which the binary s orbital separation reduces drastically due to the loss of the orbital kinetic energy @xcite , since the orbital separations of the present day binaries are orders - of - magnitude smaller than the radii that the massive primaries can reach during their evolution . when the primary star evolves and overflows its roche lobe , mass transfer proceeds on a dynamical timescale because the primary is much more massive than the secondary . the secondary can not accrete all the mass transferred onto it , and an envelope forms enshrouding the two stars . then the secondary star spirals into the envelope , and the orbital energy is used to expel the envelope ( for reviews of ce evolution , see * ? ? ? * ; * ? ? ? if surviving the ce event , the binary can continue to evolve ; otherwise the two stars would merge to be a single star . though the ce stage is too short - lived to be detected , indirect observational evidence has emerged ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? bhs are thought to be the evolutionary outcomes of sufficiently massive stars ( @xmath6 ; e.g. , * ? ? ? * ; * ? ? ? . however , the envelope of such stars are generally too tightly bound for a low - mass secondary to expel @xcite , thus a merger is more likely to take place , unless the primary has lost a significant fraction of its envelope before the ce phase through stellar winds @xcite , or abnormally high values for the ce ejection efficiency are adopted @xcite . an alternative scenario is that the secondaries in bh lmxbs are initially intermediate - mass stars , which can eject the primary s envelope more easily . observational evidence for this scenario comes from the cno - processed material in xte j1118@xmath7480 from its ultraviolet spectra @xcite . it is , however , difficult for an intermediate - mass x - ray binary ( imxb ) to evolve to short - period systems , unless some unusual angular momentum loss mechanisms are invoked ( e.g. , * ? ? ? * ; * ? ? ? meanwhile , these binaries reveal a discrepancy between the calculated effective temperatures and the observed spectral types of the donor stars . the reason lies in that nuclear burning in the center of an intermediate - mass star causes a higher effective temperature than an ordinary low - mass star , though it has lost a significant fraction of its mass during the evolution @xcite . clues to the initial mass range for the donors in bh lmxbs can be derived from the spin evolution of the bhs , since the bh spins increase with accretion @xcite . two methods , i.e. , x - ray continuum - fitting @xcite and modeling relativistic reflection @xcite have been employed to estimate the spins of about 20 bhs ( * ? ? ? * ; * ? ? ? * and references therein ) . assuming that bhs in lmxbs are born with negligible spins , @xcite showed that the initial masses of the secondaries in short - period binaries are generally @xmath8 . from another point of view , approaches have been taken in the study of ce evolution by refining the energy budget for the ejection as well as the core - envelope binding energy . for example , @xcite argued that nuclear fusion should be added into the energy sources that may contribute to the envelope ejection . @xcite proposed that a thermonuclear runaway which is caused by the mixing hydrogen from the secondary into the helium - burning shell of the primary would assist the ejection . when treating the binding energy of the envelope , @xcite pointed out that there might be mass outflows during the slow spiral - in stage , when the orbital energy is balanced with the enthalpy rather than merely the internal energy of the envelope . that treatment can lower the binding energy by a factor of @xmath9 , and help allow for a low - mass companion to survive the ce phase . it is also noted that the value of the binding energy is sensitive to the definition of the boundary between the remnant core and the ejected envelope ( e.g. , * ? ? ? @xcite advocated that , the he core of a giant would experience a thermal readjustment stage after the ce event . thus the boundary between the remnant core and the ejected envelope can be determined as the place where the hydrogen burning shell had maximal compression prior to ce evolution . another issue in the formation of bh lmxbs is the mass range for the bh progenitors . current supernova ( sn ) theories can not reliably predict which stars form bhs instead of neutron stars ( nss ) , although they are generally thought to be more massive than @xmath10 . the bh masses in the 24 xrbs have been dynamically measured to be in the range of @xmath11 to @xmath12 ( * ? ? ? * and references therein ) . however , there is statistical evidence for the presence of a dearth of nss or bhs with masses @xmath13 @xcite . this is in contrast with the traditional thought that the distribution of bh masses should decay with mass ( e.g. , * ? ? ? * ; * ? ? ? * ) , suggesting that the physics of sn explosions that lead to the formation of bhs is still unclear @xcite . recently , @xcite argued that zero - age main sequence stars with mass @xmath14 might experience an unsuccessful explosion and evolve to bhs , while stars with mass around @xmath15 experience a successful sn explosion and become nss . though this bifurcation is not well understood and may be related to the compactness parameter that describes the density profile outside the iron core @xcite , the failed sn mechanism can naturally account for the observed @xmath16 gap between nss and bhs . @xcite proposed that this mechanism can interpret the red supergiant problem and the sn rate problem , which concerns the absence of red supergiants with masses @xmath17 as the progenitors of type iip sne , and the deficit of the observed cosmic sn rate compared to the observed cosmic star formation rate , respectively . in this paper we investigate the formation of galactic bh lmxbs incorporating the influence of the aforementioned factors with both stellar evolution and binary population synthesis ( bps ) methods . in section 2 , we calculate and compare the binding energy parameters in various models , which are then used in our bps study . in section 3 , we introduce the assumptions and prescriptions adopted in our bps calculations . the results in various kinds of models are presented and compared with observations in section 4 . we summarize the results and discuss their implications in section 5 . in the standard ce model the change in the binary s orbital energy @xmath18 is responsible for the ejection of the ce , which is described by the following equations @xcite , @xmath19 and @xmath20 where @xmath21 is the binding energy of the envelope , @xmath22 the efficiency of converting the orbital energy to kinetic energy to eject the ce , @xmath23 the gravitational constant , @xmath24 the primary s mass and @xmath25 its core mass , @xmath26 the mass of the secondary , and @xmath27 and @xmath28 the orbital separations just before and after the ce phase , respectively . when only gravitational binding energy is considered , @xmath29 is given by @xmath30 if the internal energy of the stellar matter ( including both thermal energy and recombination energy ) also contributes to the binding energy , then we have @xmath31 dm,\ ] ] where @xmath32 is internal energy @xcite . recently , @xcite proposed a modified energy criterion taking into account the mass outflows during the spiral - in stage . in this so - called enthalpy model , the binding energy is expressed as @xmath33 dm,\ ] ] where @xmath34 and @xmath35 are the pressure and the density of the gas , respectively . since the @xmath36 term is non - negative , the absolute value of @xmath21 becomes smaller in this case . for convenience , @xcite introduced a parameter @xmath37 characterizing the central concentration of the donor s envelope to describe the binding energy : @xmath38 where @xmath39 is the mass of the donor star s envelope , and @xmath40 is the ratio of the roche lobe radius and the orbital separation at the onset of ce . generally , once a star fills its roche lobe , @xmath41 is taken to be the stellar radius . combining eqs . ( [ equation3])([equation6 ] ) , one can obtain different @xmath37s from the gravitational binding energy , the total energy and the enthalpy prescriptions , namely @xmath42 , @xmath43 and @xmath44 , respectively . we calculate their values using an updated version ( 7624 ) of the modules for experiments in stellar astrophysics ( mesa ) @xcite evolution code . we consider stars with initial masses in the range of @xmath45 , which securely covers the range of the stellar - mass bh progenitors ( more massive stars produce negligible bhs according to the initial mass function and will evolve to wolf - rayet ( wr ) stars with almost all the hydrogen envelopes lost due to strong stellar winds ) . the chemical composition of the stars is taken to be @xmath46 and @xmath47 . the mixing length parameter @xmath48 is set to be 2.0 . for massive stars , stellar winds may have non - ignorable influence on the @xmath37 parameter @xcite , but their loss rates have not been accurately determined . here , we adopt two prescriptions for the wind loss rate . the first one is taken from @xcite and @xcite for o and b stars in different stages , denoted as wind1 hereafter , while the other one , denoted as wind2 , adopts the maximum value of the prescriptions above in all the evolutionary stages , to be consistent with @xcite , and to set an upper limit for the influence of the stellar wind . the effect of stellar rotation is ignored , because in our calculation for lmxb formation , ce evolution usually occurs during case c mass transfer when the primary star has evolved to be a supergiant star with very slow rotation due to stellar evolution and/or tidal synchronization . our calculations show that the evolution of @xmath44 , @xmath43 , and @xmath42 generally have the similar trend in most of the evolutionary stages , i.e. , decreasing with increasing stellar radius . however , for stars less massive than @xmath49 , @xmath37s increase again when they have ascended the asymptotic giant branch and developed a deep envelope ( see also * ? ? ? as an illustration , fig . [ figure1 ] shows different @xmath37 values as a function of the stellar radius @xmath50 for a @xmath51 star , and fig . [ figure2 ] shows the @xmath44 values as a function of the stellar radius for stars with different masses . the solid and dashed lines correspond to the wind1 and wind2 prescriptions , respectively . in fig . [ figure1 ] the red , blue , and green lines describe the evolution of @xmath44 , @xmath43 , and @xmath42 respectively in each case . it is seen that the values of @xmath37 with the wind1 prescription are about three times as large as those with the wind2 prescription ( the binding energy with the wind1 prescription is half of the value with the wind2 prescription considering different envelope masses ) , indicating that the binding energy parameter for massive stars is sensitive to wind loss @xcite . we plot the core mass against the stellar radius in fig . [ figure3 ] , and list the stellar parameters along the evolutionary tracks in table [ table1 ] in the two cases . we also use an updated stellar evolution code called ev originally developed by @xcite to calculate the values of @xmath37s , and find that they are several times smaller than those with the mesa code in the wind2 case ( see also * ? ? ? * ; * ? ? ? the reason for this difference is that stars are generally more compact ( especially for stellar structure within the hydrogen burning shell ) during late evolution when modeled with the ev code . we choose to use the results with the mesa code in the following because it adopts much denser grids for stellar structure than the ev code . @xcite pointed out that a low - mass ( @xmath52 ) star is unable to eject the bh progenitor s envelope unless @xmath53 . in our calculation , this condition could be achieved for both @xmath44 and @xmath43 if the masses of the bh progenitors are @xmath54 . binaries with more massive progenitor stars are unable to survive ce evolution even if @xmath44 is adopted ( see also fig . [ figure2 ] ) . we use an updated version of the bps code developed by @xcite to study the formation of bh lmxbs . the code originates from the rapid single - star evolution code written by @xcite . it involves various kinds of processes such as mass transfer , accretion , ce evolution , sn kicks , tidal interactions , and angular momentum loss . we have made quite a few modifications of the bps code to follow the formation and evolution of xrbs @xcite . we briefly describe some considerations adopted in this work . given the binding energy parameter @xmath37 , we can estimate the post - ce separation with the following equation : @xmath55 if neither the he core nor the secondary star fills its roche lobe , the binary is regarded to survive the ce phase and can continue to evolve . generally the value of @xmath22 is less than unity , and there is indication that the ce efficiency could be low @xcite . moreover , it may vary with stellar structure and the binary parameters . indeed , it has been suggested that @xmath22 may depend on the component mass and the orbital period in the studies of the formation of post - ce white dwarf binaries ( e.g. , * ? ? ? * ; * ? ? ? * ) . however , it is not known whether this can be applied to more massive binaries , so we still adopt constant values ( 1 and 0.2 ) for @xmath22 . it is noted that @xmath22 and @xmath37 are usually combined to affect ce evolution . besides ce evolution , the sn explosions play an important role in the formation of bh lmxbs . the sn mechanisms are relevant to both the formation rate of xrbs and the mass distribution of the remnant bhs . when determining the bh masses @xmath56 , one needs to account for the measured masses of known bhs and the @xmath57 gap between the ns and bh masses without any fine - tuning of stellar mass loss . we consider both the prescriptions based on the neutrino - driven convection - enhanced sn mechanism , which is called the rapid sn mechanism , and the failed sn mechanism . in the former case the gravitational remnant mass is described by @xcite , @xmath58 here @xmath59 is the mass of the proto - compact object and @xmath60 with @xmath61 where @xmath62 is the co core mass , @xmath63 , and @xmath64 . the latter prescription is under the assumption that the bh formation is controlled by the compactness of the stellar core at the time of collapse : low compactness stars are more likely to explode as sne and produce nss , while high - compactness stars are more likely to evolve to failed sne that produce bhs @xcite . in this case we assume @xmath65 or @xmath62 , where @xmath66 is the he core masses prior to core collapse ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? the value of @xmath66 is determined as follows ( * ? ? ? * for a detailed description ) . first , the core mass @xmath67 at the end of the hertzsprung - gap is set according to the initial mass of the star ( see eq . 28 of * ? ? ? * ) . then the evolution of the core mass during the giant branch is determined by combining the power - law core mass - luminosity relation @xmath68 where @xmath69 , @xmath70 , and @xmath71 and @xmath72 depend on the stellar mass , and the energy conservation equation for hydrogen burning @xmath73 where @xmath74 is the envelope mass fraction of hydrogen and @xmath75 is the specific energy release . thus @xmath76 where @xmath77 represents hydrogen rate constant . a similar procedure is adopted to calculate the co core mass , with @xmath78 replaced by @xmath79 myr@xmath80 . in all of the cases we require that @xmath81 . we do not consider the formation of bhs by accretion - induced collapse of nss in xrbs . we also assume that a natal kick " is imparted on the newborn bhs , similar as nss . the kick velocity is set to be inversely proportional to the remnant mass , i.e. , @xmath82 , where @xmath83 is the kick velocity for nss , which follows a maxwellian distribution with @xmath84 @xcite . we consider the bh lmxbs evolved from primordial binaries with both incipient low - mass and intermediate - mass secondaries . we assume a constant star formation rate of @xmath85yr@xmath80 @xcite , and evolve @xmath86 primordial binaries with the primary mass ( @xmath24 ) ranging from @xmath87 to @xmath88 following the @xcite initial mass function , and the secondary mass ( @xmath26 ) from @xmath89 to @xmath90 . for the initial mass ratio @xmath91 , we consider both flat distribution @xmath92 with @xmath93 randomly distributed between 0 and 1 , and a twin "- like distribution with @xmath94 . the initial separation @xmath95 of the binary components varies between @xmath96 and @xmath97 , with a flat distribution for @xmath98 . all binaries are assumed to be initially in circular orbits . we stop the calculation when the donor mass in lmxbs drops down to @xmath89 , or the evolution time exceeds 15 gyr . we summarize the models and the prescriptions employed in table . [ table2 ] . we first consider the formation of bhs via the rapid sn mechanism . in this case the progenitor stars are usually more massive than @xmath99 in binaries . for such massive stars the binding energy of the envelope is so large that with @xmath100 a low - mass secondary inevitably merges with the he core of the primary , and no successful bh lmxbs can form ( see also * ? ? ? so in models a1 and a2 we set @xmath101 with @xmath102 and @xmath44 , respectively . the distributions of the donor mass of the resultant bh xrbs at current epoch are plotted in the left and right panels of fig . [ figure4 ] , respectively . to reveal the donor mass distribution in different scale , the donor masses are confined within @xmath103 and @xmath104 in the upper and lower panels of the figure , respectively . the solid and dashed red lines represent the results with the @xmath37s obtained with the wind1 and wind2 prescriptions , respectively . also displayed is the distribution of the observed bh lmxbs with the black line ( data taken from * ? ? ? * ; * ? ? ? the number of the bh xrbs is calculated by multiplying the formation rate with the duration in a specific stage . we only display the sources which are subject to the thermal - viscous instability in an x - ray irradiated accretion disk and assumed " to be transient @xcite . they include almost all bh lmxbs produced . the exceptions are the systems comprised of intermediate - mass donors or with extremely short periods , which are not relevant here . in model a1 and a2 , the donors of persistent sources locate in the mass range of @xmath105 , and their total number is comparable with that of transient sources . when @xmath106 , most bh xrbs harbor an intermediate - mass donor star ( with mass @xmath107 ) , implying that lower - mass secondaries are unlikely to survive ce evolution . the corresponding formation rate of the bh xrbs is too low ( @xmath108 yr@xmath80 ) to account for the observed systems in the galaxy . when @xmath109 , the formation rate enhances a bit ( @xmath110 @xmath111 ) , and the donor mass distribution peaks at around @xmath112 . it is seen that in both cases the distributions significantly deviate from the observed one . the calculated orbital period distributions are too long ( @xmath113 1 day ) to be consistent with observations ( see fig . [ figure12 ] below ) . these results clearly disfavors the rapid sn mechanism , so in the following we only consider the failed sn mechanism . in models b1 and b2 , we assume @xmath114 and a flat mass ratio distribution with @xmath115 and @xmath116 , respectively . we present the calculated distributions of the donor mass for the two models and compare them with observations in fig . [ figure5 ] ( left : b1 ; right : b2 ) . here the red and blue lines correspond to @xmath101 and 0.2 , respectively , and the line styles are same as in fig . [ figure4 ] . with the failed sn mechanism the masses of the bh progenitors can be lower than @xmath117 , so it is easier for low - mass secondaries to survive the ce phase . from the left panel of fig . [ figure5 ] we see that in the cases of wind1/@xmath101 , wind2/@xmath101 , and wind1/@xmath118 the donor mass peaks at @xmath119 , matching the observational data , while in the case of wind2/@xmath118 , the donor masses cluster around @xmath120 , larger than the observed peak . this is simply because more massive donor stars are required to survive the ce evolution when the product @xmath121 is lower . however , larger @xmath122 does not always necessarily lead to larger number of bh lmxbs , since the formation of a bh xrb strongly depends on the orbital separation just before and after the ce phase . with the same initial orbital periods , larger @xmath122 results in wider post - ce orbits , which may cause the subsequent roche lobe overflow to be postponed , reducing the x - ray lifetime and thus the number of xrbs . figure [ figure6 ] shows the birthrates of the bh xrbs as a function of the initial donor mass in each case . we mention that , although the bh xrb birthrates can be as high as @xmath123 yr@xmath80 , the birthrates of those with incipient low - mass donors are several times lower . we use the kolmogorov - smirnov ( ks ) test to quantitatively compare the the univariate distributions of the calculated and measured donor mass . the ks statistic @xmath72 quantifies a distance between the empirical distribution functions of the two samples . the null hypothesis that the two distributions are drawn from the same one is rejected at a significance level @xmath124 if @xmath125 @xcite . in table [ table2 ] we list the calculated values of @xmath72 and @xmath126 for all models ( except models a1 and a2 because of too few data points ) in the case of wind2/@xmath101 at @xmath127 and 0.05 . figure [ figure7 ] shows the donor mass distributions for models c1 ( left panel ) and c2 ( right panel ) under the assumptions that @xmath128 and @xmath109 . in models c1 and c2 we adopt a flat distribution of the initial mass ratio and a twin " distribution ( * ? ? ? * ; * ? ? ? * see however , cantrell & dougan 2014 ) , respectively . here we confine the donor masses within the range of @xmath129 because of the small number of xrbs with more massive donor stars . figure [ figure7 ] shows that the peak of the donor mass distribution in the cases of wind1/@xmath101 , wind2/@xmath101 , and wind1/@xmath130 is @xmath131 , but shifts towards @xmath132 in the case of wind2/@xmath118 , roughly in accordance with fig . [ figure5 ] . this is also reflected by their similar @xmath72 values . to see whether the resultant bh lmxbs evolve from incipient low- or intermediate - mass binaries , we repeat the calculation with the initial donor mass confined to be @xmath133 and @xmath134 respectively . the results in model c1 ( with the wind2 prescription ) are plotted in fig . [ figure8 ] . it is clearly seen that the total number of the bh lmxbs evolved from the primordial binaries with the initial donor mass @xmath135 is almost the same as in the left panel of fig . [ figure7 ] , so we can safely conclude that almost all the bh lmxbs in our calculation have incipient low - mass secondaries . although the bh xrbs with low - mass donors have a lower birthrate than those with intermediate - mass donors , it is compensated by their long x - ray lifetimes . figure [ figure7 ] also indicates that assuming a @xmath94 distribution ( model c2 ) leads to less efficient formation of bh lmxbs than with a flat @xmath93 distribution because of less progenitor binaries with low - mass secondary stars . figure [ figure9 ] shows the distribution of the orbital period vs. the donor mass ( left panel ) , and the distribution of the mass transfer rate vs. the donor mass ( right panel ) in model c1 . the color scale represents the relative number of the xrbs in a parameter space . it can be seen that the majority of bh xrbs are short - period ( @xmath136 day ) systems with mass transfer rates below @xmath137 . the above results are based on the hypothesis that the bh mass is equal to its progenitor s he core mass at the time of core bounce . to check the effect of another choice of the remnant mass , we construct models d1 and d2 , which are same as models c1 and c2 except with @xmath138 , respectively . the donor mass distributions are shown in fig . [ figure10 ] . we find that the formation rate in mode d1 decreases to about half that in model c1 , and the peak of the donor mass distribution extends to a broader range compared with fig . [ figure6 ] . the remnant mass influences the formation of the bh lmxbs by determining the roche lobe size after the sn explosion . figure [ figure11 ] shows the distributions of the bh masses under the two choices . the red and blue lines represent the results in models c1 and d1 , which peak at @xmath139 and @xmath140 , respectively . @xcite have analyzed the data of 16 bh x - ray transients , and found that the observed bh masses can be best described by a narrow distribution at @xmath141 , which is obviously more compatible with the result in model c1 . in figs . [ figure12 ] we show the calculated distributions of the orbital period and the effective temperature of the donor stars in model a. the color scale represents the relative number of the xrbs . we also plot the measured orbital periods and effective temperatures of the donors for galactic bh lmxbs . in fig . [ figure12 ] for models a1 ( left panel ) and a2 ( right panel ) , the observed distribution obviously deviates from the calculation , showing a tendency towards lower temperature . the reason is that , as stated before , most of the bh lmxbs in models a1 and a2 descended from imxbs , so the donor star is likely to have a higher temperature than an originally low - mass star . we get similar results for model b2 with @xmath142 . in other models the calculated orbital periods are more compatible with observations . we also use the ks test to quantitatively compare the calculated and observed distributions , and the results for models b1 , b2 , c1 , and d1 are shown in table [ table3 ] . we can see that , except model b2 with @xmath118 , other models are all acceptable . we show the distributions in the best - fit model c1 in fig . [ figure13 ] . previous studies have already demonstrated that it is hard to produce compact bhxrbs consisting of an initially low - mass companion with the standard formation scenario , since the companion star is unable to expel the ce ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * see however , yungelson & lasota 2008 ; wiktorowicz et al . 2013 ) . however , the orbital periods , the donors spectral types , and the measured spins of quite a few bhs strongly suggest that in the incipient binaries the donor stars are likely to be of low - mass . we construct a series of models to explore the plausible solutions to the above - mentioned puzzle . the factors we have taken into account include the binding energy parameter @xmath37 , the ce efficiency @xmath22 , the sn mechanisms for bh formation , and the distribution of the mass ratio . by adopting different choices of these prescriptions , we examine their roles in the formation of bh lmxbs by comparing the calculated parameter distributions with observations . our calculations show that the @xmath37-values strongly depend on the stellar mass and the evolutionary stage . generally @xmath44 is several times larger than @xmath42 and @xmath43 , and this can help the binary to avoid merging during the ce evolution without invoking unphysically high values of @xmath22 . however , we fail to reproduce the observational characteristics of galactic bh lmxbs based on the rapid sn mechanism even with @xmath44 adopted , since the bh progenitors are still too massive ( @xmath143 ) in this scenario . we then consider the alternative failed sn mechanism , and assume that the remnant bh mass equals the he or co core mass of the primary . this time we obtain more satisfactory results , with the distributions of the donor mass , temperature and orbital period for the current binaries largely consistent with observations . in particular , most bh lmxbs in this case originate from primordial binaries with incipient low - mass ( @xmath133 ) secondary stars , compatible with the requirement derived from the spin estimates of several bhs in lmxbs . finally , all models predict that most bh lmxbs are transient sources due to thermal instability in the accretion disks . thus , in the framework of the failed sn mechanism , it is possible to reproduce the galactic bh lmxbs with normal value of @xmath144 . realizing the uncertainties in the treatment of ce evolution , we also check the results by adopting a small @xmath118 in each case . our results demonstrate that , when @xmath122 is large enough for a low - mass star to survive the ce phase , the donor mass distribution is not sensitive to the choice of @xmath37 or @xmath22 , which affects the number and birthrate of the produced bh lmxbs . however , for very small @xmath122 , the donor mass distribution strongly depends on the value of @xmath37 or @xmath22 , because surviving the ce phase becomes paramount and difficult . in addition , the prescription of the bh mass with the he core mass leads to better match of the observations than with the co core mass . our results provide useful constraints on the formation process of bh lmxbs , but are subject to many uncertainties in both observational and theoretical aspects . when comparing the calculated results with observations , one needs to be cautious that this comparison is influenced by small number statistics as well as selection effects . currently there are around 20 dynamically confirmed stellar - mass bhs . it is not known whether they are correctly representative of the overall population in the galaxy . most bh lmxbs are transient sources , and we are lack of thorough understanding of their outburst characteristics ( including the peak luminosity and its relation with other parameters , and the duty cycles ) and the physical mechanisms @xcite . to make things more complicated , a new class of very faint x - ray transients were recently discovered with peak luminosities of only @xmath145 ergs@xmath80 @xcite . further more , considering the selection effect that luminous sources are more likely to be observed , we may expect that a large fraction of the lmxbs possess even lower - mass donor stars than the current sample , because they are relatively dim in x - rays even during outbursts . there are also big uncertainties in modeling the evolutionary sequences of compact star binaries . we take into account @xmath146 in calculating the binding energy of the envelope ( see also * ? ? ? * ) , but its credibility is controversial and there are arguments both for and against it in the literature , let alone the definition of the core - envelope boundary from which the binding energy is integrated ( see * ? ? ? * for a discussion ) . it is also highly uncertain whether the ce efficiency @xmath22 is constant or depends on other physical parameters . @xcite and @xcite suggested that it may decrease with increasing mass ratio , but this effect has not been confirmed in modeling the sdss post - ce binary population ( e.g. , * ? ? ? * ) and needs further verification . even if true , the physics behind it is not clear , and numerical simulations in 3d of ce evolution are eagerly required to resolve this issue . as to the sn and the bh formation mechanisms , our results base on the suggestion that stars of mass @xmath147 die in failed sne leaving a bh @xcite . the model still needs to be tested by both 3d numerical simulations of the collapse of massive stars and targeted field surveys for sne in nearby galaxies . obviously a thorough investigation on the evolutionary history of bh lmxbs needs to incorporate bps with the aforementioned input physics in a self - consistent way . we are grateful to an anonymous referee for helpful comments . this work was supported by the natural science foundation of china under grant numbers 11133001 and 11333004 , and the strategic priority research program of cas ( under grant number xdb09000000 ) . 99 natexlab#1#1 bailyn , c. d. , jain , r. j. , coppi , p. , & orosz , j. a. 1998 , , 499 , 367 belczynski , k. , wiktorowicz , g. , fryer , c. l. , holz , d. e. , & kalogera , v. 2012 , , 757 , 91 brown , g. e. , & bethe , h. a. 1994 , , 423 , 659 cantrell , a. g. , & dougan , t. j. 2014 , , 445 , 2028 casares , j. & jonker , p. g. 2014 , spa . , 183 , 223 chen , w .- c . , & li , x .- d . 2006 , , 373 , 305 clausen , d. , piro , a. l. , & ott , c. d. 2015 , , 799 , 190 coriat , m. , fender , r. p. , & dubus , g. 2012 , , 424 , 1991 davis , s. w. , blaes , o. m. , hubeny , i. , & turner , n. j. 2005 , , 621 , 372 davis , p. j. , kolb , u. , & knigge , c. 2012 , , 419 , 287 de kool , m. 1990 , , 358 , 189 de marco , o. , passy , j .- c . , moe , m. , herwig , f. , mac low m .- m . , paxton , b. 2011 , , 411 , 2277 drake , j. j. , & sarna , m. j. 2003 , , 594 , 55 eggleton , p. , , 151 , 351 eggleton , p. , , 156 , 361 fabian , a. c. , parker , m. l. , wilkins , d. r. , et al . 2014 , , 439 , 2307 fabian , a. c. , rees , m. j. , stella , l. , & white , n. e. 1989 , , 238 , 729 farr , w. m. , sravan , n. , cantrell , a. et al . 2011 , , 741 , 103 feigelson , e. d. , & jogesh , b. g. 2012 , modern statistical methods for astronomy , cambridge university press ( cambridge , uk ) fragos , t. , & mcclintock , j. e. 2015 , , 800 , 17 fryer , c. l. 1999 , , 522 , 413 fryer , c. l. , belczynski , k. , wiktorowicz , g. et al . 2012 , , 749 , 91 fryer , c. l. , & kalogera , v. 2001 , , 554 , 548 han , z. , podsiadlowski , p. , & eggleton , p. , , 270 , 121 han , z. , podsiadlowski , p. , maxted , p. f. l. , marsh , t. r. , & ivanova , n. 2002 , , 336 , 449 han , z. , podsiadlowski , p. , maxted , p. f. l. , & marsh , t. r. 2003 , , 341 , 669 haswell , c. a. , hynes , r. i. , king , a. r. , & schenker , k. 2002 , , 332 , 928 hobbs , g. , lorimer , d. r. , lyne , a. g. , & kramer , m. 2005 , , 360 , 974 horiuchi , s. , nakamura , k. , takiwaki , t. , kotake , k. , tanaka , m. 2014 , , 445 , l99 hurley , j. r. , pols , o. r. , & tout , c. a. 2000 , , 315 , 543 hurley , j. r. , tout , c. a. , & pols , o. r. 2002 , , 329 , 897 ivanova , n. 2002 , phd thesis , oxford university ivanova , n. 2006 , , 653 , l37 ivanova , n. 2011 , , 730,76 ivanova , n. , & chaichenets , s. 2011 , , 731 , l36 ivanova , n. , justham , s. , chen , x. , et al . 2013 , a&ar , 21 , 59 justham , s. , rappaport , s. , & podsiadlowski , p. 2006 , , 366 , 1415 kalogera , v. 1999 , , 521 , 723 kiel , p. d. , & hurley , j. r. 2006 , , 369 , 1152 kiziltan , b. , kottas , a. , de yoreo , m. , & thorsett , s. e. 2013 , , 778 , 66 kobulnicky , h. a. , & fryer , c. l. 2006 , , 670 , 747 kochanek , c. s. 2014 , , 785 , 28 kochanek , c. s. 2015 , , 446 , 1213 kreidberg , l. , bailyn , c. d. , farr , w. m. , & kalogera , v. 2012 , , 757 , 36 kroupa , p. , tout , c. a. , & gilmore , g. 1993 , , 262 , 545 lasota , j .- p . 2001 , , 45 , 449 lasota , j .- p . , dubus , g. , & kruk , k. 2008 , , 486 , 523 li , x .- d . 2015 , , 64 - 66 , 1 mcclintock , j. e. , narayan , r. , & steiner , j. f. 2014 , space sci . , 183 , 295 oconnor , e. , & ott , c. d. 2011 , , 730 , 70 ohlmann , s. t. , rpke , f. k. , pakmor , r. , & springel , v. 2016 , , 816 , l9 zel , f. , psaltis , d. , narayan , r. , & mcclintock , j. e. 2010 , , 725 , 1918 zel , f. , psaltis , d. , narayan , r. , & santos villarreal , a. 2012 , , 757 , 55 paczynski , b. 1976 , in iau symposium , vol . 73 , structure and evolution of close binary systems , ed . p. eggleton , s. mitton , & j. whelan ( dordrecht , d. reidel publishing co. ) , p. 75 paxton , b. , bildsten , l. , dotter , a. , et al . 2011 , , 192 , 3 paxton , b. , cantiello , m. , arras , p. , et al . 2013 , , 208 , 4 paxton , b. , marchant , p. , schwab , j. , et al . 2015 , , 220 , 15 pinsonneault , m. h. , & stanek , k. z. 2006 , , 639 , l67 podsiadlowski , p. , rappaport , s. , & pfahl , e.d . 2002 , , 565 , 1107 podsiadlowski , p. , rappaport , s. , & han , z. 2003 , , 341 , 385 podsiadlowski , p. , ivanova , n. , justham , s. , & rappaport , s. 2010 , , 406 , 840 portegies zwart , s. f. , verbunt , f. , & ergma , e. 1997 , , 321 , 207 remillard , r. a. , & mcclintock , j. e. 2006 , ar , 44 , 49 shao , y. , & li , x .- d . 2014 , , 796 , 37 shiode , j. h. , & quataert , e. 2014 , , 780 , 96 sion , e. m. , bond , h. e. , lindler , d. , et al . 2012 , , 751 , 66 smith , l. f. , biermann , p. , & mezger , p. g. 1978 , , 66 , 65 smith , n. , & arnett , w. d. 2014 , , 785 , 82 smith , n. , li , w. , filippenko , a. v. , & chornock , r. 2011 , , 412 , 1522 sukhbold , t. & woosley , s. e. 2014 , , 783 , 10 taam , r. e. & sandquist , e. l. 2000 , ar , 38 , 113 tauris , t. m. , & dewi , j. d. m. 2001 , , 369 , 170 thorne , k. s. 1974 , , 191 , 507 timmes , f. x. , woosley , s. e. , weaver , t. a. 1996 , , 457 , 834 van den heuvel , e. p. j. 1983 , in accretion - driven stellar x - ray sources , ed . w. h. g. lewin & e. p. j. van den heuvel ( cambridge and new york , cambridge university press ) , p. 303 vink , j. s. , de koter , a. , & lamers , h. j. g. l. m. 2001 , , 369 , 574 webbink , r. f. 1984 , , 277 , 355 wijnands , r. , int zand , j. j. m. , rupen , m. , et al . 2006 , , 449 , 1117 wiktorowicz , g. , belczynski , k. , & maccarone , t. j. 2013 , arxiv:1312.5924v2 wong , t .- w . , valsecchi , f. , ansari , a. , et al . 2014 . , 790 , 119 woosley , s. e. , & weaver , t. a. 1995 , , 101 , 181 xu , x .- j . , & li , x .- d . 2010a , , 722 , 1985 xu , x .- j . , & li , x .- d . 2010b , , 716 , 114 yungelson , l. r. , & lasota , j .- 2008 , , 488 , 257 zhang , s. n. , cui , w. , & chen , w. 1997 , , 482 , l155 zorotovic , m. , schreiber , m. r. , gsicke , b. t. et al . 2011 , , 536 , l3 with the stellar radius @xmath50 for a @xmath148 star with initial mass of @xmath149 . the the red , blue and green lines represent @xmath44 , @xmath43 , and @xmath42 , and the solid and dashed lines represent the results with the wind1 and wind2 prescriptions , respectively.[figure1],scaledwidth=100.0% ] values as a function of the stellar radius for stars with different mass . the solid and dashed lines represent the results with the wind1 and wind2 prescriptions , respectively . [ figure2],scaledwidth=100.0% ] star . here , the boundary of the core and the envelope is determined to be the maximal compression point in the hydrogen burning shell . the solid and dashed lines represent the results with the wind1 and wind2 prescriptions , respectively . [ figure3],scaledwidth=100.0% ] .the stellar parameters for a @xmath51 star at different evolutionary stages . columns 2 - 4 and 5 - 7 correspond to the wind1 ( left ) and wind2 ( right ) prescriptions , respectively [ table1 ] [ cols="^,^,^,^,^,^,^",options="header " , ]

one of the most popular methods for numerical simulation of quantum field theories is the monte - carlo integration over all possible field configurations , with the integration weight being proportional to @xmath0 , where @xmath1 is the action of the theory . however , such method might become not very efficient when the fields are strongly correlated at large distances , for example , in the vicinity of quantum phase transitions , or when the path - integral weight becomes non - positive . the latter situation is typical , e.g. , for field theories at finite chemical potential . recently it has been realized that these difficulties can be avoided or significantly reduced if one performs the perturbative expansion of the theory around some `` free '' action @xmath2 , and sums up the resulting series , instead of directly evaluating the path integral . such summation can be also performed using the monte - carlo procedure , and this alternative simulation method is usually called `` diagrammatic monte - carlo '' @xcite . diagrammatic monte - carlo turned out to be very efficient for numerous problems in statistical and condensed matter physics , and allowed to obtain many interesting results with precision which was unattainable for other methods . however , a systematic application of diagrammatic monte - carlo to field theories which are relevant for high - energy physics is so far hindered mainly because expansions in powers of coupling constant , which lead to the conventional diagrammatic technique due to feynman , typically yield only asymptotic series . such series can not be directly summed and therefore can not be sampled by a monte - carlo procedure . in practical simulations , this non - convergence problem is typically avoided by finding a suitable strong - coupling expansion . it is then argued that in a finite volume this expansion can be continued to the weak - coupling domain @xcite . examples of such expansions which were used for numerical simulations are the strong - coupling expansion in @xmath3 and @xmath4 lattice sigma - models and in abelian gauge theories , and the aizenman random current representation @xcite for the @xmath5 theory @xcite . unfortunately , the structure of strong - coupling expansions differs significantly for different theories . moreover , in many cases they are quite complicated and their structure is not sufficiently well understood . in particular , it has not been realized yet how to systematically sample strong - coupling expansion diagrams for non - abelian lattice gauge theories or for sigma - models with @xmath6 target space . some progress in this direction has been recently achieved for field theories in the large-@xmath7 limit @xcite . another way to treat the non - convergent weak - coupling expansions has been proposed recently in @xcite . the basic idea is to construct a sequence of better and better approximations to the original path integral , each of which has a convergent expansion . however , so far the utility of this construction was demonstrated only for the zero - dimensional theory . it is also not clear how to generalize the construction of @xcite to lattice field theories with compact variables , such as non - abelian lattice gauge theories . one can conclude that in the context of numerical simulations in high - energy physics diagrammatic monte carlo is so far a promising , but not a universal tool . in the present paper we propose a novel simulation method which , on the one hand , inherits the advantageous features of diagrammatic monte - carlo , and , on the other hand , can be used to investigate theories with asymptotic weak - coupling expansions and does not require the detailed knowledge of the structure of the perturbative expansion . the method is based on the stochastic interpretation of schwinger - dyson equations , which are typically much easier to derive than the general form of the coefficients of perturbative series . such stochastic interpretation has been considered recently in @xcite for large-@xmath7 quantum field theories , and a somewhat similar approach was discussed quite a long time ago in @xcite ( see @xcite for a more detailed discussion of the methods used in these papers ) . similarly to the `` worm algorithm '' @xcite , which is an essential ingredient in the diagrammatic monte - carlo , the method samples open diagrams which correspond to field correlators , rather than closed diagrams which correspond to the partition function of the theory . another distinct feature is that the diagrams are sampled directly in the momentum space . asymptotic series can be treated within this method using the standard resummation tools , such as the pade - borel resummation or expansion over the basis of some special functions @xcite . for example , in the case of factorially divergent series , our method incorporates the borel transform ( or , more generally , the borel - leroy transform ) of the field correlators in a natural way . in order to illustrate the applicability of our method , here we consider the theory in which schwinger - dyson equations take probably the simplest nontrivial form , namely , the theory of a one - component scalar field with quartic interaction . perturbative series in this theory are known to diverge factorially , and we use pade - borel resummation to recover physical results . as a test of the method , we study the scale dependence of the renormalized coupling constant of the theory in different space - time dimensions , and confirm numerically the triviality of the theory in five and four space - time dimensions and the instability of the trivial fixed point in three dimensions @xcite . for simplicity , we work only in the phase with unbroken @xmath8 symmetry . application of our algorithm to the phase with broken symmetry will be discussed briefly in the concluding section , and will be studied separately elsewhere . we hope that the simulation strategy , which we illustrate here on the simplest nontrivial example , can be easily generalized to other quantum field theories . the paper is organized as follows : in section [ sec : sds_phi4 ] we introduce the basic definitions and write down the schwinger - dyson equations for the @xmath5 theory in momentum space . in section [ sec : lineq_stochastic ] we provide a general description of the stochastic method which we use to solve these equations . then in section [ sec : stochastic_sd_solution ] we formulate the simulation algorithm which stochastically estimates the coefficients of the perturbative expansion of the field correlators in powers of the bare coupling constant . in section [ sec : series_resummation ] we consider a practical method to extract physical observables from the numerical data produced by this algorithm . while sections [ sec : sds_phi4 ] - [ sec : stochastic_sd_solution ] are essential to understand the presented method , section [ sec : series_resummation ] is more technical . in section [ sec : phys_res ] we present and briefly discuss some physical results which we obtained using our algorithm . finally , in section [ sec : conclusions ] we make some concluding remarks and discuss further extensions and generalizations of our approach . we consider the theory of a one - component scalar field @xmath9 in @xmath10-dimensional euclidean space with quartic interaction . the action of the theory is : @xmath11 the disconnected field correlators in momentum space are defined as follows : @xmath12 note that only even - order correlators are nonzero . by @xmath13 we denote the corresponding connected correlators ( that is , the correlators which contain only connected feynman diagrams ) . due to momentum conservation , all correlators also contain a factor @xmath14 . it is also convenient to define the correlators @xmath15 from which this factor is omitted . we define the renormalized mass @xmath16 and the wave function renormalization constant @xmath17 from the behavior of the two - point correlator at small momenta @xcite : @xmath18 the renormalized coupling constant @xmath19 is related to the one - particle irreducible four - point correlator @xmath20 at zero momenta . for the theory ( [ phi4_action ] ) it is proportional to the connected four - point correlator with truncated external legs : @xmath21 we note that our definition of the coupling constant differs from the one that is used in most textbooks by a factor of @xmath22 . this definition is more convenient for the analysis of schwinger - dyson equations . schwinger - dyson equations for the correlators ( [ phi4_correlators_def ] ) in momentum space read : @xmath23 @xmath24 where the arguments of the correlator in the first summand on the r.h.s . of ( [ phi4_sd_disconnected ] ) are all the momenta except @xmath25 and @xmath26 . equations ( [ phi4_sd_disconnected_n2 ] ) and ( [ phi4_sd_disconnected ] ) were obtained by variation of the field correlators and the action over @xmath27 . similar equations can be obtained for any argument of field correlators @xmath28 , but the resulting system of equations turns out to be redundant . to obtain a complete system of equations , it is sufficient to consider the variation over @xmath27 only . we thus arrive at a system of functional linear inhomogeneous equations for an infinite set of unknown functions @xmath29 . according to the general theorems of linear algebra , the solution of such equations is unique , if it exists . a straightforward way to solve equations ( [ phi4_sd_disconnected_n2 ] ) , ( [ phi4_sd_disconnected ] ) is to truncate an infinite set of equations at some correlator order and to discretize the continuum momenta . in this case , however , the required computational resources quickly grow with the number of correlators which are studied . such infinite systems of linear equations can be more efficiently solved using stochastic methods . in the next section we provide a general description of a specific stochastic method which , in our opinion , is most convenient for the solution of schwinger - dyson equations in quantum field theories . we consider a system of linear inhomogeneous equations of the following form : @xmath30 where @xmath31 is the element of some space @xmath32 and @xmath33 denotes summation or integration over all elements of this space . the coefficients @xmath34 are assumed to be positive , while @xmath35 can be of any sign . @xmath34 and @xmath35 are also assumed to satisfy the inequalities @xmath36 for any @xmath37 . the source function @xmath38 is also assumed to be positive . factorization of the coefficients of the equation ( [ lineq0 ] ) into @xmath34 and @xmath35 is to a large extent arbitrary , and can be chosen in some optimal way for any particular problem . if the space @xmath32 in ( [ lineq_ineq ] ) contains an infinite number of elements , a deterministic approximate solution of ( [ lineq0 ] ) would require the truncation of this space to some finite number of elements , and it is _ a priori _ not known which elements can be discarded . an alternative to the deterministic solution is the stochastic solution , for which @xmath39 is proportional to the probability of occurence of the element @xmath31 in some random process . this idea dates back to the works of von neumann and ulam ( described in @xcite , see also @xcite for a more recent review ) . in this case , the solution automatically incorporates the importance sampling . namely , those elements of space @xmath32 for which @xmath39 is numerically sufficiently small are automatically truncated , since the random process can not reach them in a finite number of iterations . such methods for the solution of large systems of linear equations are widely used , e.g. , in engineering and control design , for problems related to partitions of large graphs etc . since the basic idea behind these methods is very simple , and the number of works on the subject is vast , we have decided to present here a concise formulation of the method which we found most useful , without giving any specific reference to the literature . consider the markov process with states specified by one element of the space @xmath32 and one real number @xmath40 . we specify this process by the following [ alg : rp_abstract_desc ] at each iteration , do one of the following : evolve : : : with probability @xmath34 change from the current state @xmath41 to the new state @xmath31 and multiply @xmath40 by @xmath35 . restart : : : otherwise go to the state with @xmath42 and a random @xmath43 , which is distributed with probability @xmath44 . the `` restart '' action is also used to initiate the random process . the first inequality in ( [ lineq_ineq ] ) ensures that the probability of the `` evolve '' action does not exceed unity , and the second inequality ensures that @xmath40 remains bounded and thus have a well - defined average over the stationary state of the markov process . now let @xmath45 be the stationary probability distribution for this random process , that is , the probability to find the process in the state @xmath46 measured over a sufficiently large number of iterations of algorithm [ alg : rp_abstract_desc ] . a general form of the equation governing the stationary probability distributions of markov processes is @xmath47 , where @xmath48 is the probability of transition from the state @xmath49 to the state @xmath50 . for the random process specified above such an equation reads : @xmath51 we note that @xmath52 is the average probability of choosing the `` restart '' action . let us denote @xmath53 . by integrating ( [ lineq_markov ] ) over @xmath40 one can check that @xmath54 is the solution of the original equation ( [ lineq0 ] ) . in practice , in order to find @xmath39 one should simulate the random process specified by the algorithm [ alg : rp_abstract_desc ] , and sum up the factors @xmath40 over a sufficiently large number of iterations separately for each @xmath31 , dividing the results by a total number of iterations : @xmath55 where @xmath56 and @xmath57 are the values of @xmath31 and @xmath40 at @xmath58th iteration . here @xmath59 can be measured simultaneously with @xmath39 as @xmath60 , where @xmath61 is the number of `` restart '' actions during @xmath62 iterations . at a first sight , one can try to apply the method described in section [ sec : lineq_stochastic ] directly to equations ( [ phi4_sd_disconnected_n2 ] ) and ( [ phi4_sd_disconnected ] ) . the space @xmath32 should be then the space of sequences of momenta @xmath63 for any @xmath64 . however , a simple analysis shows that the inequalities ( [ lineq_ineq ] ) for equations ( [ phi4_sd_disconnected_n2 ] ) and ( [ phi4_sd_disconnected ] ) can not be satisfied , since the number of summands in the first term on the r.h.s . of ( [ phi4_sd_disconnected ] ) grows linearly with the number @xmath65 of field variables in the correlator . as will become clear from what follows , this growth is in fact the manifestation of the factorial divergence of the coefficients of the perturbative series . such structure of schwinger - dyson equations for one - component scalar field should be contrasted with scalar field theories in the large-@xmath7 limit , where the number of planar diagrams grows only in geometric progression , and where schwinger - dyson equations in factorized form at sufficiently small coupling constant can be interpreted as equations for the stationary probability distributions of the so - called nonlinear random processes @xcite . in order to overcome this problem , let us assume that the correlators ( [ phi4_correlators_def ] ) can be formally expanded in power series in the coupling constant @xmath66 : @xmath67where the coefficients @xmath68 will be specified later . now we insert the expansion ( [ correlators_prob_redef ] ) into the schwinger - dyson equations ( [ phi4_sd_disconnected_n2 ] ) and ( [ phi4_sd_disconnected ] ) and collect the terms with different powers of @xmath66 . this yields the following equations for the functions @xmath69 : @xmath70 @xmath71 equations ( [ stochastic_eq_n2 ] ) and ( [ stochastic_eq ] ) are also linear inhomogeneous equations , but on a larger functional space - the unknown functions @xmath69 now depend also on @xmath72 . we can now try to choose the coefficients @xmath68 in ( [ correlators_prob_redef ] ) so as to cast these equations in the stochastic form ( [ lineq0 ] ) with coefficients which satisfy the inequalities ( [ lineq_ineq ] ) . the space @xmath32 now should contain sequences @xmath63 of @xmath73 momenta in @xmath10-dimensional space and a nonnegative integer number @xmath72 . the sum @xmath33 should be understood as summation over @xmath65 and @xmath72 and integration over @xmath65 momenta : @xmath74 thus , altogether the space of states of the markov process that we would like to construct consists of an ordered sequence of momenta of arbitrary length @xmath63 , a positive integer number @xmath72 and a real number @xmath40 . comparing the form of equations ( [ stochastic_eq_n2 ] ) and ( [ stochastic_eq ] ) with the general form ( [ lineq0 ] ) , we can now construct markov process which will solve these equations , as discussed in section [ sec : lineq_stochastic ] . again , we specify it by the following [ alg : rp_phi4 ] at each iteration , do one of the following : add momenta : : : with probability @xmath75 , where @xmath76 , add a pair of momenta @xmath77 to the current sequence of momenta , inserting the first momenta at the beginning of the sequence and the second - between the @xmath50th and @xmath78th elements of the sequence , where @xmath50 is chosen at random between @xmath79 possibilities . the momentum @xmath80 is distributed within the @xmath10-dimensional sphere of radius @xmath81 with the probability distribution @xmath82 . do not change @xmath72 and @xmath40 . create vertex : : : with probability @xmath83 replace the three first momenta @xmath25 , @xmath84 and @xmath85 in the sequence by their sum @xmath86 . multiply @xmath40 by @xmath87 and increase @xmath72 by one . restart : : : otherwise restart with a sequence which contains a pair of random momenta @xmath77 with the probability distribution @xmath82 , and with @xmath88 and @xmath89 . since the factor @xmath87 in the `` create vertex '' action does not exceed unity , the second inequality in ( [ lineq_ineq ] ) is satisfied . let us check whether we can also satisfy the first inequality , that is , whether the total probability of `` add momenta '' and `` create vertex '' actions can be made less than one for any sequence of momenta and for any @xmath72 . this can only be achieved if both ratios @xmath90 and @xmath91 are finite for any @xmath65 and @xmath72 . for the first ratio it is only possible if @xmath68 grows not slower than @xmath92 at large @xmath65 . in this case the second ratio can only be bounded for all @xmath65 and @xmath72 if @xmath68 grows as @xmath93 at large @xmath65 , @xmath72 . we thus conclude that equations ( [ stochastic_eq_n2 ] ) and ( [ stochastic_eq ] ) indeed can be interpreted as the equations for the stationary probability distribution of some markov process , if the coefficients @xmath68 grow sufficiently fast . the factorial divergence of the perturbative series is then absorbed in the coefficients @xmath68 . it is interesting that such a simple analysis of schwinger - dyson equations reveals the divergence of the perturbative series in a straightforward way , without the need to explicitly calculate any diagrams ! in algorithm [ alg : rp_phi4 ] we have also introduced the ultraviolet cutoff @xmath81 , which is necessary in order to normalize the probability distribution of the momenta which are created when the `` add momenta '' action is chosen . in our simulations we set @xmath94 , so that all masses are measured in units of @xmath81 ( `` lattice units '' ) . a more self - consistent way to introduce such cutoff would be probably to start with lattice theory in the coordinate space and assume that all momenta belong to the first brillouin zone @xmath95 . according to the standard renormalization - group arguments , a particular choice of the cutoff prescription should result only in tiny corrections of order @xmath96 to the physical results . therefore we use the isotropic cutoff scheme , which leads to a much more efficient numerical algorithm . indeed , random momenta with isotropic probability distribution can be easily generated by the standard mapping of the one - dimensional probability distribution @xmath97 to the uniform probability distribution on the interval @xmath98 } $ ] . on the other hand , with lattice discretization in coordinate space the random momenta should be generated with the anisotropic probability distribution @xmath99 . this can be done either by using the metropolis algorithm or by discretizing the momenta , which require much more computational resources . an interpretation of our cutoff scheme in terms of feynman diagrams will be given below . algorithm [ alg : rp_phi4 ] thus stochastically samples the coefficients of the perturbative expansion of field correlators of the theory ( [ phi4_action ] ) , which are reweighted by the factors @xmath68 . each state of the markov process defined by this algorithm corresponds therefore to some ( in general , disconnected ) feynman diagram with @xmath65 external legs and @xmath72 vertices . combinatorial growth of the number of diagrams is compensated by the growth of the coefficients @xmath68 . the action `` add momenta '' corresponds to adding a bare propagator line to the diagram , without attaching its legs anywhere - thus the number of external legs is increased by two , and the diagram order is not changed . the action `` create vertex '' corresponds to the creation of one more vertex by joining three external legs of the diagram and attaching a new external leg to the joint . the number of external legs is thus decreased by two and the order of the diagram is increased by one . since we add new momenta in pairs @xmath77 , the sum of momenta on all the external legs of any generated diagram is always identically zero . the constraint @xmath100 is thus automatically satisfied . by a slight modification of algorithm [ alg : rp_phi4 ] , one can also easily trace whether the generated diagram is connected or disconnected . upon the `` add momenta '' action one should mark the two newly created legs by some unique label , and upon the `` create vertex '' action , the labels of the legs which are being joined should be replaced by a single new label . in this way , one can directly measure the four - point connected diagram which enters the definition of the renormalized coupling constant ( [ renorm_coupling_def ] ) . the `` restart '' action simply erases any diagram which was constructed before , and initiates the construction of a new diagram , which starts with just a single bare propagator . since the maximal achievable diagram order is equal to the maximal number of `` create vertex '' actions between the two consecutive `` restart '' actions , it is advantageous to minimize the probability of `` restarts '' . the rate of `` restart '' actions @xmath59 for our algorithm [ alg : rp_phi4 ] thus plays the role similar to the reject probability in the standard metropolis - based monte - carlo : for maximal numerical efficiency it is advantageous to maximally reduce it . on the other hand , in order to implement importance sampling with maximal efficiency , the factor @xmath40 in algorithms [ alg : rp_abstract_desc ] and [ alg : rp_phi4 ] should be as close to unity as possible . for each particular implementation of algorithm [ alg : rp_abstract_desc ] , one should find optimal balance between the rate of `` restart '' actions and the efficiency of importance sampling by adjusting the factors @xmath35 in ( [ lineq0 ] ) . in algorithm [ alg : rp_phi4 ] we have already chosen these factors to be @xmath87 for the `` create vertex '' action , and unity for the `` add momenta '' action . in the language of feynman diagrams , such a prescription can be interpreted as follows . the weight of each feynman diagram is proportional to the kinematical factor @xmath101 where @xmath102 , @xmath103 are independent momenta circulating in loops and @xmath104 , @xmath105 can be expressed as some linear combinations of @xmath102 and the momenta of the external legs . in algorithm [ alg : rp_phi4 ] we in fact perform monte - carlo integration over the independent momenta @xmath102 , generating them randomly and independently at the `` add momenta '' action with the probability distribution proportional to @xmath106 . our isotropic ultraviolet cutoff scheme thus consists in limiting the integrations over independent momenta @xmath102 to @xmath107 . the factor @xmath40 in algorithm [ alg : rp_phi4 ] is then proportional to the product of propagators involving the dependent momenta @xmath104 . the weight ( [ diag_weight ] ) is obtained by averaging @xmath40 over random @xmath102 . an alternative is to perform a metropolis - like integration , with the probability of the `` create vertex '' action being proportional to @xmath108 and with @xmath40 being identically equal to one . however , we have found that such an alternative prescription increases the rate of `` restart '' actions quite significantly , and results in a less optimal performance of the algorithm . therefore , we consider only the first choice , implemented in algorithm [ alg : rp_phi4 ] . let us now consider the coefficients @xmath68 more closely . as it was already shown , @xmath68 should grow as @xmath93 at large @xmath65 , @xmath72 . on the other hand , if @xmath68 grow too fast , higher - order diagrams will be strongly suppressed . let us assume that @xmath68 is proportional to @xmath109 times some functions which grow exponentially with @xmath72 and @xmath65 . by minimizing the rate of `` restart '' actions for all @xmath65 at a fixed @xmath72 , we find the optimal value @xmath110 . therefore , we choose the following form of the coefficients @xmath68 : @xmath111 with some @xmath31 and @xmath41 . the sum of the probabilities of the `` add momenta '' and `` create vertex '' actions for such a choice of @xmath112 is : @xmath113 the maximal value of the total probability is reached for @xmath89 . minimization with respect to @xmath31 then shows that the total probability ( [ coeff_ptot ] ) does not exceed one only if @xmath114 for @xmath115 , the total probability ( [ coeff_ptot ] ) is equal to one if @xmath116 . if we set @xmath115 and @xmath117 , as we actually did in our simulations , the probabilities of both `` add momenta '' and `` create vertex '' actions do not depend on the bare mass @xmath118 or on the space - time dimensionality @xmath10 . such a choice also minimizes the rate of `` restart '' actions @xmath59 . hence , the autocorrelation time of the markov process specified by algorithm [ alg : rp_phi4 ] also does not depend on the physical parameters of the theory ( [ phi4_action ] ) . in practice , it does not exceed several iterations . we see that our algorithm does not suffer from the critical slowing - down in the sense of the usual monte - carlo algorithms , where autocorrelation time typically strongly increases close to the continuum limit . however , this absence of critical slowing down is compensated by the increase of the computational time which is required to obtain sufficient statistics in the low - momentum region . indeed , the volume of this region decreases as @xmath119 as the infrared cutoff @xmath120 goes to zero , and hence the probability for the momenta @xmath121 to get within this region also quickly decreases . it is also interesting to note that the inclusive probabilities @xmath122 to obtain a diagram with @xmath65 legs , irrespectively of the order @xmath72 , are proportional to the borel - leroy transforms of the field correlators @xmath29 . this means that the factors @xmath123 in the perturbative expansion of the correlators are replaced by @xmath124 . thus one can say that formally algorithm [ alg : rp_phi4 ] stochastically estimates the borel - leroy transform of field correlators for @xmath41 in the range @xmath125 . however , in order to recover physical results , one should integrate the borel - leroy transform in the range @xmath126 , therefore it is difficult to use this statement for practical calculations . a much more convenient method is to analyze the dependence of the coefficients @xmath69 on @xmath72 , as described in section [ sec : series_resummation ] . for numerical experiments , we have implemented algorithm [ alg : rp_abstract_desc ] as a c program . source code of this program is available at @xcite . we have used the ` ranlux ` random number generator at luxury level 2 @xcite . all numerical data to which we refer below were obtained using this code . we have set @xmath115 , @xmath117 . numerically we found that with such choice of parameters the average rate of `` restart '' actions is @xmath127 . + in order to test the performance of algorithm [ alg : rp_phi4 ] , we first consider zero - dimensional scalar field theory with the action @xmath128 , for which the expansion coefficients in ( [ correlators_prob_redef ] ) can be easily obtained in an analytic form . on fig . [ fig : d0series_ratio ] we plot the ratios of the expansion coefficients @xmath129 of the connected @xmath65-point functions to their exact values . these results were calculated with @xmath130 iterations of our algorithm [ alg : rp_phi4 ] . one can see that up to approximately @xmath131 numerical results differ from exact values by less than @xmath132 , while at larger @xmath72 statistical errors become quite significant . + next , on fig . [ fig : order_dist ] we plot the probabilities of encountering the two- and four - point diagrams of order @xmath72 per one iteration as a function of @xmath72 for the four - dimensional theory . the label `` all momenta '' means that the diagrams were added to statistics independently of the momenta of their external legs . the label @xmath133 , where @xmath134 for the two - point diagrams and @xmath135 for the four - point diagrams , means that the diagrams were additionally weighted by the factor @xmath136 ( see also section [ sec : series_resummation ] ) . the data was obtained for @xmath137 iterations of algorithm [ alg : rp_phi4 ] for @xmath138 and @xmath139 . we see that when the diagrams are added to statistics irrespectively of the momenta of external legs , at large order @xmath72 the probabilities of encountering both the two- and four - point diagrams of order @xmath72 are almost equal and decay as @xmath140 , where @xmath141 . this value is very close to the probability @xmath142 of `` create vertex '' action in algorithm [ alg : rp_phi4 ] , since the probability of `` add momenta '' action is suppressed at large @xmath72 . imposing the infrared cutoff results in a strong kinematical suppression of the probabilities , but they still decay as @xmath140 at large @xmath72 . the fits of the form @xmath143 are plotted on fig . [ fig : order_dist ] with solid lines . algorithm [ alg : rp_phi4 ] stochastically estimates the coefficients of perturbative expansion of the correlators ( [ phi4_correlators_def ] ) , reweighted by the factors @xmath68 . in order to measure some physical observable one should be able to re - sum somehow the factorially divergent perturbative series ( [ correlators_prob_redef ] ) . in addition , the zero - momentum limit of the correlators ( [ phi4_correlators_def ] ) should be taken to measure the renormalized parameters of the theory from ( [ renorm_mass_def ] ) and ( [ renorm_coupling_def ] ) . + + let us first discuss the zero - momentum limit of the field correlators . we will first take the zero - momentum limit of each expansion coefficient @xmath69 in ( [ correlators_prob_redef ] ) , and then consider the resummation of the series ( [ correlators_prob_redef ] ) . as discussed above , algorithm [ alg : rp_phi4 ] produces sequences of momenta of the external legs of feynman diagrams , and @xmath144 is proportional to the probability of all the momenta in the sequence being zero . to calculate this probability numerically , one should measure the probability for all the momenta @xmath28 to belong to some small region near the point @xmath145 , and then extrapolate the ratio of this probability to the volume of the region to the zero size of this region . here we use soft infrared cutoff , and define for any function @xmath146 of @xmath65 momenta @xmath147 where @xmath148 and we also assume that @xmath146 contains the factor @xmath149 . this factor should be omitted when taking the zero - momentum limit , as in the definitions of the renormalized parameters ( [ renorm_mass_def ] ) and ( [ renorm_coupling_def ] ) . the zero - momentum limit of @xmath146 corresponds then to the limit @xmath150 . in order to measure the renormalized parameters @xmath16 , @xmath19 and @xmath17 from the behavior of correlators ( [ renorm_mass_def ] ) and ( [ renorm_coupling_def ] ) , we consider the regularized zero - momentum limit ( [ ir_cutoff_def ] ) @xmath151 of the quantity @xmath152 and tune the renormalized mass @xmath153 so that the deviation of @xmath151 from a constant value is minimized for @xmath154 . this constant is by definition the wave function renormalization constant @xmath17 . the functions @xmath151 are plotted for different bare masses and for different space - time dimensions on fig . [ fig : resummed ] on the left . horizontal solid lines are the values of @xmath17 , and vertical solid lines correspond to @xmath155 . one can see that the limit @xmath156 is indeed well - defined in this case , for all the values of the bare mass @xmath118 and for all space - time dimensions . in order to calculate @xmath151 , we have to calculate first its expansion coefficients @xmath157 , which are defined similarly to ( [ correlators_prob_redef ] ) . taking into account the definition ( [ ir_cutoff_def ] ) , we measure these coefficients by summing the quantities @xmath158 where @xmath159 , separately for different diagram orders @xmath72 over sufficiently large number of iterations of algorithm [ alg : rp_phi4 ] . then @xmath160 where the first factor arises due to normalization of the source term in ( [ phi4_sd_disconnected_n2 ] ) and @xmath161 denotes averaging over a sufficiently large number of iterations of the markov process specified by algorithm [ alg : rp_phi4 ] , as described in ( [ averaging_formula ] ) . similarly , in order to measure the renormalized coupling constant from ( [ renorm_coupling_def ] ) , we should measure the one - particle irreducible four - point correlator @xmath20 at zero external momenta . at small external momenta we use ( [ renorm_mass_def ] ) and ( [ renorm_coupling_def ] ) and write @xmath162 now one can take the regularized zero - momentum limit of the expansion coefficients of ( [ four_point_func_approx ] ) by summing the quantities @xmath163 separately for connected diagrams of different orders @xmath72 over sufficiently large number of iterations of algorithm [ alg : rp_phi4 ] , as in ( [ averaging_formula ] ) . here we have defined the kinematical factors @xmath164 , @xmath165 as follows : @xmath166 then the expansion coefficients of the regularized zero - momentum limit of the truncated connected four - point correlator ( [ four_point_func_approx ] ) are given by @xmath167 + the resummed function @xmath168 is plotted for different bare masses and for different space - time dimensions on fig . [ fig : resummed ] on the right ( to make comparison with ( [ renorm_coupling_def ] ) easier , we divide it by @xmath22 ) . since the low - momentum region for the four - point correlator is much stronger suppressed kinematically then for the two - point correlator , the limit @xmath156 is now not so well - defined , especially at small bare mass @xmath118 . in order to extrapolate @xmath168 to zero , we fit several ( from 3 to 5 ) data points with smallest @xmath120 by a linear function , and assume that the intercept of this linear function is the required limit @xmath169 . slanting solid lines on the plots on the right of fig . [ fig : resummed ] are these linear fits , and vertical solid lines again correspond to @xmath155 . clearly , such extrapolation procedure introduces quite large systematic errors into our measurements of the renormalized coupling constant ( [ renorm_coupling_def ] ) . however , as we shall see below , even with such a crude extrapolation we get reasonable results for the scale dependence of @xmath19 . in order to illustrate how the soft infrared cutoff ( [ ir_cutoff_def ] ) reproduces the zero - momentum limit , on fig . [ fig : low_order_contrib ] we also show the ratios of the expansion coefficients of the regularized zero - momentum limit of the connected @xmath65-point functions @xmath157 and @xmath170 to the results of analytical calculations up to @xmath171 . namely , we calculate analytically perturbative contributions to the two - point function @xmath172 and the connected four - point function @xmath173 up to @xmath171 and multiply the external legs by @xmath174 , where @xmath153 is calculated from our numerical data , as described below . then we either set all the external momenta to zero ( the corresponding points on the plot are shown with full symbols ) or integrate over them with the weight @xmath175 ( these points are shown with empty symbols ) . we set @xmath176 for the two - point function and @xmath134 for the four - point function . first we note that when both the analytical and the numerical results are weighted with the factor @xmath175 , the ratio is very close to one , which suggests that the loop integrals in ( [ diag_weight ] ) are indeed accurately reproduced by our algorithm [ alg : rp_phi4 ] for sufficiently large external momenta . on the other hand , the exact zero - momentum limit differs from the regularized result ( [ ir_cutoff_def ] ) . as discussed above , for the four - point function we have to take quite large ir cutoff in order to gain sufficient statistics , and the zero - momentum limit is then accessed using linear extrapolation . to get a deeper insight into possible problems with the regularized zero - momentum limit ( [ ir_cutoff_def ] ) of the two - point function , let us consider 1-particle reducible diagrams with two external legs and with @xmath72 insertions of 1-particle irreducible self - energy diagrams . the contribution of such diagrams is proportional to the kinematical factor @xmath177 where @xmath178 is the self - energy . if we neglect the dependence of @xmath178 on @xmath80 ( which is indeed weak for the lowest - order perturbative contributions ) and fix the infrared cutoff in ( [ ir_cutoff_def ] ) to some finite @xmath120 , at sufficiently large @xmath179 , a simple estimate shows that the regularized infrared limit ( [ ir_cutoff_def ] ) of ( [ nested_diagrams ] ) differs from the exact zero - momentum limit by a factor @xmath180 . in our simulations the minimal value of the infrared cutoff is @xmath176 . according to the above arguments , this allows to re - sum the diagrams with up to @xmath181 insertions of self - energy diagrams . however , even for smaller @xmath72 we expect that the contributions of the rapidly changing kinematical factors of the form similar to ( [ nested_diagrams ] ) are the main source of systematic errors in our simulations . according to fig . [ fig : low_order_contrib ] , they can be as large as @xmath182 for @xmath183 and @xmath171 . presumably , these errors can be significantly reduced by implementing some resummation procedure for one - particle reducible diagrams . after having discussed the zero - momentum limit , let us describe the integral borel - leroy transformation which we use for the resummation of the series ( [ correlators_prob_redef ] ) . for generality , in the following we will denote by @xmath184 any @xmath65-point correlator , probably multiplied by some function of momenta or integrated over all the momenta with some weight , as in ( [ g2_average_final ] ) and ( [ g4_average_final ] ) . correspondingly , by @xmath185 we denote the coefficients of expansion of @xmath184 in powers of the bare coupling constant @xmath66 , reweighted by the factors @xmath68 , as in ( [ correlators_prob_redef ] ) . inserting the explicit form of the factors @xmath68 from ( [ coeff_def ] ) into the series ( [ correlators_prob_redef ] ) and omitting the dependence on momenta , as discussed above , we obtain : @xmath186 using the integral representation of the gamma - function @xmath187 and changing the integration variable to @xmath188 , we get : @xmath189 in real simulations , the coefficients @xmath185 are known only up to some finite maximal order . a standard resummation strategy in this case is to construct the pade approximant of the function @xmath190 , that is , to approximate @xmath191 by some rational function of @xmath192 . in this case the expansion coefficients ( [ correlators_prob_redef ] ) are approximated by the sum of several exponents : @xmath193 where @xmath194 is the order of the tree - level diagrams which contribute to the connected @xmath65-point correlator : @xmath195 and @xmath196 . inserting the expression ( [ expansion_coeff_fits ] ) into the integral ( [ resummation ] ) , we obtain : @xmath197 thus each exponent in ( [ expansion_coeff_fits ] ) corresponds to a simple pole of the borel - leroy transform . integration over @xmath192 can be now performed analytically , and we obtain for the two - point correlator : @xmath198 where @xmath199 . for the four - point correlator , the integration yields : @xmath200 + the expansion coefficients @xmath201 and @xmath202 for the four - dimensional theory , defined by ( [ g2_average_final ] ) and ( [ g4_average_final ] ) , are plotted on fig . [ fig : order_contrib ] for different values of the bare mass @xmath118 . for this plot , we also omit the @xmath72-independent factor @xmath203 in ( [ coeff_def ] ) . the renormalized mass @xmath16 used to calculate ( [ g2_average_final ] ) and ( [ g4_average_final ] ) corresponds to @xmath204 . solid lines are the fits of the form ( [ expansion_coeff_fits ] ) , with three exponents for the two - point correlators and with one exponent - for the four - point correlators . + due to statistical errors in the coefficients @xmath185 , the standard pade approximant constructed on all the available data points turns out to be very unstable - small errors in the data lead to very large deviations of the approximant and to appearance of multiple spurious poles . in order to obtain more stable results , we find the optimal number of exponents and the values of @xmath205 and @xmath206 by fitting the expression ( [ expansion_coeff_fits ] ) to the numerical data . a technical difficulty is that such fits typically result in badly conditioned minimization problems , if the number of exponents exceeds two . fortunately , there are specific methods which work well for this particular problem . they are based on the singular value decomposition of the hankel matrix or the so - called page matrix @xcite . we found that for our data the fits based on the hankel matrices are optimal . let us briefly describe this fitting procedure . define the hankel matrix @xmath207 and the shifted hankel matrix @xmath208 as @xmath209 where @xmath210 is the maximal order to which the expansion coefficients are known and @xmath211 is the floor function . the hankel matrix can be decomposed as @xmath212 , where @xmath213 is the diagonal matrix with positive elements which are assumed to decrease from left to the right . we now form another matrix @xmath214 of rank @xmath7 , with @xmath215 , where @xmath7 is the required number of exponents in ( [ expansion_coeff_fits ] ) . the eigenvalues of @xmath216 are then the optimal values of the coefficients @xmath206 in ( [ expansion_coeff_fits ] ) @xcite . when @xmath206 are known , the optimal values of @xmath205 can be easily found by a simple linear regression . in our simulations we have used such maximal number of exponents @xmath7 , for which all @xmath206 in ( [ expansion_coeff_fits ] ) are still real and positive , so that all the poles of the borel - leroy transforms of field correlators ( [ resummation_afterfit ] ) lie on the real @xmath192 axis at @xmath217 . we have found that in this case the positions of all the poles are numerically stable . on the other hand , if we allow also for poles at @xmath218 or for poles off the real axis , their positions turn out to be numerically very unstable . we therefore disregard them as numerical artifacts . there are also theoretical arguments @xcite that the borel image of field correlators in bare perturbation theory should not have poles at positive real @xmath192 . a disadvantage of the fitting method described above is that it does not take into account the statistical errors in the data when finding the optimal values of @xmath206 - the weights of data points do not depend on their errors . therefore , after having found the optimal number of parameters and their values in ( [ expansion_coeff_fits ] ) from the singular value decomposition of the hankel matrix , we use this values as an initial guess for the standard minimization - based fitting algorithm . in most cases , the optimal values which take into account the errors in the data turn out to be very close to the ones that were found from the hankel matrices , and the minimization procedure quickly converges . the resulting @xmath219 is then typically of order of unity . we also note that we have used only the coefficients with relative error below @xmath132 for fitting . the positions of the poles of the pade approximants of the borel - leroy transforms of the functions @xmath220 and @xmath221 on the negative @xmath192 axis are shown on fig . [ fig : poles ] as a function of the bare mass @xmath118 . in order to illustrate the uncertainties in pole positions , in fig . [ fig : poles ] we show a scatter plot of poles for @xmath222 statistically independent data sets , each obtained with @xmath223 iterations of algorithm [ alg : rp_phi4 ] . for the two - point correlator , our fitting procedure reproduces three distinct poles , and for the four - point connected correlator , for which numerical errors are more significant , from one to two distinct poles can be seen . note that when the precision of the numerical data is not sufficient to find two distinct poles , our fitting procedure yields only one pole which is situated between the two poles which would be found if the precision of the data would be higher . this can be clearly seen for the smallest value of the bare mass @xmath224 . at small bare mass the positions of the poles are also less stable numerically , and hence larger statistics is required to find them with good precision . in this section we present the results of the measurements of the renormalized parameters of the theory - the renormalized mass @xmath16 , the renormalized coupling constant @xmath19 and the wave function renormalization constant @xmath17 as defined in ( [ renorm_mass_def ] ) and ( [ renorm_coupling_def ] ) . the resummation procedure and the regularized zero - momentum limit , which were used to obtain these results , are described in section [ sec : series_resummation ] . all the results presented on figs . [ fig : order_contrib ] - [ fig : rgflows ] were obtained by averaging over @xmath222 independent runs of algorithm [ alg : rp_phi4 ] , each consisting of @xmath223 iterations for @xmath225 and @xmath226 iterations for @xmath227 . for @xmath227 , such increase of the number of iterations was motivated by stronger kinematical suppression of low - momentum region in higher space - time dimension . averaging over independent runs was made in order to accurately estimate the statistical errors in the resummed correlators ( [ resummation_final_g2 ] ) and ( [ resummation_final_g4 ] ) . generation of each data set with @xmath223 iterations took several hours on a single 2 ghz cpu , which is comparable to the computer time which was required to produce similar results using the `` worm '' algorithm ( several core - months for @xmath228 data points in different dimensions @xcite ) . from fig . [ fig : order_dist ] one can see that with such statistics from @xmath222 to @xmath229 orders of perturbative expansion of field correlators can be analyzed in the small - momentum region ( we use only coefficient with relative errors below @xmath132 for resummation , as discussed above ) . + + on fig . [ fig : mr_vs_l ] we illustrate the dependence of the renormalized mass @xmath16 on the bare coupling constant @xmath66 for different space - time dimensions and for different bare masses @xmath118 . in order to demonstrate that our resummation procedure yields the results which agree with the lowest orders of perturbation theory at small @xmath66 , on fig . [ fig : mr_vs_l ] we also plot the one - loop result for the renormalized mass , which indeed fits the data for small @xmath66 . fig . [ fig : zr_vs_l ] shows the dependence of the wave function renormalization constant @xmath17 on the bare coupling constant . @xmath17 is close to unity in the whole range of coupling constants and for all dimensions @xmath10 , which agrees with the results of direct monte - carlo simulations @xcite . two - loop perturbative calculation of the self - energy shows that @xmath230 , with deviation from unity not exceeding @xmath231 for @xmath232 . on the other hand , our calculations show that @xmath233 . most likely this difference is due to systematic errors in our approximation of the zero - momentum limit of field correlators , which were discussed above in detail . + finally , on fig . [ fig : rgflows ] we illustrate the scale dependence of the renormalized coupling constant . namely , we fix the bare coupling constant @xmath66 ( we use @xmath234 for @xmath235 and @xmath236 for @xmath237 ) and change the bare mass @xmath118 . the renormalized coupling constant @xmath19 is then studied as a function of the renormalized mass @xmath16 in units of uv cutoff . this is , of course , equivalent to fixing the value of @xmath16 in physical units and changing the ultraviolet cutoff @xmath81 . according to the renormalization group arguments @xcite , when the continuum limit of the theory is approached and the renormalized mass @xmath16 in units of uv cutoff tends to zero , in space - time dimension @xmath138 or larger the renormalized coupling should tend to zero . hence the continuum limit of the @xmath5 theory in @xmath238 dimensions is a free theory of massless scalar fields . our data confirms this triviality conjecture : for both @xmath138 and @xmath227 the renormalized coupling clearly decreases with the renormalized mass . for @xmath138 the solid line on fig . [ fig : rgflows ] is the result of integration of the one - loop @xmath239-function , which implies that @xmath19 should approach zero very slowly , at a logarithmic rate . our results are consistent with such behavior within error range , although at small bare mass @xmath19 seems to decrease faster than logarithmically . this systematic deviation from logarithmic scaling is probably due to large systematic errors in the measurement of @xmath19 at small @xmath118 , as discussed in section [ sec : series_resummation ] . for @xmath227 , at small masses the renormalized coupling goes to zero almost linearly , in agreement with the dimensional analysis . linear fits of @xmath19 on several data points at small renormalized mass are also shown as solid lines on fig . [ fig : rgflows ] . on the other hand , for @xmath237 renormalization - group arguments predict that the trivial fixed point at @xmath240 , @xmath241 is unstable @xcite . our data agrees with this statement : the renormalized coupling in this case quickly grows as @xmath16 goes to zero . at large renormalized mass @xmath16 the renormalized coupling tends to its bare value @xmath66 for all bare masses and space - time dimensions . in this paper we have presented a novel simulation method which stochastically samples open feynman diagrams with probability which is proportional to their weight times some re - summing combinatorial factor . in this respect , our method is similar to the `` worm '' algorithm by prokofev and svistunov @xcite , which is often used in the framework of diagrammatic monte - carlo . the basic idea behind our approach is the stochastic perturbative solution of schwinger - dyson equations , which form an infinite system of linear inhomogeneous equations . thus it is not necessary to know explicitly the structure of each term in the perturbative expansion , and the transition probabilities are not subject to any detailed balance condition . in contrast to the `` worm '' algorithm , in our algorithm the number of external legs of diagrams is not fixed and also becomes a random variable . with only a minor modification of the algorithm , one can consider either disconnected or connected diagrams . in order to illustrate this general idea , we have applied it to study the running of the renormalized coupling constant in the scalar field theory with quartic interaction . with our numerical algorithm we were able to obtain the coefficients of perturbative expansions in powers of the bare coupling constant up to 15th order . resulting series were then resummed using a specially adapted pade - borel - leroy resummation procedure . we have confirmed that the coupling constant approaches zero in the continuum limit in four and five space - time dimensions , and grows in three space - time dimensions . performance of our algorithm in terms of computer time is comparable to the reported performance of the `` worm '' algorithm for the same theory @xcite . let us also note that the algorithm of @xcite significantly slows down in the weak - coupling limit @xmath242 , while the precision of our algorithm increases in this limit . disadvantages of our method are the need for an external resummation procedure and the small signal - to - noise ratio in the low - momentum region . the latter is analogous to slowing down of the conventional monte - carlo simulations with increasing lattice volume . the presented algorithm clearly allows for many improvements . for example , instead of expanding the correlators in powers of the bare coupling constant @xmath66 , as in ( [ correlators_prob_redef ] ) , one could try to expand them in the basis of some specially constructed functions of @xmath66 ( see e.g. chapter 16.5 of @xcite ) . as well , one can consider other choices of the re - summing coefficients @xmath243 in ( [ correlators_prob_redef ] ) than ( [ coeff_def ] ) . for example , @xmath243 can be proportional to the coefficients of the expansion of the averages @xmath244 in powers of the coupling constant @xmath66 in a zero - dimensional scalar field theory with the action @xmath245 . preliminary calculations show that such choice allows to explore even higher orders of perturbation theory , although the resummation procedure becomes somewhat more involved . it is also possible to speed up the simulations by analytically calculating the expansion coefficients @xmath69 for some small @xmath72 . in particular , if the expansion coefficient @xmath246 for the two - point function is known for some finite @xmath247 , one can modify algorithm [ alg : rp_phi4 ] by starting at @xmath247 and generating the first momenta in the sequence @xmath248 with the probability distribution proportional to @xmath246 . since @xmath72 always increases in algorithm [ alg : rp_phi4 ] , only diagrams of order @xmath249 will be sampled with such a modification . with small changes , the method can be also applied to quantum field theories in the large-@xmath7 limit @xcite . in this case the perturbative series are expected to diverge only exponentially , and by analyzing the behavior of large - order expansion coefficients one can locate phase transitions of gross - witten type . since the aim of the present paper is to illustrate the general method on the simplest possible nontrivial example , we do not consider here these potential improvements . let us also discuss briefly the application of algorithm [ alg : rp_phi4 ] to the @xmath5 theory with spontaneous symmetry breaking . a detailed study of this case will be presented elsewhere . since the possibility to interpret the bare propagator as the probability distribution of the momentum is absolutely crucial for the formulation of algorithm [ alg : rp_phi4 ] , it is not possible to set @xmath250 without modifying the algorithm . instead , one should redefine the field variable @xmath251 , where @xmath252 corresponds to one of the stable minima of the potential , and write schwinger - dyson equations in terms of the field @xmath253 , which now has the physical mass . note that as long as @xmath8 symmetry remains spontaneously broken , the system will not be able to jump to other nontrivial minimum of the potential ( as it would typically happen with the standard monte - carlo in a finite volume ) . the reason is that our method in fact operates in an infinite volume limit , and the infrared cutoff is imposed only when collecting the statistics ( see section [ sec : series_resummation ] ) . we hope that the presented approach can be easily generalized to other quantum field theories . it can be especially advantageous for those theories , for which the all - order perturbative expansion is difficult to construct in an explicit form or leads to asymptotic series , and hence the `` worm '' algorithm can not be applied in a straightforward way . of primary interest is the extension to non - abelian gauge theories with fermions , where diagrammatic monte - carlo can potentially help to tackle the notorious sign problem at finite chemical potential ( see , e.g. , @xcite ) . in non - abelian gauge theories schwinger - dyson equations can be written in terms of gauge - invariant quantities wilson loops . such formulation of schwinger - dyson equations is known as migdal - makeenko loop equations @xcite . the configuration space of the random process which solves these equations should contain loops ( that is , closed sequences of links ) on the lattice , and the basic transformations on this space should be the merging and the modification of loops ( for examples of such loop transformations see the algorithm for solving the weingarten model , described in @xcite ) . the probability that such random process produces some loop @xmath254 should be then proportional to the wilson loop @xmath255 . the main problem with such approach is that straightforward stochastic interpretation of loop equations on the lattice leads to the strong - coupling expansion @xcite , which is not analytically connected with the continuum limit of the theory at weak coupling . let us mention , however , that some attempts to explore the phase structure of qcd at finite chemical potential basing on the lowest - order strong - coupling expansion have been already reported in the literature , and quite interesting results were obtained @xcite . work on the application of the presented method to loop equations in non - abelian gauge theories is in progress . the author is grateful to m. i. polikarpov , yu . m. makeenko , a. s. gorsky and n. v. prokofev for interesting and stimulating discussions . this work was partly supported by grants rfbr 09 - 02 - 00338-a , rfbr 11 - 02 - 01227-a , a grant for the leading scientific schools no . nsh-6260.2010.2 , by the federal special - purpose programme personnel of the russian ministry of science and education and by a personal grant from the fair - russia research center ( frrc ) . , * * , ( ) , http://arxiv.org/abs/cond-mat/9804097[arxiv:cond-mat/9804097 ] ; , http://link.aps.org/doi/10.1103/physrevlett.87.160601 [ * * , ( ) ] ; , , , , * * , ( ) , http://arxiv.org/abs/cond-mat/0605350[arxiv:cond-mat/0605350 ] ; , , , , _ _ , , http://arxiv.org/abs/0802.2923[arxiv:0802.2923 ] . , * * , ( ) , http://arxiv.org/abs/0908.0284[arxiv:0908.0284 ] ; , * * , ( ) , http://arxiv.org/abs/0902.3100[arxiv:0902.3100 ] ; , * * , ( ) , http://arxiv.org/abs/1001.2231[arxiv:1001.2231 ] ; , * * , ( ) , http://arxiv.org/abs/1011.1359[arxiv:1011.1359 ] ; , , , _ _ ( ) , http://arxiv.org/abs/1101.3452[arxiv:1101.3452 ] .

with the commissioning of the large hadron collider ( lhc ) , it will soon be possible to test models of multiparticle production in hadron - hadron collisions up to an energy of 14 tev . charged particle densities , @xmath6 , especially the values at mid - rapidity and their dependence on center - of - mass energy @xmath7 ( c.m.s . ) are important for understanding the mechanism of hadron production and the interplay of soft and hard scattering contributions in the lhc energy range . the rate of parton - parton and multi parton - parton ( mpi ) scattering are strongly correlated to the observed particle multiplicity ( related also to _ initial entropy _ and _ initial energy density _ generated in the collision process ) . new data on inclusive charged particle distributions from the lhc in @xmath8 collisions have become available @xcite . these results complement previous data on @xmath8 and @xmath9 collisions taken at lower energies @xmath7 = 0.02 - 1.96 tev @xcite . many of these measurements have been used to constrain phenomenological models of soft - hadronic interactions and to predict properties at higher energies @xcite . the recent lhc data may lead to a better theoretical understanding based on a quantum chromodynamics ( qcd ) approach @xcite . the heavy ion jet interacting generator ( hijing ) @xcite and hijing / bb v1.10 models @xcite have been used extensively to study particle production in @xmath8 collisions and to determine the physical properties of the ultra - dense matter produced in relativistic heavy - ion collisions . in the lund @xcite and dual - parton ( dpm ) @xcite models multi qcd strings or flux tubes were proposed to describe soft multiparticle production in longitudinal color fields . color exchange between high @xmath10 partons in the projectile and target create confined color flux tubes of tension ( @xmath11 gev / fm ) , that must neutralize through pair production or color singlet hadronization approximately uniformly in rapidity . in nucleus - nucleus collisions , the @xmath12 enhancement of the local parton density of high @xmath10 partons allows for higher color casimir representations to be excited . those flux tubes with stronger longitudinal color fields than in average @xmath8 reactions have been called color ropes @xcite and naturally have higher string tensions @xcite . recently , an extension of color glass condensate ( cgc ) theory has proposed a more detailed dynamical `` glasma '' model @xcite of color ropes . in the hijing model @xcite the soft beam jet fragmentation is modeled by simple diquark - quark strings as in the lund model with multiple gluon kinks induced by soft gluon radiations . hard collisions are included with standard perturbative qcd ( pqcd ) as programmed in the pythia generator @xcite . however , hijing differs from pythia by the incusion of a geometric scaling multiple jet production model . thus this model contains both longitudinal field induced _ soft _ beam jet multiparticle production and collinear factorized pqcd based _ hard _ multiple jet production for @xmath13 gev / c . a systematic comparison with data on @xmath8 and @xmath9 collisions in a wide energy range @xcite revealed that minijet production and fragmentation as implemented in the hijing model provide a simultaneous and consistent explanation of several effects : the inclusive spectra at moderate transverse momentum ( @xmath14 ) ; the energy dependence of the central rapidity density ; the two particle correlation function ; and the degree of violation of koba - nielsen - olesen ( kno ) scaling @xcite , @xcite up to tevatron energy ( @xmath15 tev ) . however , the model failed to describe the dependence of the mean value of transverse momentum ( @xmath16 ) on charged particle multiplicity ( @xmath17 ) . wang argued @xcite that by requiring high @xmath17 within a limited pseudorapidity ( @xmath18 ) range one necessarily biases the data towards higher @xmath14 minijets , hence the observed increase of @xmath16 versus ( vs. ) @xmath6 @xcite . this effect has also been associated with the presence of transverse flow of the hadronic matter @xcite and was proposed as possibly due to _ quark - gluon plasma _ ( qgp ) formation already in @xmath8 collisions . initial states of color gauge fields produced in high - energy heavy - ion collisions have also recently been discussed in ref . decay of a strong color electric field ( scf ) ( @xmath19=@xmath20 v / m ) due to the schwinger mechanisms @xcite plays an important role at the initial stage of heavy - ion collisions at ultra - relativistic energies . a thermalization scenario based on the analogy between schwinger mechanisms and the hawking - unruh effect has been proposed @xcite . it was also suggested that the back - reaction and screening effects of quark and anti - quark pairs on external electric field could even lead to the phenomenon of plasma oscillations @xcite . recently , the schwinger mechanism has been revisited @xcite and pair production in time - dependent electric fields has been studied @xcite . it was concluded that particles with large momentum were likely to have been created earlier , and for very short temporal widths ( @xmath21 , where the compton time @xmath22 ) and as a consequence the schwinger formula could underestimate the reachable particle number density . in previous papers , we have shown that the dynamics of strangeness production in @xmath8 and au + au collisions at relativistic heavy ion collider ( rhic ) energies deviates considerably from calculations based on schwinger - like estimates for homogeneous and constant color fields and point to the contribution of fluctuations of transient strong color fields ( scf ) @xcite . in this paper we extend our study of the dynamic consequences of scf in the framework of the hijing / bb v2.0 model @xcite to particle production in hadron - hadron collisions at lhc energies . we explore dynamical effects associated with long range coherent fields ( _ i.e _ , strong longitudinal color fields , scf ) , including baryon junctions and loops @xcite , with emphasis on the new observables measured in @xmath8 collisions at the lhc by the alice , atlas and cms collaborations . our study aims to investigate a broad set of observables sensitive to the dynamics of the collisions , covering both longitudinal and transverse degree of freedom . in addition , this study is intended to provide the @xmath8 baseline to future extrapolations to lhc studies of proton - nucleus ( @xmath23 ) and nucleus - nucleus ( @xmath24 ) collisions . in hijing / bb v2.0 , in addition to conventional quark - diquark longitudinal electric fields , novel color flux topology junction anti - junction ( jj ) loops are also implemented . in a dual superconductor model of color confinement for the three - quark positioning in a y - geometry , the flux tubes converge first toward the center of the triangle and there is also another component that runs in the opposite direction ( like a _ mercedes star _ ) @xcite . unlike the conventional diquark - quark implemented in lund and the hijing model @xcite , the hijing / bb v1.10 @xcite model allows the diquark - quark to split with the three independent flux lines tied together with an @xmath25 a junction and terminating on three delocalized fundamental casimir quarks . we introduced @xcite a new version ( v2.0 ) of hijing / bb that differs from hijing / bb v1.10 @xcite in its implementation of additional more complex flux topologies via junction anti - junction ( jj ) loops . we parametrize the probability that a junction loop occurs in the string . moreover , we enhance the intrinsic ( anti)diquark - quark @xmath14 kick ( by a factor @xmath26 = 3 ) of all ( @xmath27-qq ) strings that contain one or multiple jj loops . the reason for this is the mechanism behind the dynamic of diquark - quark breaking ( see ref . @xcite for details ) . in string fragmentation phenomenology , it has been proposed that the observed strong enhancement of strange particle production in nuclear collisions could be naturally explained via strong color field ( scf ) effects @xcite . for a uniform chromoelectric flux tube with field ( _ e _ ) , the pair production rate @xcite per unit volume for a ( light)heavy quark ( @xmath28 ) is given by @xmath29 where @xmath30 ( diquark ) , @xmath31 ( strange ) , @xmath32 ( charm ) or @xmath33 ( bottom ) . the _ current quark masses _ are @xmath34 = 0.45 gev @xcite , @xmath35 gev , @xmath36 gev , and @xmath37 gev @xcite . the _ constituent quark masses _ of light non - strange quarks are @xmath38 = 0.23 gev , of the strange quark is @xmath39=0.35 gev @xcite , and of the diquark is @xmath40 gev @xcite . note that @xmath41 is the effective string tension in low energy @xmath8 reactions with @xmath42 gev / fm and @xmath43 , @xmath44 are the second order casimir operators ( see ref . an enhanced rate for spontaneous pair production is naturally associated with `` _ strong chromo - electric fields _ '' , such that @xmath45 1 _ at least some of the time_. in a strong longitudinal color electric field , the heavier flavor suppression factor @xmath46 varies with string tension via the well known schwinger formula @xcite , since @xmath47 for @xmath48 , @xmath31 , @xmath32 or @xmath33 and @xmath49 , @xmath50 . in our calculations , we assume that @xmath51 = 0.5 gev , @xmath52 = 0.28 gev , @xmath53 = 1.30 gev . therefore , the above formula implies a suppression of heavier quark production according to @xmath54 : @xmath50 : @xmath55 : @xmath31 : @xmath32 @xmath56 1 : 1 : 0.02 : 0.3 : 10@xmath57 for the vacuum string tension @xmath1 = 1 gev / fm . for a color rope , on the other hand , if the _ average string tension _ value @xmath58 increases , the suppression factors @xmath46 increase . we show below that this dynamical mechanism improves considerably the description of the strange meson / hyperon data at the tevatron and at lhc energies . saturation physics is based on the observation that small - x hadronic and nuclear wave functions , and , thus the scattering cross sections as well , are described by the same internal momentum scale known as the _ saturation scale _ ( @xmath59 ) . a recent analysis of @xmath8 data up to lhc 7 tev has shown that , with the @xmath60 factorized ( glr ) gluon fusion approximation @xcite , the growth of the @xmath5 can be accounted for if the saturation scale grows with c.m.s . energy as @xmath61 with @xmath62 . the saturation scale is also increasing with atomic number as @xmath63 @xcite . it was argued that the effective string tension ( @xmath58 ) of color ropes should scale with @xmath64 @xcite . however , in hijing the string / rope fragmentation is the only soft source of multiparticle production and multiple minijets provide a semi - hard additional source that is computable within collinear factorized standard pqcd with initial and final radiation ( dglap evolution @xcite ) . in order to achieve a quantitative description , within our hijing / bb framework we will show that combined effects of hard and soft sources of multiparticle production can reproduce the available data in the range @xmath65 tev only with a reduced dependence of the effective string tension on @xmath7 . we find that the data can be well reproduced taking @xmath66 where @xmath67 = 1 gev / fm is the vacuum string tension value , @xmath68 = 1 gev@xmath69 is a scale factor , and @xmath70 is adjusted to give @xmath71 gev / fm at the rhic energy @xmath72 tev . our phenomenological @xmath73 is compared to @xmath74 in fig . [ fig : kappa_vs_cgc ] , where @xmath58 = 1.40 gev / fm at @xmath7 = 0.017 tev increases to @xmath58 = 3.14 gev / fm at @xmath7 = 14 tev . [ h ! ] in @xmath8 collisions ( eq . [ eq : kappa_sup ] ) . the values @xmath75 ( eq . [ eq : larry10 ] ) are from cgc model fits to lhc data from ref . [ fig : kappa_vs_cgc ] , title="fig:",height=226 ] 0.5 cm the energy dependence of the string tension leads to a variation of the diquark / quark suppression factors , as well as the enhanced intrinsic transverse momentum @xmath60 . these include i ) the ratio of production rates of diquark - quark to quark pairs ( diquark - quark suppression factor ) , @xmath76 ; ii ) the ratio of production rates of strange to non - strange quark pairs ( strangeness suppression factor ) , @xmath77 ; iii ) the extra suppression associated with a diquark containing a strange quark compared to the normal suppression of strange quark ( @xmath78 ) , @xmath79 ; iv ) the suppression of spin 1 diquarks relative to spin 0 ones ( apart from the factor of 3 enhancement of the former based on counting the number of spin states ) , @xmath80 ; and v ) the ( anti)quark ( @xmath81 ) and ( anti)diquark ( @xmath82 ) gaussian width . as an example we plot in fig . [ fig : mg - fig1s_u ] the energy dependence of the suppression factor @xmath83 , when the string tension values @xmath73 are taken from eq . [ eq : kappa_sup ] in comparison with the values predicted using @xmath75 ( eq . [ eq : larry10 ] from the cgc model fit @xcite ) . [ h ! ] , using @xmath73 from eq . [ eq : kappa_sup ] and using @xmath74 from ref . @xcite are compared . ( note that this schwinger suppression factor is not used in the cgc model ) . [ fig : mg - fig1s_u ] , title="fig:",height=226 ] 0.5 cm the contributions of multiple jets to the multiplicity distributions in @xmath8 and @xmath9 collisions have been studied in detail in ref . @xcite . within the hijing model , one assumes that nucleon - nucleon collisions at high energy can be divided into soft and hard processes with at least one pair of jet production with @xmath84 . a cut - off scale @xmath85 in the transverse momentum of the final jet production has to be introduced below which the interaction is considered non - perturbative and can be characterized by a finite soft parton cross section @xmath86 . the inclusive jet cross section @xmath87 at leading order ( lo ) @xcite is @xmath88 where , @xmath89 depends on the parton - parton cross section @xmath90 and parton distribution functions ( pdf ) @xmath91 . the summation runs over all parton species ; @xmath92 and @xmath93 are the rapidities of the scattered partons ; @xmath94 and @xmath95 are the light - cone momentum fractions carried by the initial partons . the factor @xmath96 accounts for the next - to - leading order ( nlo ) corrections to the leading order jet cross section . in the default hijing model ( v.1.383 ) , the duke - owens parameterization @xcite of pdfs in nucleons is used . with the duke - owens parameterization of pdfs , an energy independent cut - off scale @xmath972 gev/@xmath32 and a constant soft parton cross section @xmath98 mb are sufficient to reproduce the experimental data on total and inelastic cross sections and the hadron central rapidity density in @xmath99 collisions @xcite . our results have been obtained using the same set of parameters for hard scatterings as in the latest version of hijing ( v1.383 ) . charged hadron multiplicity measurements are the first results of the lhc physics program . data for @xmath8 collisions were reported by the alice , atlas , and cms collaborations @xcite . the new data at mid - rapidity for non single diffractive interactions ( nsd ) and inelastic scattering ( inel ) are shown in fig . [ fig : pp_dn_ch_0 ] , which includes also similar results at lower energies . the main result is an observed sizeable increase of the central pseudorapidity density with c.m.s . energy . the main contribution to the multiplicity comes from soft interactions with only a small component originating from hard scattering of the partonic constituents of the proton . in contrast to the higher @xmath14 regime , well described by pqcd , particle production in soft collisions is generally modeled phenomenologically to describe the different @xmath8 scattering processes : elastic scattering ( el ) , single diffractive ( sd ) , double diffractive dissociation ( dd ) , and inelastic non - diffractive scattering ( nd ) . experimentally , minimum bias events are a close approximation of nsd interactions , _ i.e. _ , @xmath100 , where @xmath101 is the total cross section . the selection of nsd events is energy dependent and differs somewhat for different experimental triggers . the event selection of inelastic processes ( inel ) includes sd interactions : @xmath102 . the data therefore must be corrected for the sd component , involving model dependent calculations . the results of the model calculations for both nsd ( left panel ) and inel ( right panel ) are also shown in fig . [ fig : pp_dn_ch_0 ] . solid lines depict results including the strong color field ( scf ) effects whereas the results without scf effects are shown as dashed lines . without scf effects the model strongly overestimates the central charged particle density . the absolute value and energy increase of the central rapidity density are well reproduced assuming the energy dependent string tension given in eq . [ eq : kappa_sup ] . at higher lhc energies ( 2.36 and 7 tev ) , a discrepancy of 10 - 15 % is observed . [ h ! ] and @xmath9 interactions for non - single - diffractive ( nsd ) ( left panel ) and inelastic ( inel ) ( right panel ) interactions as a function of c.m.s . energy . the solid and dashed lines are the results with and without scf , respectively . the data are from refs . @xcite ( left panel ) and from refs . @xcite ( right panel ) . only statistical error bars are shown . the open stars at 7 and 14 tev ( alice fit ) , are obtained by a power law fit to lower energy data from refs . @xcite , @xcite . [ fig : pp_dn_ch_0 ] , title="fig:",height=151 ] 0.5 cm as the colliding energy increases , the rate of multiple parton interactions ( mpi ) also increases , producing a rise in the central multiplicity . the increase with energy in our phenomenology is due to the interplay of the increased mini - jet production in high colliding energy with scf effects . for an increase of strangeness suppression factors due to an increase of string tension with energy ( @xmath103 ) , the model predicts a decrease of produced pions due to energy conservation . lower values of @xmath73 imply smaller values for strangeness suppression factors , therefore a higher multiplicity of mesons ( mostly pions ) . [ t ! ] + @xmath104 ( @xmath105 ) collisions at various c.m.s . the solid histograms are the results with scf and jj loops . the data are from refs . @xcite ( ua5 ) , @xcite ( star ) , @xcite ( cdf ) , @xcite ( alice ) , @xcite ( cms ) . only statistical error bars are shown . [ fig : pp_dn_ch_eta ] , title="fig:",width=325,height=325 ] 0.5 cm changing the effective value @xmath73 = 2.89 gev / fm to @xmath73 = 2.0 gev / fm results in an increased multiplicity of 11% at 7 tev where the effect is greatest . in addition , the multiplicity depends also on the value of the cut - off parameter @xmath85 . low values of @xmath85 imply high rates of parton - parton scattering and hence high levels of particle multiplicity . evidently for increasing values of @xmath85 the opposite is expected . at 7 tev , where the effect is also the largest , changing @xmath85 by @xmath106 gev results in a change of only @xmath107 for the central pseudorapidity density . data on the charged particle pseudorapidity distribution are also available over a limited @xmath18 range @xcite @xcite . these are presented in fig . [ fig : pp_dn_ch_eta ] where we also include results obtained at 0.2 tev by the ua5 @xcite and star @xcite collaborations , and with cdf results @xcite obtained at the tevatron for @xmath9 collisions at 1.8 tev . consistent with the discussion above , a scenario with scf effects ( solid histograms ) reproduces the measured multiplicity distributions well . at all energies considered , theoretical calculations predict a central dip at mid - rapidity that is consistent with the observations . at 0.9 and 2.36 tev , the shape of the distribution measured by the alice collaboration is very well reproduced , while the cms results show a much flatter distributions than calculated ( also the case at 7 tev ) . data over a larger rapidity range are needed to determine the shape of the falling density in the fragmentation region . for completeness , predictions at 10 and 14 tev , the higher lhc energies , are also shown . the measured transverse momentum distributions for nsd events over an energy range @xmath108 tev are shown in fig . [ fig : pp_ch_pt_all ] . these recent measurements are performed in the central rapidity region and cover a wide @xmath14 range ( @xmath109 gev / c ) , where both hard and soft processes are expected to contribute . the data of atlas and cms are measured in larger pseudorapidity intervals ( @xmath110 ) . in contrast , alice and cdf measurements are in a very central region ( @xmath111 and @xmath112 , respectively ) . the calculation takes into account the difference in acceptance , but as can be concluded from fig . [ fig : pp_dn_ch_eta ] , this difference in pseudorapidity range has a negligible effect on the measured cross section . [ h ! ] , title="fig:",width=325,height=325 ] 0.5 cm the model calculations including scf effects describe the data well at @xmath113 tev , but lead to a harder spectrum that observed at higher energy . the model gives a fair description of the spectral shape at low @xmath14 but overestimate the data at high @xmath14 . the discrepancy is highest ( up to a factor of three ) at @xmath114 tev . we do not understand the source of this discrepancy and await higher @xmath14 data to draw a firm conclusion . however , in our phenomenology this could indicate that jet quenching , _ i.e. _ suppression of high @xmath14 particles like that observed at rhic energies in nucleus - nucleus collisions , could also appear in @xmath8 collisions in events with large multiplicity . this overestimation of high @xmath14 yield leads also to a similar overestimation of the mean transverse momentum ( @xmath16 ) and of the correlation of mean @xmath16 as a function of n@xmath115 , which we do not discuss here . within our model we generate events with different numbers of mini - jets up to some maximum values . high numbers of mini - jets lead to higher multiplicity events . for events with ten mini - jets ( the maximum assumed in our calculation ) , the central charged particle pseudorapidity density could increase up to @xmath56 20 and the total multiplicity could be greater than 150 at tevatron energy ( 1.8 tev ) . the measurements of two particle correlations over the entire azimuth could reveal the jet structure related to high @xmath14 particles @xcite . the study of these correlations in events with high multiplicity could help us to draw a firm conclusion with regard to a possible jet quenching phenomenon in @xmath8 collisions . the @xmath8 single particle inclusive @xmath14 spectra measurements are important for understanding collision dynamics , since the various particles show different systematic behavior , as observed at rhic energy @xcite . detailed theoretical predictions for single inclusive hadron production ( including strange hyperons ) are discussed in this section . baryon - to - meson ratios ( b / m ) are experimental observables that can be used at the lhc for investigating multi - parton interactions and helping to understand the underlying physics @xcite . [ h ! ] = 1.8 tev of baryon and mesons transverse momentum at mid - rapidity ( @xmath116 ) ( a ) and their ratios for single strange particles ( @xmath117)/@xmath118 ( b ) in minimum bias @xmath9 collisions . the solid and dashed lines have the same meaning as in fig . [ fig : pp_dn_ch_0 ] . experimental results at @xmath7 = 1.8 tev are from ref . @xcite ( cdf collaboration ) . error bars include only the statistical uncertainties . the ratios ( b ) have been calculated by us , dividing the spectra in the panel ( a ) . [ fig : pp_ks0_l0_18 ] , title="fig:",height=151 ] 0.5 cm unexpectedly high b / m ratios observed in @xmath119 collisions have been discussed in terms of recombination and coalescence mechanisms @xcite . such high ratios at intermediate @xmath14 were also reported in @xmath8 collisions at rhic @xcite and at the tevatron @xcite . in @xmath8 collisions , however , a coalescence / hadronization scenario is not favored due to low phase space density in the final state . our hijing / bb model , with scf effects included , provides an alternative dynamical explanation of the heavy - ion data at rhic energies . we have shown that the model also predicts an increasing yield of ( multi)strange particles , thereby better describing the experimental data @xcite . [ h ! ] ( a ) and their ratio ( b ) for minimum bias @xmath9 collisions at @xmath7 = 7.0 tev . the solid and dashed lines have the same meaning as in fig . [ fig : pp_dn_ch_0 ] . the experimental results ( a ) are from ref . @xcite(cms collaboration ) . error bars include only the statistical uncertainties . the ratios ( b ) have been calculated by us , dividing the spectra in the panel ( a ) . [ fig : pp_ks0_l0_7 ] , title="fig:",height=151 ] 0.5 cm figure [ fig : pp_ks0_l0_18 ] ( b ) shows a comparison of model predictions with cdf experimental data @xcite of the strange baryon - to - meson ratio @xmath120 at 1.8 tev . the particle @xmath14 spectra are shown in the panel ( a ) . the measured ratio is fairly well described within our phenomenology . the larger string tension parameterization results in a predicted increase of the ratio @xmath120 by a factor of @xmath56 10 at the tevatron energy . the @xmath14 spectra show that the increased ratio is due almost entirely to an increase of the @xmath121 cross section that is well described with @xmath122 . the model underestimates the kaon production by 15 - 25 % . the model predictions at 7 tev , currently the maximum energy where there are data @xcite , are shown in fig . [ fig : pp_ks0_l0_7 ] ( b ) . the model gives a good description of hyperon production ( @xmath123 ) , for which an increase by a factor of ten is still predicted if scf effects are considered . however , our calculations underestimate by approximately a factor of two the yield of @xmath118 and , as a consequence , result in a higher ratio than that observed . this result needs further investigation ( theoretical and experimental ) on heavy flavour production at this energy . we note that the models pythia @xcite and energy - conserving partons off - shell remnants and spliting of partons ladders ( epos ) @xcite can not reproduce the observed high b / m ratio ( see fig . 6 and fig . 7 from ref . the preliminary data reported by the cms collaboration indicate high hyperon yield . comparisons with pythia results show that this model significantly underestimates the hyperon yields in @xmath8 collisions at 0.9 and 7 tev @xcite . the strange particle ratios could also be the manifestation of new collective phenomena . in the epos model such an increase is obtained if the production of a _ mini - plasma _ is considered in @xmath8 collisions @xcite , @xcite . if confirmed by future measurements , the study of these observables could open a perspective on new physics in @xmath8 interactions . similar conclusions are obtained from the study of the proton / pion ( @xmath124 , @xmath125 ) ratios where data exist at lower energies @xcite . these data , shown in figure [ fig : pp_pi_pbar_18 ] , are limited to low @xmath126 gev / c . adding scf effects results in a very sizable increase of the ratio and our calculations provide a good description of the data in the measured range . however , as the calculations indicate , to draw a final conclusion measurements at intermediate and high @xmath14 are needed . the solid and dashed lines have the same meaning as in fig . [ fig : pp_dn_ch_0 ] . experimental results for @xmath127 at @xmath7 = 0.54 tev ( left panel ) and at @xmath7 = 1.8 tev ( right panel ) are from ref . @xcite ( e735 collaboration ) . the results at mid - rapidity at @xmath7 = 0.54 tev ( left panel ) are from ref . @xcite ( ua2 collaboration ) . error bars include only the statistical uncertainties . [ fig : pp_pi_pbar_18 ] , title="fig:",height=151 ] 0.5 cm to the extent that the lhc experiments are able to identify hadron species , such data will provide vital input to validate this interpretation . the model predictions at lhc energies for the @xmath14 dependence of the @xmath128 ratio are shown in fig . [ fig : pp_pi_pbar_14 ] . an enhancement up to the highest lhc energy and a weak energy dependence , with a saturation that sets in for a c.m.s . energy @xmath129 tev , is predicted . note that preliminary data at 0.9 tev reported by the alice collaboration for @xmath14 spectra of pions ( @xmath130 ) and protons ( @xmath104 ) @xcite cover to @xmath131 gev / c . the model results , with scf effects included ( dot - dashed histogram in fig . [ fig : pp_pi_pbar_14 ] ) are consistent with a @xmath132 ratio derived from the spectra reported by alice at 0.9 tev in ref . @xcite . [ h ! ] ) for minimum bias events at mid - rapidity at lhc energies . the upper curves correspond to calculations which include the effects of scf and jj loops . the lower curves corresponds to calculations without scf effects . the results are at 0.9 tev ( dot dashed - histograms ) , at 7 tev ( dashed histograms ) , at 10 tev ( dotted histograms ) and at 14 tev ( solid histograms ) . [ fig : pp_pi_pbar_14 ] , title="fig:",height=226 ] 0.5 cm in our approach , the dynamical mechanism that leads to such high values of b / m ratios is scf appearing at the initial stage of the interaction . the scf mechanism strongly modifies the fragmentation processes ( strangeness suppression factors ) and thus results in a huge increase of ( strange)baryons . this interpretation is also supported by more sophisticated theoretical calculations , in a scenario in which a time - dependent pulse for the initial strength of the color field is considered @xcite . the large enhancement of the baryon - to - meson ratios demonstrates that scf could play an important role in multiparticle production in @xmath8 collisions at lhc energies and that high energy density fluctuations can reach very high densities , potentially comparable to those reached in central au + au collisions at rhic energies @xcite . from the study of the baryon - anti - baryon asymmetry one can learn about the mechanism of baryon number transport . baryon production via the conventional default quark - diquark mechanisms in the lund string model are known to be inadequate even in @xmath133 phenomenology . this is one of the main reasons for our continued exploration of alternative baryon junction mechanisms . the details of the new implementation of jj loops in hijing / bb v2.0 are described in ref . @xcite sec ii a. in hijing / bb v2.0 the main two mechanisms for baryon production are quark - diquark ( @xmath134 ) string fragmentation and junction anti - junction ( jj ) loops @xcite in which baryons are produced approximately in pairs . in a junction loop a color flux line splits at some intermediate point into two flux lines at one junction and then the flux lines fuse back at an anti - junction somewhere further along the original flux line . the distance in rapidity between these points is chosen via a regge distribution @xcite . we assume that , out of the non single diffractive nucleon - nucleon ( nn ) interaction cross section ( @xmath135 ) , a fraction @xmath136 of the events excite a junction loop . the probability that the incident baryon has a jj loop in p(a)+a collisions after @xmath137 simulated binary collisions is given by @xmath138 where @xmath139 = 17 mb , @xmath140 is parametrized in the model , and the total inelastic nucleon - nucleon cross sections are calculated @xcite . these cross sections imply that a junction loop occurs in @xmath8 collisions with a rather high probability ( at rhic energy @xmath141 for @xmath142 = 42 mb ) . taking a constant value for @xmath139 results in a decrease with energy of the probability @xmath143 , due to a faster increase with energy of @xmath142 relative to @xmath140 . the actual probability is modified also by string fragmentation processes for which we consider a threshold cutoff mass @xmath144 gev / c@xmath69 in order to have enough kinematic phase space to produce bb pairs . we have investigated the sensitivity of the results to the value of parameters @xmath139 and @xmath145 and found no significant variation on pseudorapidity distributions of charged particles for 15 mb @xmath146 25 mb and for 4 gev / c@xmath69 @xmath147 6 gev / c@xmath69 . baryon number transport is quantified in terms of the rapidity loss ( @xmath148 , where @xmath149 and @xmath150 are the rapidity of incoming beam and outgoing baryon , respectively ) and has been discussed within our model phenomenology for @xmath119 collisions in refs . it was shown that hijing / bb v1.0 overestimate the stopping power and give a mild energy dependence of net - baryons at mid - rapidity . the energy dependence of net - baryons at mid - rapidity per participant pair within hijing / bb v2.0 is proportional to @xmath151 , similar to the dependence predicted in ref . @xcite , with the assumption that jj is the dominant mechanism . this dependence is obtained assuming the following parameters @xcite : @xmath152 gev@xmath69 ( the usual parameter of regge theory ) , @xmath153 ( the reggeon intercept of the trajectory ) , and @xmath154 ( where @xmath155 ) for the pomeron intercept . [ h ! ] distributions ( a ) at mid - rapidity for the @xmath156 ratio at @xmath7 = 0.9 tev ; rapidity distributions are shown in part ( b ) . the results are from three possible scenarios : without contributions from scf and jj loops ( dotted histograms ) , including only the effect of jj loops ( dashed histograms ) and including both effects ( scf and jj loops ) ( solid histograms ) . the data are from ref . @xcite ( alice collaboration ) . only statistical errors are shown . [ fig : pp_rap_pbar_pt ] , title="fig:",height=151 ] 0.5 cm recently , the alice collaboration reported results @xcite on mid - rapidity anti - proton - to - proton ratio in @xmath8 collisions at @xmath157 and 7 tev and equivalently the proton - anti - proton asymmetry , @xmath158 . these data could be used to constrain regge inspired model descriptions of baryon asymmetry . the authors state that , within statistical errors , the observed @xmath156 ratio shows no dependence on transverse momentum or rapidity in the limited measured acceptance ( @xmath159 ; @xmath160 gev / c ) . in fig . [ fig : pp_rap_pbar_pt ] ( a ) are compared the hijing / bb v2.0 model predictions with the published data at 0.9 tev . our model predicts negligible dependence on @xmath14 for @xmath161 gev / c , and a slight @xmath14 dependence at higher @xmath14 , where both effects ( jj loops and scf ) could contribute . the full calculation is shown by the solid histogram . the dotted line is the prediction without jj loops and scf effects while the dashed line includes the effect of jj loops only . over the measured ranges the rapidity distribution of the ratio @xmath156 distribution is not sensitive to the various scenarios presented . the scenario with combined effects results in a wider rapidity distribution at @xmath162 ( b ) . the narrow structure observed near @xmath163 has no physical significance and we believe that it is likely due to a numerical artifact of our current implementation of fragmentation scheme . separate proton and anti - proton rapidity distributions are , however , much more sensitive to scf effects , as seen in fig . [ fig : pp_p_pbar_rap ] . the model predicts a substantial increase ( by a factor of @xmath164 ) for p(@xmath105 ) , when scf are taken into account . due to the high cutoff mass @xmath165 gev / c@xmath69 the effect of jj loops is very small over the entire rapidity region . analysis of the 7 tev data leads to similar conclusions , but shows less effect at high @xmath14 . [ h ! ] = 0.9 tev . the histograms have the same meaning as in fig . [ fig : pp_rap_pbar_pt ] . [ fig : pp_p_pbar_rap ] , title="fig:",height=151 ] 0.5 cm over the measured range the feed - down - corrected data for the @xmath156 ratio rises from @xmath166 at 0.9 tev to @xmath167 at 7 tev @xcite . although the measured mid - rapidity ratio is close to unity there is a small but significant excess of protons over anti - protons corresponding to an asymmetry of @xmath168 and @xmath169 at @xmath170 and 7 tev , respectively . within our model using a regge intercept of @xmath171 = 1/2 results in a proton - anti - proton asymmetry @xmath172 = 0.032 ( 0.002 ) at 0.9 tev ( 7 tev ) . the result overestimates the value of asymmetry at 0.9 tev . these values show that a fraction of the baryon number associated with the beam particles is transported over rapidity intervals of less than seven units . however , a better understanding of the baryon transport would require measuring the rapidity dependence of the asymmetry over a larger range . in ref . @xcite we discuss the predicted rapidity correlation length for production of baryon - anti - baryon pairs within our model . the predicted rapidity correlation length ( @xmath173 depends upon the value of the regge intercept @xmath174 @xcite . a value of @xmath175 leads to rapidity correlations in the range @xmath176 , while a value @xmath177 @xcite is associated with infinite range rapidity correlations . this kind of analysis and future measurements at lhc energies will help us to determine better the specific parameters characterizing jj loops used in the calculations of baryon number transport and/or baryon - anti - baryon asymmetry while helping to establish the validity of regge inspired models . we have studied the influence of strong longitudinal color fields and of possible multi - gluon dynamics ( _ gluon junctions _ ) in particle production in @xmath8 collisions , with a focus on rhic , to tevatron , and lhc energies . we have investigated a set of observables sensitive to the dynamics of the collisions , covering both longitudinal and transverse degree of freedom . a detailed comparison with newly available experimental data from the lhc has been performed . we found that the inclusion of the multiple minijet source limits the growth of the string tension @xmath73 to be approximately only linear as a function of saturation scale @xmath59 ( eq . [ eq : kappa_sup ] ) , in contrast to recent approaches @xcite where @xmath73 scales as @xmath75 ( eq . [ eq : larry10 ] ) . the reason for this is that in the cgc model the collinear factorized minijet mechanism is suppressed by geometric scaling to much higher @xmath14 . future measurements at lhc energies ( @xmath7 = 7 and 14 tev ) , extended to high @xmath14 , will help to clarify the validity of this mechanism . we have shown that scf could play an important role in particle production at mid - rapidity in @xmath8 collisions . our calculations show that high - energy density fluctuations in @xmath8 collisions at lhc can reach densities comparable to those reached in central nuclear ( @xmath119 ) collisions at rhic . a large enhancement of the ( strange)baryon - to - meson ratios that persists up to the highest lhc energy can be explained as an effect of scf that appears at the initial stage of the interaction . the mechanisms of hadron production are very sensitive to the early phase of the collisions , when fluctuations of the color field strength are highest . strong color field effects are modeled by varying the effective string tension that controls the @xmath178 and @xmath179 pair creation rates and strangeness suppression factors . scf , therefore , may modify the fragmentation processes with a resultant huge increase of ( strange)baryons . we show that both jj loops and scf effects could play an important role in baryon ( anti - baryon ) production at mid - rapidity in @xmath8 collisions at lhc energies . introducing a new jj loop algorithm in the framework of hijing / bb v2.0 leads to a consistent and significant improvement in the description of the recent experimental results for proton - anti - proton and for baryon - anti - baryon asymmetry in comparison to the older versions hijing / b or hijing / bb v1.0 @xcite . we have shown that baryon number transport is suppressed for @xmath180 , a result that is confirmed by recent alice measurements @xcite . the present study is limited to the effect of initial state baryon production via possible junction dynamics in strong fields . it would be very useful to consider a generalization of back reaction effects @xcite to the case not only of pair production relevant for mesons but to the more difficult three string junction configurations needed to describe baryon production . a greater sensitivity to scf effects is expected also for open charm and bottom production @xcite . in particular , measurements of rapidity and @xmath14 distributions for particles involving charm and bottom quark , would provide an important test of the relevance of scf fluctuations , helping us to determine values of the suppression factors @xmath46 ( where q = qq , @xmath31 , @xmath32 , @xmath33 ) , which have strong dependence on the main parameters of qcd ( the constituent and current quark masses ) and on the system size . even though the success of this procedure has been clearly illustrated here , a fuller understanding of particle production and especially of ( multi)strange particles in ultra - relativistic @xmath8 collisions at the lhc remains an exciting open question , and will continue to challenge many theoretical ideas . 0.2 cm * acknowledgments : * we thank s. das gupta and s. jeon for useful discussions and continued support . we thank p. levai for helpful discussions and suggestions throughout this project . vtp acknowledges computer facilities at columbia university , new york , where part of these calculations were performed . this work was supported by the natural sciences and engineering research council of canada . this work was also supported by the division of nuclear science , of the u. s. department of energy under contract no . de - ac03 - 76sf00098 and de - fg02 - 93er-40764 . t. pierog , s. porteboeuf , i. karpenko and k. werner , to appear in the proceedings of 45th rencontre de moriond , qcd and high energy interactions , 13 - 20 march 2010 , la thuile , aosta valley , italy , arxiv:1005.4526 [ hep - ph ] . e. eichten , i. hinchliffe , k. d. lane and c. quigg , rev . mod . * 56 * , 579 ( 1984 ) [ addendum - ibid . * 58 * , 1065 ( 1986 ) ] . d. w. duke and j. f. owens , phys . rev . d * 30 * , 49 ( 1984 ) . d. acosta _ et al . _ , cdf collaboration , phys . d * 65 * , 072005 ( 2002 ) . h. ricaud , a. kalweit and a. maire , j. phys . g * 37 * , 094049 ( 2010 ) . r. j. fries , b. muller , c. nonaka and s. a. bass , phys . c * 68 * , 044902 ( 2003 ) . v. greco , c. m. ko and p. levai , phys . c * 68 * , 034904 ( 2003 ) . r. c. hwa and c. b. yang , phys . c * 67 * , 034902 ( 2003 ) . b. i. abelev _ et al . _ [ star collaboration ] , phys . c * 75 * , 064901 ( 2007 ) . d. e. acosta _ et al . _ [ cdf collaboration ] , phys . d * 72 * , 052001 ( 2005 ) . cms collaboration , cern cms physics analysis summary qcd-10 - 007 ( 2010 ) , http://cdsweb.cern.ch/record/1279344 ; keith a. ulmer for the cms collaboration , arxiv:1012.1605 [ hep - ex ] . k. werner , t. hirano , i. karpenko , t. pierog , s. porteboeuf , m. bleicher and s. haussler , nucl . suppl . * 196 * , 36 ( 2009 ) . k. werner , i. karpenko and t. pierog , j. phys . ser . * 230 * , 012026 ( 2010 ) .

the cms experiment at the large hadron collider ( lhc ) announced @xcite recently that , for standard model ( sm ) with four quark generations ( 4 g ) , the higgs boson in the mass range of 120 to 600 gev is excluded at the 95% c.l . " . the atlas experiment is in agreement @xcite . a common inference is that the 4 g itself is practically ruled out . afterall , cms also reported the most stringent bounds on the @xmath2 and @xmath3 quarks to date : @xmath20 gev @xcite and @xmath21 gev @xcite , both at 95% c.l . , which are rather close to the unitarity bound ( ub ) of 500550 gev @xcite . however , in as much as 4 g may not exist , an intriguing possibility @xcite is that electroweak symmetry breaking ( ewsb ) itself might be triggered by , or related to , the strong yukawa couplings of the @xmath2 and @xmath3 quarks . can the ub violation of strong @xmath22 scattering @xcite be related to the ub violation of strong @xmath23 and @xmath24 scattering ? to pursue such questions is a major purpose of the lhc , and is well within its abilities . if a relatively light higgs particle emerges soon at the lhc with sm cross sections , then 4 g would truly be in trouble @xcite . but , the exclusion statement of sm/4 g higgs might well get extended to the sm higgs itself with 2011 - 2012 lhc data . for higgs particle beyond 600 gev or so , one enters the strong @xmath22 scattering domain , and the higgs " becomes a broad object @xcite , which requires both high energy ( 14 tev ) and high luminosity to explore . for such a heavy higgs boson , if the @xmath2 or @xmath3 quarks were however found below 500550 gev or so , then the yukawa sector may not be strongly coupled enough to link with the strongly coupled higgs sector " . thus , we have in mind the scenario where neither the ( sm - like ) higgs boson , nor the 4 g quarks , are found below 600 gev and 500550 gev , respectively . it is important to remember that new @xmath25 violating phases associated with 4 g quarks may link to @xcite the baryon asymmetry of the universe ( bau ) . thus , the existence of a very heavy 4 g may touch both ewsb and bau , which are two of the greatest problems in particle physics . this provides strong motivation for continuing the pursuit of 4 g in this volatile time . if the higgs boson is heavy ( and fat " ) , while the 4 g quarks are above the ub , then whether the large yukawa coupling induces @xmath23 condensation @xcite or not , it would be important to explore possible _ bound states _ of this strong coupling . this is not just about potentially interesting lhc phenomenology , but may be necessary to provide a guide for the search of ultraheavy chiral quarks beyond ub . the main purpose of this paper is to explore lower lying bound states of strong yukawa coupling , and the associated properties . however , by venturing above the ub , one immediately encounters the perils of the breaking down of perturbation theory . thus , in lieu of genuine nonperturbative approaches , such as @xcite lattice field theory ( lft ) , our work is only of an illustrative kind . one aid to the study is a new , heavy isospin . if 4 g quarks exist , by which we mean a sequential left - handed doublet and a pair of right - handed singlets under the weak interaction , the @xmath26 , @xmath27 variables @xcite or electroweak precision tests ( ewprt ) require the @xmath2 and @xmath3 to be nearly degenerate ( and here , independently , a heavier higgs is also required @xcite ) . of course , some small splitting is needed to satisfy @xmath27 , but in this paper we will treat the @xmath2 and @xmath3 as degenerate , hence one has a new isospin " . this isospin , in contrast to the chiral limit of @xmath28 and @xmath29 quarks under qcd , is in the opposite limit of degenerate _ we thus borrow the notation of isovector @xmath30 , @xmath31 ( or @xmath32 $ ] ) and isoscalar @xmath33 , @xmath34 ( or @xmath35 ) etc . , for the heavy @xmath36 mesons " . as we will see , unlike technicolor , the @xmath31 " meson does not play a major role for yukawa bound states , nor does the @xmath33 . another interesting point is that , if the yukawa interaction is the dominant binding mechanism , since it is color blind , the @xmath36 mesons , unlike the qcd situation , come not only in color singlets , but color octets as well @xcite . we find the states @xmath37 , @xmath38 , @xmath10 and @xmath39 , where the subscript indicates the color representation , to be of phenomenological interest . in particular , @xmath37 and @xmath10 may be accessible in the near future . in sec . ii , we bring forth the issues of strongly coupled relativistic bound states . we contrast the necessarily relativistic @xmath40 coupling of weak nambu goldstone boson ( ng ) , or longitudinal vector boson exchange , to the coulombic qcd bound state , as well as scalar exchange . because the higgs scalar should be treated as heavy now , its effect is less prominent than ng exchange . we compare the traditional relativistic expansion @xcite with a relativistic bethe - salpeter approach @xcite , and illustrate why the yukawa bound state involves highly relativistic motion of its constituents . an issue appears regarding the treatment of @xmath41-channel ng exchange , towards strong yukawa binding . this , plus other issues , forces us to compromise in the study of the possible spectrum at this stage , and we restrict ourselves to @xmath42 in the range of 500700 gev , i.e. not far above the ub . in this way , the relativistic expansion provides a partial guide , while we offer a peek beyond and consider binding energies of order @xmath43 gev or more , for the possible spectrum around and above the tev scale . this leads us to focus on , from the production point of view , an @xmath37 state ( effectively a @xmath44 " , or gluon - prime ) , as well as possibly the @xmath10 and @xmath38 states in its decay final state . in sec . iii we turn to exploring the production and decay properties . we survey the key parameters needed : the binding energy , the vector pseudoscalar and octet singlet mass splittings , the vector meson decay constant @xmath45 , and the quark mixing element @xmath13 . production is mostly through @xmath46 and depend only on @xmath45 , but the annihilation , transition , and free quark decays involve all these parameters , where the numerical values we use are only illustrative . we find , in general , that the @xmath37 is relatively narrow , but has a host of decay final states , which might therefore elude early detection . we offer some discussion of the phenomenology at lhc in sec . iv , touching briefly on _ deep _ bound states , i.e. the possibility of binding energy approaching @xmath42 itself , for @xmath42 beyond 700 gev . we end with a conclusion and outlook . when the unitarity bound is reached for very heavy chiral quarks , it means that some _ tree level _ @xmath23 and @xmath24 scattering cross sections will violate unitarity , or conservation of probability , in the high energy limit . pointed out over 30 years ago @xcite , it is remarkable that we are now at the doorsteps to probe whether such new heavy chiral quarks exist . many , if not most people , tacitly take the ub as a kind of ceiling for sm - like chiral quarks to make sense . but in reality , crossing the ub just implies that the yukawa couplings are becoming so strong , they are turning nonperturbative . we have seen how the remarkable theory of qcd turns nonperturbative in the infrared , resulting in the rich phenomena of hadrons . we will not dwell on deeper short - distance ( uv ) gauge theories like technicolor , but just take the large yukawa couplings @xcite at face value : if chiral quarks @xmath47 ( a left - handed doublet of @xmath2 and @xmath3 in our case ) exist at or above the ub , what could be the emergent phenomena ? consider first heavy quarkonia bound by qcd . since qcd is perturbative at short distance , we have the familiar coulombic bound states with a @xmath48 enhancement factor , where @xmath49 is the velocity . this is the domain of non - relativistic qcd ( nrqcd ) , where one expands in velocity , which is of order @xmath50 . the nr nature makes good contact with the familiar atomic systems . exchanging higgs bosons brings in the yukawa couplings , which has been considered in the literature . for our case , we will treat the higgs boson as above @xcite 600 gev and heavy , which suppresses the binding effect due to higgs exchange . however , ng or longitudinal vector boson exchange ( transverse gauge boson exchange has coupling constant @xmath51 or @xmath52 , hence subdominant and largely ignored by us ) also couples with the yukawa coupling strength , but it involves the @xmath40 , which couples the upper and lower components of the massive dirac quark . since the lower component vanishes when the heavy quark is at rest , ng exchange is suppressed in the nr limit . conversely , the coupling to high momentum heavy quarks is large , the more so the heavier the quark . this reflects the derivative coupling nature of longitudinal vector bosons . the upshot is that , yukawa interactions between very heavy quarks are large when these quarks are in relativistic relative motion , i.e. with momentum @xmath53 . with this insight , and ignoring the higgs exchange for the moment , if the bound state formation is dominated by qcd , then the nr nature of the bound state ( @xmath54 ) actually suppresses the effect of yukawa coupling . however , as the yukawa coupling increases with @xmath42 , although the qcd - bound system becomes even more nonrelativistic , at some point ng exchange would ( perhaps suddenly ) take over , and one would find the bound state system turns ultrarelativistic . we shall illustrate with two different perspectives , one a traditional relativistic expansion @xcite , the other a relativistic bethe - salpeter approach @xcite . a standard approach in considering bound state phenomena is to make a relativistic expansion around the leading potential . the higgs potential for very heavy quarks , in the context of forming bound states , was considered a long time ago @xcite . the relativistic corrections were recently calculated in ref . the scattering amplitudes for @xmath17-channel higgs , neutral and charged ng ( called fictitious scalar in ref . @xcite ) , and gluon exchange , as well as @xmath41-channel ng exchange , were computed . @xcite , however , did not put in @xmath41-channel higgs and gluon exchange , even though both color singlet and octet @xmath23 configurations were discussed . touching both @xmath55 gev and @xmath56 cases , a variational approach was used to estimate the size @xmath57 ( equivalent to wavefunction ) and binding energy for @xmath58 , @xmath59 , color singlet and octet states . we will not repeat what was already done here , but just give some salient features . as a control on validity of the relativistic ( @xmath60 or @xmath61 ) expansion , the authors of ref . @xcite required @xmath62 ( @xmath63 is the relative momentum ) , which translates into @xmath64 . since @xmath17-channel gluon exchange gives the coulomb potential , it is clear that one has a coulombic qcd bound state when @xmath42 is not yet too large , with bohr radius @xmath65 . the low mass higgs case is no longer tenable with 4 g , both by direct search @xcite , and indirectly @xcite from ewprt due to the heaviness of 4 g quarks . for the heavy higgs case illustrated with @xmath56 , as one can see from fig . 8 of ref . @xcite , the radii @xmath57 of the @xmath38 , @xmath66 and @xmath67 ( the subscript @xmath68 stands for color singlet ) states decrease rather slowly below @xmath65 as @xmath42 increases . the @xmath39 radius rises slowly above @xmath65 due to the extra repulsion it receives from @xmath41-channel ng exchange . the yukawa effect is subdominant compared with qcd , where inspection of the binding energies offer further support : they rise slowly from @xmath69 gev around @xmath70 gev , to @xmath71 gev at @xmath72 gev , with @xmath39 only slightly lower . however , just before @xmath42 reaches 400 gev , radius @xmath57 for @xmath38 drops precipitously , while the binding energy rises sharply from around @xmath73 gev , to around @xmath43 gev at @xmath74 gev . the condition @xmath64 is violated shortly above @xmath75 gev . the kink " around 400 gev is the point where the ng exchange has wrestled the mechanism for binding away from the usual nr qcd binding . the sudden drop of the size of the bound state is due to tapping the large attraction at large momentum @xmath76 for the bound quark ( besides the @xmath40 coupling of the ng bosons , the heavy higgs also defines a very short range for the strong yukawa coupling to be effective ) . the resulting larger binding energy overcomes the much increased kinetic energy . for @xmath66 , the onset is delayed until @xmath77 gev or so , with binding energy of @xmath78 gev at @xmath79 gev . for color octet states , which are not bound by qcd , the binding energy for @xmath37 , due purely to yukawa coupling , turns on sharply around @xmath80 gev , rising to @xmath81 gev for @xmath79 gev . the @xmath82 state turns on much later , around @xmath83 gev . but , unlike @xmath39 , because there is no @xmath41-channel repulsion , @xmath10 is degenerate with @xmath37 . it should be noted that the yukawa effects of neutral and charged ng exchange is weaker but constructive for @xmath37 , while the converse is true for @xmath10 , so this degeneracy could be accidental . furthermore , this degeneracy should be lifted by @xmath41-channel gluon annihilation , which would raise @xmath84 but was not considered by the authors of ref . @xcite . as we shall soon see , the vector channel also has @xmath26- and @xmath85-wave mixing . although identifying the @xmath38 as the lightest color singlet , and @xmath37 ( and @xmath10 ) as the lightest color octet , it is ironic that the relativistic expansion breaks down almost immediately after the strong yukawa binding becomes potent . but it does illustrate that one needs a genuine relativistic approach in treating strong yukawa binding . we turn to such an approach that is in principle nonperturbative , but carrying its own dubious features : the bethe - salpeter ( bs ) equation . a long time ago , around the time the ssc ( superconducting super collider ) was under construction but then unfortunately canceled , the authors of ref . @xcite pursued the bs equation approach to the relativistic bound states of very heavy sequential 4th generation quarks . it consists of a ladder sum of the scattering amplitudes that appear in the relativistic expansion . in the heavy isospin limit and treating @xmath86 , a clear isospin reorganization separates into @xmath87 and @xmath88 mesons " . for @xmath36 meson with total momentum @xmath89 and relative momentum @xmath76 , one has a set of integral equations , with loop momentum @xmath90 , where @xmath91 is the momentum exchange in @xmath17-channel . however , for @xmath41-channel annihilation contribution , the annihilation momentum is @xmath89 itself , and the integral over loop momentum @xmath90 carries no @xmath76 dependence , giving a possibly divergent constant . to remedy this , ref . @xcite took a fixed @xmath76 subtraction at @xmath92 . in this way , all the @xmath41-channel diagram contributions were eliminated from the bs equation . this includes even the @xmath41-channel gluon exchange for the octet , isosinglet vector channel , which was not considered in ref . @xcite , as the authors concerned themselves with color singlet states only . in terms of mathematical physics , to set up integral equations to be solved in a self - consistent way , the subtraction at constant @xmath76 seems reasonable . however , as admitted by the authors of ref . @xcite in a footnote , the @xmath41-channel ng exchange leads to repulsion . thus , in discussing bound state solutions , there is the issue of the physical correspondence , and therefore the range of validity ( in @xmath42 ) for implementing the subtraction . in the relativistic expansion , one clearly would not drop the @xmath41-channel diagrams . our purpose here is not to make a full treatment of the bs equation , as it is only a ladder sum of @xmath17-channel exchange diagrams , with higher order corrections ignored . furthermore , while the bs equation is relativistic , its solution depends very much on the approximations and ansatz made . ref . @xcite illustrated with covariant gauge ( but employing a weak coupling relation between spatial and temporal spinor components ) , the instantaneous approximation with positive frequency potential only , as well as keeping both positive and negative frequency potentials . although the numerical solutions share common features , the bound state mass ( @xmath93 ) values differ . starting all from @xmath94 for low @xmath42 , they decrease smoothly as @xmath42 increases , without exhibiting the kink seen in the relativistic expansion . as such , the bs approach is an improvement . but , as a common feature , once the binding energy @xmath95 becomes significant ( e.g. @xmath43 gev or so ) , at high enough @xmath42 , the low lying mesons collapse . that is , the binding energy becomes so large such that the total mass drops to zero . ( solid ) and @xmath96 ( dashed ) potentials in eq . ( 1 ) for the @xmath39 state , for @xmath97 gev , and @xmath98 gev , plotted along @xmath99 . both vanish for @xmath92 , or zero relative motion , but go to rather large values for large @xmath76 , with @xmath96 turning on rather sharply for @xmath100 gev . , title="fig:",width=302 ] having bound state solutions numerically turning tachyonic for a strongly coupled system is not particularly astounding . for qcd ( and likewise for qed ) , quarkonium masses calculated at fixed order could also vanish at large enough coupling strength . the system has turned fully relativistic with strong coupling , and the familiar bohr - schrdinger solution is no longer valid . for the bs equation , however , it is already relativistic . the collapse of the meson state may be related to the symmetry breaking itself @xcite , but because of the approximate nature of the bs equation , as well as the numerical approximations made in its solution , we refrain from dwelling further on this . rather , we wish to use the bs equation and its numerical study to compare and contrast with the previously discussed relativistic expansion , to project what may really happen for relativistic , strong yukawa bound states , before any collapse " , such as illustrated in ref . @xcite , could occur . for this purpose , we note that with the _ subtracted _ bs equation , hence with the @xmath41-channel repulsion removed , the @xmath39 turns out to be the most attractive system , i.e. the lightest bound state : the ng , gluon and higgs exchange are all attractive , as can be seen also from ref . the behavior for @xmath38 is indeed similar , but the formalism is more complicated for the @xmath101 system , where the @xmath26- and @xmath85-wave channels are coupled . note that ref . @xcite used ( sometimes tacitly ) a higgs mass around 100 gev , which is no longer valid . in our mind , we are interested in the bound state dominated by yukawa coupling , i.e. by ng exchange , hence we will view gluonic exchange as correction , with higgs exchange perhaps even milder . this matches to what one finds in the relativistic expansion , that the yukawa binding suddenly turns on , bringing on a rapid rise in binding energy . in effect , we use the bs equation approach to check , and probe beyond , the kink " seen in the relativistic expansion . the equation for @xmath39 is the most compact . in the impulse approximation , integrating over @xmath102 gives the amplitude @xmath103 , where @xmath104 is the relative momentum . keeping both positive and negative frequency amplitudes @xmath105 , one has the coupled equations , @xmath106 , \label{bspi1}\end{aligned}\ ] ] where @xmath93 is the eigenvalue , and @xmath107 . the potentials @xmath108 , where we have absorbed a factor of @xmath109 into its definition as compared to ref . @xcite , arise from @xmath17-channel diagrams as described earlier . let us understand the @xmath110 and @xmath111 potentials . the less familiar one is @xmath111 , which couples @xmath112 to @xmath113 , while @xmath110 is diagonal " . in the limit that @xmath114 , one simply has @xmath115 , where @xmath116 is analogous to the expectation value of the potential ( @xmath117 is like @xmath118 ) of nonrelativistic quantum mechanics . on the other hand , if @xmath119 , then @xmath120 , hence the @xmath111 contribution is more suppressed than the corresponding @xmath110 contribution when it is weak . the potentials @xmath121 are symmetric in @xmath76 and @xmath90 , and is steepest along @xmath99 . we plot @xmath122 in fig . [ vpm ] , for heavy @xmath98 gev , and for @xmath75 , 500 , 600 , 700 gev . we have checked that the @xmath123 potential drops by a factor of two or more as @xmath124 moves from 100 to 700 gev , but @xmath96 is rather insensitive to @xmath124 . we see from fig . [ vpm ] that both the @xmath123 and @xmath96 potentials are suppressed for low relative momentum @xmath76 , agreeing with the relativistic expansion view . as @xmath76 increases , @xmath123 increases relatively slowly , reaching @xmath125 at @xmath126 ( 700 ) gev for @xmath74 ( 700 ) gev . but @xmath96 turns on more sharply , reaching beyond @xmath127 at @xmath128 gev , and @xmath129 at @xmath130 gev . so what does the strength of @xmath131 mean ? take the @xmath132 equation , the binding energy @xmath133 is equal to @xmath134 , the kinetic energy due to motion , plus the right hand side of eq . ( [ bspi1 ] ) , which should be the potential " . let us normalize @xmath90 by @xmath42 . since @xmath135 is steepest for @xmath99 , we see that @xmath136 for @xmath137 means the potential " is comparable to @xmath42 in strength when the momentum is comparable to the rest mass , and the kinetic energy due to motion is @xmath138 . we can now sense why there is a tendency for the @xmath39 state to collapse already for @xmath42 not far above 500 gev in the numerics of ref . @xcite . we have illustrated how the binding energy for the ( heavy ) isovector , color singlet @xmath140 state , which we call @xmath39 , could be already moving towards collapse for @xmath42 as low as 500 gev ( this number from ref . @xcite contains the attraction due to light higgs exchange ) , if the bs approach of eq . ( [ bspi1 ] ) holds true . however , this state receives a repulsive @xmath41-channel ng annihilation contribution . furthermore , as the tendency towards collapse approaches , the ladder sum bs equation may no longer be sufficient . even within the bs approach , where eq . ( [ bspi1 ] ) gives rise to the earliest collapse , if one drops the negative frequency amplitude , the collapse is delayed @xcite by almost a factor of two in @xmath42 . in the covariant gauge but using a weak coupling relation between temporal and spatial spinor components , the collapse of @xmath39 occurs slightly before the positive frequency only case . it is not clear at what @xmath42 the collapse truly occurs numerically . in any case , we would not touch the collapse here , as it is not yet well understood . it seems prudent , then , to consider binding energies of @xmath43 to @xmath141 gev , but not more , to make a preliminary study of possible phenomena at lhc , running at @xmath142 tev . a similar equation as eq . ( [ bspi1 ] ) holds for the isosinglet , color singlet @xmath140 state , the @xmath67 . although gluon and higgs exchange are attractive , ng exchange turns repulsive @xcite . thus , @xmath67 is much heavier than @xmath39 , and can not be a low - lying state . turning to vector mesons , ref . @xcite found , similar to the relativistic expansion of ref . @xcite , that the isoscalar @xmath38 has the tightest binding , though it is weaker than ( @xmath41-channel subtracted ) @xmath39 . this concurs with our earlier observation that yukawa effects are weaker but constructive for vector , while stronger but destructive for pseudoscalar . however , there is no repulsive @xmath41-channel effect for @xmath38 . on the other hand , checking the formalism , we find that there is @xmath143-@xmath144 mixing , resulting in a set of more complicated coupled bs equations . from the numerics of ref . @xcite , we expect that for @xmath42 in the range of 500700 gev , binding energies of order @xmath43 to @xmath141 gev is reasonable for @xmath38 . the @xmath66 state , analogous to @xmath67 in receiving the repulsive ng exchange , is far less binding , which is also seen from the relativistic expansion . turning to color octet counterparts , one can treat gluon exchange as a perturbation when binding energies are much large than qcd bound states . given that the bound state should be rather small , the repulsion due to gluon exchange should be larger than typical qcd binding . but without fully solving for the bound state , octet - singlet splitting is uncertain . however , there should be no doubt that @xmath10 and @xmath37 would be the low - lying octet states , in agreement with the relativistic expansion . the octet @xmath145 and @xmath82 are likely not ( or at best weakly ) bound , hence we do not consider these . to summarize , the states to keep in mind are @xmath39 , @xmath38 , @xmath10 and @xmath37 . there are possibly other states with different quantum numbers , but in general they would not be lighter , while likely possessing more complicated properties . the absence of @xmath31 states is distinct from qcd - like gauge theories such as technicolor . for @xmath42 in the range of 500 to 700 gev , staying short of very tight binding , these states would probably populate the 1 to 1.4 tev range , with binding energies of order @xmath43 to @xmath141 gev . the ordering of the spectra , according to ref . @xcite ( which did not actually consider octet states ) , would be @xmath146 . since we are concerned with lhc phenomenology and the heavy quark search program in the near future , we should consider briefly issues regarding production : * @xmath39 and @xmath38 can not be produced via @xmath147 fusion , but can be produced via weak drell - yan processes , hence have a weak production cross section ; * @xmath10 can be pair produced by @xmath147 and @xmath7 scattering @xcite , but it is heavier and less efficient at 7 tev . this leaves @xmath37 , which has the same quantum numbers as the gluon , that is the most attractive in the near future in terms of production cross section . it can not be produced by @xmath147 fusion , as two massless vectors can not forge a massive vector ( yang s theorem ) particle , hence the production is limited to @xmath7 fusion . in the next section , we turn to the production and decay properties of the @xmath37 meson . note that if an @xmath145 or @xmath67 state existed , such as for qcd binding , it could tap @xmath147 production @xcite . in this section , we discuss the production and decay of the @xmath148 meson , which we have portrayed as the likely leading harbinger of 4 g bound - states in our scenario , where a very heavy quark _ isospin _ symmetry prevents the production of some mesons directly at hadron colliders . for the numerical study of the production cross - sections and decay rates , the following parameters are needed as input : the decay constant @xmath149 , the 4 g quark masses @xmath150 ( we assume degeneracy of @xmath2 and @xmath3 as a simplifying approximation , hence an isospin symmetry ) , the mass of resonances , _ i.e. _ the binding energy of resonances , and finally the quark mixing elements @xmath151 . once the constituents as well as their interactions are fixed , the decay constant and binding energy are the consequence of the dynamics of the system . as we stressed previously , in the range of strong yukawa couplings we consider , their estimation has to be done by lft approach @xcite for more quantitative understanding . however , such an endeavor is beyond the scope of this paper . instead we parameterize them in our phenomenological study . the decay constant of @xmath148 is defined through @xmath152 which we parameterize by a dimensionless parameter @xmath153 . the mass @xmath84 is the most important , as we discuss the production and decay of @xmath148 . we also assume @xmath154 is the dominant quark mixing element , and ignore mixings with lighter generations . at hadron colliders.,width=113 ] the dominant production mechanism for @xmath148 is , as illustrated in fig . [ fig : prod ] , via @xmath155 annihilation . because one can not fuse two massless gluons into a massive vector boson , gluon - gluon fusion is not operative . we have also computed the higher order @xmath156 process as a check . at the parton level , the total cross section is @xmath157 convolving with the parton luminosity @xmath158 we get the hadronic cross section , @xmath159 we plot on the left - hand side of fig . [ fig : pro ] the cross section for inclusive @xmath148 production at the lhc with @xmath160 tev . we use cteq6l @xcite parton distribution functions ( pdfs ) , and set the renormalization and factorization scales to @xmath161 . we use three values of @xmath162 , 0.03 and 0.01 for illustration . although we do not have any suggestive estimation for the decay constant , since the yukawa bound state is highly relativistic , we expect the larger ( smaller ) value of @xmath163 corresponds to a stronger ( weaker ) bound meson . we note that the decay constant divided by meson mass for usual @xmath31 , @xmath164 and @xmath165 are @xmath166 . however , we are not dealing with usual qcd - bound meson production @xcite , so we leave @xmath163 as a parameter . the cross section is proportional to @xmath167 and decreases with increasing @xmath84 . because of our ignorance of the decay constant @xmath149 , the cross section @xmath168 ranges from pb to fb , for @xmath84 ranging from 900 to 1400 tev . the plot extends to 2 tev , since it depends only on @xmath163 , which we view should be experimentally determined . as the gluon density at the lhc is large , we have checked the higher order @xmath156 scattering process , and find the contributions to be quite small for the region of our interest , @xmath169 tev . for comparison , we also plot the open @xmath23 pair production cross section at lo and nlo @xcite , as a function of @xmath170 , matching ( approximately ) to @xmath171 on the same plot . the cross section should be multiplied by two to take into account the production of the degenerate 4 g doublet . in nlo calculation , we use cteq6 m pdfs @xcite . to see the uncertainty of the cross section , as well as the size of nlo correction , we vary the scales in each calculation from @xmath172 to @xmath173 . the uncertainty for lo ( nlo ) prediction is expressed as the dotted ( hatched ) band . at the lhc running at 7 tev ( left ) and 14 tev ( right ) for various @xmath153 values . the open @xmath23 cross sections at lo and nlo are also plotted for comparison.,width=306 ] at the lhc running at 7 tev , the @xmath148 production cross section with @xmath174 is about the same order as twice the open @xmath175 production cross section in the lower mass region , but exceeds the latter in the higher mass region . on the other hand , when @xmath163 is smaller , the cross section could be well below the open production cross section . thus , @xmath148 production can be interesting if @xmath163 is sizable , such as of @xmath176 order . note that @xmath37 is produced through @xmath7 , while open @xmath23 is produced dominantly through @xmath147 fusion . we plot on the right - hand side of fig . [ fig : pro ] the @xmath148 cross section for lhc at @xmath177 tev , which is an order of magnitude larger than those for lhc at 7 tev . the open production cross section grows relatively larger than the @xmath148 production because of increase in gluon luminosities . the decay channels of @xmath148 we consider are * annihilation decay : @xmath178 , @xmath179 ; @xmath180 , @xmath181 ; * free - quark decay : @xmath182 , @xmath183 ; * meson transition : @xmath184 ; @xmath185 . in the following , we discuss each decay mode separately . first , @xmath148 can decay into dijets or @xmath179 by the co - annihilation of the 4 g quarks inside the bound state through qcd , as shown in fig . [ fig : anni ] . these are the reverse processes of the production mechanism , therefore the existence of these decay modes is robust . into light quark pair ( dijets ) or @xmath179 . , width=113 ] the decay partial width in this mode is proportional to @xmath167 . the two - body decay width is calculated at born level to be @xmath186 where @xmath187 is the number of light quark flavors , and @xmath188 is the velocity of the top quark in the @xmath148 rest - frame . due to the number of light quark flavors , the decay partial width into dijets is larger than that into @xmath179 by @xmath189 . analogous to @xmath190 production , we have also estimated the three - body @xmath191 decay rate , following the tree - level calculation of ref . @xcite , and found that it can be ignored . another type of annihilation decay channel is caused by weak boson exchange , where an off - diagonal quark mixing element acts on one of the bound quarks , as shown in fig . [ fig : weak](a ) . this is rather similar to the binding mechanism , where the quark mixing elements are always within 4 g . once the cross - generation interaction occurs , the recoil due to the energy release from the mass difference between the 4 g quarks and lower generations would eject the lighter quark and destroy the bound state , followed by subsequent decay of the leftover 4 g quark . the lower generation quark mass is too light such that it can not bind with the heavy 4 g quark by yukawa coupling . the decay width is calculated in terms of the decay constant in eq . ( [ eq : dec ] ) using the fierz identity for the current products , giving ( @xmath192 ) @xmath193 where @xmath194.\end{aligned}\ ] ] with @xmath195 . the charge conjugate decays have the same partial width as above . we have actually performed a full calculation , but set @xmath196 , or weak coupling @xmath51 , to zero at the end , as we are concerned with longitudinal vector boson exchange . we have also not distinguished between @xmath197 and @xmath84 , where the latter provides the scale parameter . the decay rate depends on both the off - diagonal quark mixing element @xmath154 as well as @xmath163 . in this mode , @xmath148 decays into on - shell @xmath8 or @xmath9 ( and conjugate ) . if we restrict to @xmath198 and @xmath199 for the decay of 4 g quarks , the final state all end up as @xmath200 . the signal is similar to @xmath179 production , but the kinematical distribution differs from the standard model counterpart . a second type of decay mode is induced by the decay of the constituent quarks , as illustrated in fig . [ fig : weak](b ) . it is similar to the weak boson exchange decay discussed just before , and quantum mechanically speaking , the exchanged boson escapes the system as an on - shell particle . we call this the free " quark decay mode , even though the decaying quark is bound . the decay partial width in this mode depends crucially on @xmath13 and @xmath12 , but not strongly on the structure of @xmath148 . this last statement would no longer hold when one enters the realm of deeply bound states , where binding energy is much larger than the @xmath43 gev adopted here . inside the bound state system , the decay of the constituents is suppressed by phase - space and time dilatation effects @xcite . that is , the decaying quark constituent is off - shell . however , for simplicity , we ignore these effects in our rate calculation , and use @xmath201 = \gamma_{t'}+\gamma_{b'},\end{aligned}\ ] ] where @xmath202 , @xmath203 are given at born level as @xmath204 with @xmath205 or @xmath206 , and @xmath207 with @xmath208 as defined earlier . note that we have kept @xmath196 here , since the decay process is quite similar to the familiar top quark decay . the decay width of the 4 g quarks is suppressed by the small @xmath13 and @xmath12 , but grows rapidly with the 4 g quark mass . if @xmath148 decays through @xmath2 , the final state would be @xmath200 , and @xmath209 if decay is through @xmath3 . the search of these signal can be along the standard 4 g quark search strategy @xcite , except that , if heavy quark mass could be reconstructed , then for example one @xmath210 pair has a lower mass ( due to binding energy ) than the other @xmath210 pair from on - shell @xmath2 decay @xcite . , and @xmath211 . , title="fig:",width=132 ] , and @xmath211 . , title="fig:",width=132 ] finally , a third class of decay is for @xmath148 to turn into other resonances . we consider the two channels of @xmath212 and @xmath184 . the other possible channel @xmath213 is forbidden by the heavy isospin . the partial width for these decays depend on the mass difference as well as the transition amplitude of these resonances . the meson transition to @xmath185 would open only if @xmath214 . for @xmath215 , @xmath148 can decay into @xmath216 or @xmath217 through the off - shell @xmath15 boson . however , the partial width would be negligibly small . we can write down a general vector to pseudoscalar transition amplitude via ( color singlet ) vector or axial - vector currents as , @xmath218,\end{aligned}\ ] ] where @xmath219 , @xmath220 are 4-momentum of @xmath148 , @xmath221 respectively , @xmath222 is the polarization vector of @xmath148 , and @xmath223 and @xmath224 are form factors in @xmath225 a straightforward calculation gives ( @xmath192 ) @xmath226 with @xmath227x(1-x^2+y^2)\lambda^3 { \nonumber}\\ & \ ; + \frac{1}{4}|a_2+a_3|^2\lambda^5 + 2|v^2| \ , y^2\lambda^3 , \label{eq : hpw}\end{aligned}\ ] ] where @xmath228 as already defined . here , @xmath196 has to be kept , since transverse @xmath15 emission has a @xmath61 phase space factor , while longitudinal @xmath15 emission has a @xmath229 factor and more suppressed . using the assumption that a free quark inside a meson interacts with currents , the form factors are reduced to @xmath230 , @xmath231 , @xmath232 and @xmath233 . for simplicity , we use this limit in our numerical calculation , and eq . ( [ eq : hpw ] ) reduces to @xmath234 where @xmath235 . a vector to vector transition amplitude via color octet vector current is parameterized as @xmath236,\end{aligned}\ ] ] where @xmath222 , @xmath237 are the polarization vectors of @xmath148 , @xmath238 , respectively , and @xmath239 are form factors in @xmath240 . a straightforward calculation gives ( @xmath192 ) @xmath241 where we take the scale of the strong coupling constant to be at the mass of @xmath148 , and @xmath242 the second step follows from taking the free - quark limit as described above , which reduces the form factors to @xmath243 and @xmath244 . we use this result , which is highly suppressed by phase space , for our numerical estimation . to perform numerical studies of the branching ratios and decay widths , we finally have to specify the numerical values of the following parameters : @xmath163 , @xmath13 ( @xmath245 ) and @xmath246 for @xmath247 ( @xmath248 for @xmath249 ) . without a full solution to the relativistic strong coupling bound state problem , however , it is difficult to ascertain the values for @xmath153 as well as the mass splittings , and @xmath84 itself . we therefore examine four sets of parameters as a survey , * case 1 : @xmath174 , @xmath250 gev , @xmath251 ; * case 2 : @xmath252 , @xmath250 gev , @xmath251 ; * case 3 : @xmath174 , @xmath253 gev , @xmath251 ; * case 4 : @xmath174 , @xmath250 gev , @xmath254 . these are chosen simply to emphasize large variety of possible dominant decay channels . in all the cases we set the binding energy of @xmath148 to @xmath255 gev , which is much larger than qcd binding . a different choice of the @xmath148 binding energy changes our results only modestly . for the choices specified in case 1 , we assume the larger decay constant , @xmath256 gev mass difference in the meson spectrum , and @xmath13 is set to the nominal current upper limit @xcite . in case 2 , we examine the smaller decay constant . case 3 is for larger mass difference , and case 4 is when @xmath13 is more suppressed . we will discuss the two different mass differences ( vector pseudoscalar and octet singlet ) as variations . we plot in fig . [ fig : br ] the branching ratio of various @xmath148 decays as a function of @xmath84 for cases 1 to 4 . in case 1 , the dominant decay modes are the transition decay into @xmath185 , especially for lighter mass region , and free quark decay , i.e. via the decay of 4 g constituent quark for heavier mass region . the branching ratios of free quark decay and the @xmath13-dependent annihilation ( @xmath15 boson exchange ) decay increase with @xmath84 , because @xmath257 reflects a larger yukawa coupling . the @xmath7 is of order 10% and drops slightly at higher @xmath84 , with @xmath179 branching ratio a factor of 5 lower , at the percent level . the transition decay into @xmath258 is at the percent level or less . in case 2 , because of the small decay constant , the annihilation decay channels @xmath8 , @xmath9 , @xmath7 and @xmath259 are suppressed . in this case , free quark decay and transition decay into @xmath185 are the two predominant modes . in case 3 , the large mass differences enhance the branching ratio of the transition decays , and the @xmath185 mode dominates . the other transition decay into @xmath258 can also be enhanced , especially in the lighter mass region . it could be that the mass difference of only @xmath260 is large , i.e. when the mass spectrum is like @xmath261 . if so , @xmath185 could be considerably suppressed , and @xmath258 would be more prominent , especially for low @xmath84 . in case 4 , the free quark decay and the @xmath13 induced annihilation decay are suppressed , due to small @xmath13 . the decay width of 4 g quarks is also suppressed for the same reason . in this case , the transition decay into @xmath185 dominates , and the annihilation decay into dijets can be sub - dominant with branching ratio at ten percent order . however , this sensitively depends on the @xmath262 mass difference , as well as the decay constant . if @xmath185 becomes kinematically suppressed , dijets would be dominant . let us summarize some general features regarding branching ratios . basically , the transition decay into @xmath185 is large , because of the large yukawa coupling and no suppression effect by bound state deformation . this decay mode can be more enhanced if the mass difference is large , but much suppressed if the mass difference is small , especially if less than @xmath196 , as we have seen . free quark decay has sizable contribution for the heavier mass region , if @xmath13 is close to the current upper limit of 0.1 . 0.13 cm and @xmath221 for the four parameter sets . the upper curves are for @xmath148 , where case 3 is enhanced by a larger @xmath262 mass difference . the lower curves , plotted 100 gev less in mass , are for @xmath221 , where cases 1 and 2 are barely distinguishable.,title="fig:",width=283 ] in fig . [ fig : wid ] , we show the total decay width of @xmath148 as a function of @xmath84 for the four parameter sets . the decay width increases with @xmath84 , from a few gev to around 10 gev . for case 3 , due to the rapid decay into @xmath185 from a relatively large @xmath262 mass difference , the width is at 20 gev range , and increases mildly with @xmath84 . still , compared with its tev scale mass , @xmath148 is a heavy but narrow meson resonance . we see that the binding energy , therefore strong yukawa dynamics , and flavor structure through @xmath13 all play crucial role for the eventual phenomenology one expects at lhc . to be able to address lhc phenomenology , we need to treat @xmath38 and @xmath10 further , as they may appear in @xmath37 decay final state . from fig . [ fig : br ] we see that in general @xmath263 is the leading decay . note , however , that we have assumed @xmath264 gev . the rate would drop sharply as this vector scalar splitting diminishes , and becomes practically negligible when @xmath15 turns virtual . on the other hand , if the strong binding found by ref . @xcite in the bethe salpeter approach with @xmath41-channel subtracted has any bearing , then @xmath10 may be deeper bound than @xmath37 . for this situation , case 3 stands as an illustration , where @xmath265 decay would be preeminent . in contrast , the process @xmath266 is never more than 10% , and more typically at @xmath267 order or smaller . the exception would be if @xmath268 is of order @xmath196 or less , but @xmath269 is sizable ( plot of case 3 in fig . [ fig : br ] , but with @xmath265 removed ) . viewing this exception as unlikely , we relegate the discussion of @xmath38 to a concurrent discussion of weak drell yan production . but we need to address how @xmath10 decays . it is interesting that @xmath270 vanishes because it is a @xmath271 transition , which we have verified by direct computation . the @xmath272 exchange diagram of fig . [ fig : weak](a ) is absent for charged @xmath273 ( i.e. @xmath274 and @xmath275 mesons ) because of isospin , while the @xmath41-channel annihilation is absent by the octet / isovector nature . the upshot is that we are left with only two decay processes : the familiar free quark decay , and a new type of decay , @xmath276 , where the @xmath15 is transverse . the latter is an inverse process of @xmath265 , with @xmath51 replacing @xmath148 . however , the annihilation rather than transition nature implies that @xmath276 is relatively suppressed . for the four parameter sets , ranging from predominance ( case 4 ) , to below @xmath267 ( case 2).,width=272 ] direct computation gives @xmath277 where @xmath278 is the @xmath10 decay constant normalized by @xmath10 mass . the rate of @xmath279 is suppressed by @xmath280 compared to the rate of @xmath281 of eq . we have checked explicitly that longitudinal @xmath15 emission again vanishes . we plot the @xmath10 width in fig . [ fig : wid ] , and the @xmath282 branching fraction , in fig . [ fig : pi8 ] . the width is at gev order , narrower than @xmath37 , but could be much smaller if a small @xmath13 suppresses the free quark decay widths . thus , @xmath279 decay branching fraction is below 10% , and much smaller for case 2 ( suppressed by a smaller decay constant ) . for case 4 , the small @xmath13 case , @xmath279 could dominate . into dijets , for the four parameter sets at the lhc 7 tev . cms upper limit @xcite for the cross section of dijet resonance production is also plotted.,width=264 ] we now discuss the possible phenomenology , as well as other issues . let us first comment on the dijet decay of @xmath148 , which could appear as a dijet resonance . absence of dijet features in tevatron and lhc searches constrains or rules out any model with particles in tev range . in fig . [ fig : jj ] , we plot the total cross section times dijet branching ratio for @xmath148 production at the lhc 7 running at tev , as a function of @xmath84 for the four parameter sets . we compare with the cms dijet resonance search @xcite with @xmath283 fb@xmath284 data at @xmath160 tev . we note that the data includes acceptance cuts for dijets , but our model numbers do not , which makes the comparison conservative . we find the cross section times dijet branching ratio are at least an order of magnitude lower than the current upper limit for all four cases , with case 4 the largest . even with @xmath285 channel removed making @xmath281 the leading decay , one is still below the cms limit . let us call this special situation case @xmath286 . our results show that , while a narrow resonance signal might start to show up with a considerably larger data set , it could show up soon for case @xmath286 . that is , if @xmath285 decay is forbidden , while free quark and exchange decays are suppressed by small @xmath13 . the need for these two conditions to be met , however , makes this possibility not particularly likely . in general , the @xmath285 decay is the dominant @xmath148 decay mode , unless it is kinematically suppressed by @xmath262 being too close to , or smaller than , @xmath196 . we have investigated the decay of @xmath10 itself , and found that it is dominated by free quark decay , with @xmath287 subdominant for cases 13 . it is important now to spell out the isospin nature : isosinglet @xmath288 , @xmath289 with a @xmath290 ratio , where the subscript @xmath41 indicates a relatively soft vector boson , and @xmath291 stands for the @xmath292 boson , as we have ignored heavy isospin breaking . thus , the signature for @xmath293 leads to @xmath294 , @xmath295 , and @xmath296 , plus c.c . except for the last case with presence of a @xmath292 boson ( the identification via dileptons would be costly in branching fraction ) , the additional @xmath297 which is relatively soft makes it an even more complicated signature than direct open @xmath23 production . the same sign dilepton approach @xcite remains a good one , but more jets would be present . note that the @xmath292 is usually vetoed against in same sign dilepton studies . however , if boosted @xmath15 and top jets can be exploited to isolate the @xmath298 and @xmath299 ( both in @xmath300 final state ) , then the associated soft @xmath301 could be an extra tag for @xmath5 production . besides the relatively low @xmath302 @xmath15 , the 7-jet system has rich imbeddings of @xmath15 , @xmath16 and @xmath17 jets . it can be part of the @xmath303 search program , which would in any case be a background . but the multi - jet system is of rather high mass , and with different signature content , that one may be able to separate . if the total jet mass resolution is good , one may discover both the @xmath273 in multijets with a soft @xmath15 tag , with the @xmath273 and the @xmath304 reconstructing to @xmath37 . the lhc experiments should also look for a @xmath292 boson associated with high jet multiplicity , perhaps with a hint of unusual @xmath305 plus multijets as backdrop . given that @xmath221 is very narrow , one could look for a relatively soft @xmath292 recoiling from 6 or 8 hadronic jets with very high , but relatively specific mass , containing substructures of multiple boosted @xmath15 jets or boosted top jets . the whole event , @xmath306 6 or 8 jets , would also reconstruct to a narrow resonant mass . one might think that @xmath307 multijets can be similarly pursued . however , eq . ( [ eq : opw ] ) does not apply to @xmath308 : the @xmath309 term vanishes with @xmath310 . the photonic decay involves heavy quark spin flip , hence suppressed by @xmath311 . treating @xmath312 , since we allow @xmath313 gev at best , the radiative rate is @xmath314 . since a larger @xmath315 ( case 3 ) gives a larger total @xmath148 width , we find that @xmath316 is always below the percent level . however , photon detection does not suffer from the factor of 0.06 as for @xmath317 detection . hence , the lhc experiments might also consider @xmath307 multijet studies . case 4 offers yet again an intriguing signature , assuming @xmath285 decay dominance . from fig . [ fig : pi8 ] we see that @xmath287 is dominant , as free quark decay is suppressed by a small @xmath13 . one therefore has a unique signature of @xmath318 ( @xmath319 and @xmath320 ) . here , one @xmath15 is soft , and the other hard , with @xmath302 greater than 500 gev , accompanied by a similarly hard gluon . both @xmath15 s tend to be transverse , but the hard @xmath15-jet and the gluon jet would form a rather narrow resonance ! this case offers dramatic signature and should be quickly searched for . the production cross section , of course , is modulated by @xmath149 ( see fig . [ fig : pro ] ) . case 4 , which is the limiting case of small @xmath13 , has better likelihood than the even more special case 4@xmath321 discussed earlier . leaving case 4 , i.e. if @xmath13 is closer to 0.1 , the next prominent decay compared with @xmath285 in general is free " quark decay , i.e. one of the bound @xmath2 or @xmath3 quarks decays , dissolving the bound state system . this could be practically the only other decay mode , if the exchange and annihilation decays are suppressed by a small decay constant ( case 2 ) . the signature is @xmath322 , @xmath323 ( plus c.c . ) , where the notation implies the associated @xmath324 or @xmath325 decays on - shell , but the @xmath210 and @xmath326 are decay products of a bound , somewhat off - shell @xmath2 or @xmath3 quark , which is not too different from open @xmath23 production . unless the on - shell nature is used in the direct search , the search limits would not be affected . however , once 4 g quarks are discovered , one should check whether , for some fraction of the events ( depending on resonance vs open @xmath23 production ratio ) , one of the heavy quarks is in fact off - shell , which would be an indication for bound state phenomena . we remark that open @xmath23 production at the lhc is dominantly through @xmath147 fusion , while the @xmath148 resonance production is through @xmath7 fusion . there is little resonance phenomena in @xmath327 via yukawa effects . in fact , in the @xmath328 channels , it is even repulsive . thus , even above the unitarity bound , standard search can continue , except that there may be some anomalies " as we have discussed , if any thing is found at all . we have already dealt with the special case of dijet resonance for case 4 . dijets from @xmath329 tend to be subdominant in all other cases ( i.e. 13 ) , but it could be at 10% level . it would provide a spectacular dijet resonance signal , and definite measurement @xcite of resonance mass , and spin if there is good signal over background . if the branching ratio could be measured in some way , one could access the important decay constant . in general , a resonance would also appear in @xmath259 ( boosted top jets ) , with cross section 1/5 the dijet resonance . finally , there is also the exchange decay to @xmath8 and @xmath9 , which is a subdominant channel typically below 10% . it mimics single @xmath2 ( @xmath3 ) " production , and can be studied that way . but an associated boosted top , or a high @xmath302 @xmath16-jet that tags a resonant @xmath326 , could be quite distinct . we offer some remarks on the @xmath330 @xcite and @xmath331 @xcite anomalies at the tevatron . naively , one might think that the presence of resonance production of @xmath259 could be relevant for @xmath330 . but @xmath148 has same quantum numbers as the gluon , i.e. the coupling to @xmath259 is fully vector . thus , it can not generate @xmath330 suggested by tevatron data . for the @xmath331 anomaly , the yukawa bound resonances are so massive , they can have nothing to do with it . this brings us to comparing with technicolor ( tc ) models . low scale tc has been invoked @xcite for the @xmath331 anomaly suggested by cdf . our @xmath148 and @xmath221 are yukawa bound @xmath23 mesons with an operative heavy isospin , from degenerate chiral quark doublet @xmath47 not too far above the unitarity bound . thus , our @xmath148 and @xmath221 mesons are much heavier than those in low scale tc . for more generic tc models @xcite , since walking " is generally required , the technipion @xmath332 tends to be closer to the technirho @xmath333 in mass such that @xmath334 is absent , while ( near ) degeneracy of @xmath335 and @xmath336 with @xmath333 is also often invoked . the signature for these technimesons are typically @xmath337 , @xmath338 and @xmath339 . thus , not only the spectrum is rather different absence of @xmath31 and @xmath340 mesons the decay signature is also in strong contrast . the bound states due to strong yukawa coupling , which follow simply from the existence of new heavy chiral quarks without assuming new dynamics , should be easily distinguished from technicolor . the yukawa - bound ultraheavy mesons are also quite distinct from qcd bound states . not only there is the absence of @xmath33 ( where @xmath147 fusion would be possible ) and @xmath31 type mesons , they have a much larger binding energy , and are much smaller in size . this is brought about by not only a strong coupling constant , but facilitated by a @xmath40 coupling due to goldstone or longitudinal vector bosons ; the @xmath341 higgs boson , being heavy , would in fact be subdominant . thus , the tight bound states involve ultrarelativistic motion of its very heavy constituents , hence somewhat counterintuitive . by far we have not attempted any actual solution of the bound state problem here . we therefore chose to remain close to the unitarity bound , considering bound state masses not more than 1.5 tev . we have chose to parameterize with a few key parameters . our numerics , and associated phenomenology , should be viewed as only illustrative , with the _ key parameters to be determined by experiment_. what mass scale would nature actually choose , if she so chooses to offer an extra chiral doublet above the existing three generations ? it may be related to the electroweak symmetry breaking through @xmath36 condensation . it could in principle lead to very deeply bound states , with binding energy approaching @xmath42 order or more . we have only scratched the surface of yukawa - bound heavy mesons , the treatment of which would require genuine nonperturbative methods , such as @xcite lattice field theory . paradoxically , it is not impossible that heavier quark masses than considered here could result in lower heavy meson masses . again , experiment might take the lead here . with the experimental limits on sequential chiral 4th generation already at 500 gev , i.e. at the doorsteps of the unitarity bound , we have considered the possibility of new @xmath36 mesons bound by strong yukawa couplings . comparing a relativistic expansion approach ( which indicated nonapplicability ) , versus a relativistic bethe salpeter equation approach , we chose to illustrate what might appear in early data of lhc running , i.e. bound states just above the tev scale , but with relatively complicated decay final states . electroweak precision tests suggest a new ( heavy ) isospin symmetry , such that the leading production would be a color octet , isosinglet vector meson , which we call @xmath148 . it can be produced via @xmath4 fusion , through an unknown decay constant , @xmath149 . for decay , the key parameters besides @xmath149 are the quark mixing element @xmath13 , and the mass differences @xmath262 and @xmath269 where @xmath221 is a heavy color octet pion " and @xmath238 a color singlet we find the leading decay is likely @xmath285 , while the other transition channel into @xmath258 is relatively suppressed . the other leading decay is free " , or constituent , quark decay . illustrating with four cases for large / small decay constant , nominal / suppressed mass differences or @xmath13 , together with the two decay channels of free quark decay and @xmath11 decay of @xmath221 , we considered the possible lhc phenomenology . we find in general the @xmath37 to be narrow compared to its mass . the special case of small @xmath13 leads to two possible spectacular signatures . one is @xmath342 , where a massive back - to - back @xmath11 system is accompanied by a relatively soft @xmath15 . this mode could become suppressed if @xmath262 is close to or less than @xmath196 . then , one could have a dijet resonance close to current lhc limits , and a narrow dijet resonance could appear soon . in the general case , dijet signal is suppressed by a decay branching ratio . other than the above two ( perhaps unlikely ) spectacular signals , the generic leading decay is @xmath343 , @xmath344 , followed by free quark decay of @xmath221 ( @xmath287 is typically subdominant ) . this leads to possible multijet signals with an associated relatively soft @xmath15 or @xmath292 tag , where the multijet system is very massive , and with multiple @xmath16 , @xmath15 and @xmath17 jet substructures . if such massive multijet systems can be studied , one could possibly reconstruct both the @xmath221 and @xmath148 resonances . assuming single channel dominance , one can measure the meson decay constant by the total cross section . if @xmath285 is suppressed by kinematics , however , the likely leading decay would be by the constituent heavy quark decay , which is very similar to standard @xmath23 signal , except one heavy quark decays somewhat off - shell . since in any case the leading @xmath327 fusion does not exhibit resonance phenomena , _ the current 4th generation @xmath18 and @xmath19 search can continue beyond the unitarity bound_. but if 4 g quarks are discovered , then some good fraction of the events would have one quark decaying below threshold , indicating bound state phenomena . one , of course , would have to disentangle also @xmath285 , as already discussed . we have provide some definite signatures for yukawa - bound heavy @xmath23 mesons in the 1 to 1.5 tev range . but our study is only of precursory nature . as lhc energy increases , and with higher luminosity , it could uncover new chiral quarks above the unitarity bound , with new unusual bound states . one could probe into the truly nonperturbative regime , which our results only offer a glimpse of what might happen . there may be a host of new heavy mesons awaiting us beyond the horizon . 0.3 cm * acknowledgement*. we thank k .- f . chen , a. djouadi , k. ishiwata , b. kniel , t. kugo , m.b . wise , b .- young and c .- p . yuan for discussions , and k. hagiwara for encouraging comments . wsh thanks the national science council for academic summit grant nsc 99 - 2745-m-002 - 002-asp , and te and hy are supported under nsc 100 - 2811-m-002 - 061 and nsc 100 - 2119-m-002 - 001 . this has been explored , in the context of the nambu jona - lasinio model , by w.a . bardeen , c.t . hill and m. lindner , phys . d * 41 * , 1647 ( 1990 ) . the authors credit the idea to y. nambu , bootstrap symmetry breaking in electroweak unification " , efi preprint 89 - 08 ( 1989 ) . for a recent brief review on the 4th generation , see b. holdom _ et al . _ , pmc phys . a * 3 * , 4 ( 2009 ) . b.w . lee , c. quigg and h.b . thacker , phys . rev . lett . * 38 * , 883 ( 1977 ) ; phys . d * 16 * , 1519 ( 1977 ) . kribs , t. plehn , m. spannowsky and t.m.p . tait , phys . d * 76 * , 075016 ( 2007 ) ; h .- j . he , n. polonsky and s .- f . su , phys . d * 64 * , 053004 ( 2001 ) ; v.a . novikov , l.b . okun , a.n . rozanov and m.i . vysotsky , jetp lett . * 76 * , 127 ( 2002 ) [ pisma zh . fiz . * 76 * , 158 ( 2002 ) ] . j. pumplin _ et al . _ , jhep * 0207 * , 012 ( 2002 ) . r. baier and r. rckl , z. phys . c * 19 * , 251 ( 1983 ) . et al . _ , d * 35 * , 3366 ( 1987 ) [ erratum - ibid . d * 38 * , 1632 ( 1988 ) ] . p. nason , s. dawson and r.k . ellis , nucl . b * 303 * , 607 ( 1988 ) . a. petrelli _ et al . _ , nucl . b * 514 * , 245 ( 1998 ) . y. sumino _ et al . _ , phys . d * 47 * , 56 ( 1993 ) . m. jezabek , j.h . khn and t. teubner , z. phys . c * 56 * , 653 ( 1992 ) . s. chatrchyan _ et al . _ [ cms collaboration ] , phys . b * 701 * , 204 ( 2011 ) . this effect is examined for the top quark case in the @xmath345 resonance region , in y. sumino and h. yokoya , jhep * 1009 * , 034 ( 2010 ) . t. aaltonen _ et al . _ [ cdf collaboration ] , phys . lett . * 106 * , 171801 ( 2011 ) . note that this has been refuted by the d0 collaboration , v.m . abazov _ et al . _ [ d0 collaboration ] , _ ibid . _ * 107 * , 011804 ( 2011 ) .

in this paper we revisit an old and all - important problem of turbulent modeling . the problem , first formulated in terms of images borrowed from kinetic theory by prandtl ( 1925 ) , was later developed by kolmogorov ( 1942 ) , launder and spaulding ( 1974 ) and many of their followers . the success of these semi - empirical approaches can hardly be overestimated : with development of powerful computers turbulent modeling became an important part of scientific and engineering design process . still , these models , involving low -order time - derivatives and low powers of dimensionless rate of strain , are less effective in describing strongly sheared and rapidly distorted flows where the dimensionless rate of strain is not too small . to improve performance , various non - linear models , pioneered by speziale ( 1987 ) , accounting for the next order in powers of the rate of strain , have been proposed . since the second - order trancation of a lacking - small -parameter expansion is problematic , these models had a mixed success . ( the role of the large - eddy simulations ( les ) in description of complex flows will be discussed in conclusions . ) the attempts to systematically derive turbulence transport models from the navier - stokes equations ( ns ) , based on renormalized wyld s perturbation expansions ( rng ( yakhot/ smith et al ( 1992 ) , rubinstein and barton ( 1990 ) ) , double expansion ( yakhot et al ( 1992 ) ) , dia ( yoshizawa ( 1987 ) ) , were relatively successful and the resulting equations found their place in engineering . it became clear soon that resummation of the series was impractical and the approach was restricted to the low- orders . it has been shown recently by h. chen et al ( 2004 ) that the turbulence models , derived by application of the chapman - enskog expansion ( ce ) to a _ model _ kinetic ( boltzmann - bgk ) equation written for hydrodynamic modes , are similar to those previously obtained from the wyld ( 1961 ) expansion , applied directly to the navier - stokes equations . the main result of chen et al ( 2004 ) can be stated as follows : the perturbation theory , leading from kinetic equation to the navier - stokes ( ns ) equations , generates the well - known turbulence models if , instead of the relaxation time @xmath4 of kinetic theory , one writes @xmath5 and represents the temperature @xmath6 as @xmath7 . here @xmath8 and @xmath9 are the mean turbulent kinetic energy and dissipation rates , respectively . thus , the formal superficial similarity between the two perturbation expansions has been established . this result means that , in principle , the models of an arbitrary non - linearity and complexity , which we even can not explicitly write down , are contained in a simple kinetic equation . in this work , we would like to reformulate the procedure developed in chen ( 2004 ) by taking into account both microscopic and hydrodynamic modes . the coarse - grained kinetic equation for hydrodynamic modes is then obtained by integrating out the fast microscopic variables . to achieve this goal , we had to substantially redefine the expansion procedure and , as a result , obtain a qualitatively new class of turbulence models . kinetic theory for a low - density gas is based on the boltzmann equation @xmath10 for a single particle distribution function @xmath11 , where collision integral @xmath12 if the system hamiltonin is known , this equation can be derived directly from the liuville theorem . here the element of the phase -space volume @xmath13 and @xmath14 is the relative velocity of colliding particles . if the intermolecular distance is @xmath15 where @xmath16 is the length - scale of intermolecular interaction , the strait particle trajectories between collisions are assumed in deriving expression ( 1.2 ) . the probability density functions in ( 1.2 ) are @xmath17 and @xmath18 . since in accord with the theory of elastic collisions , the velocities @xmath19 are expressed in terms of @xmath20 , the integration in ( 1.2 ) is carried out over the phase space @xmath21 only . the relaxation -time approximation ( rta ) : @xmath22 where @xmath23 is the distribution function in thermodynamic equilibrium and @xmath24 is a properly chosen relaxation time , is an often used anztaz , mentioned in landau and lifshitz ( 1981 ) as a `` rough estimate of the ( boltzmann ) collision integral '' ( 1.2 ) . still , recent implementations of this approximation in the so called `` lattice boltzmann '' numerical codes , led to a remarkable success in simulating a wide variety of extremely complex fluid flows . as of today , the approach has been tested on basically all examples of classic laminar flows ( benzi ( 1992 ) , chen ( 1998 ) , succi ( 2001 ) ) . the robustness , speed and simplicity of the method , made it an attractive tool for both theoretical investigations and engineering design . even more spectacular , and somewhat unexpected , is success enjoyed by the lattice boltzmann method in simulating complex turbulent flows ( h. chen ( 2003 ) ) . in this application , the relaxation time @xmath24 is expressed in terms of hydrodynamic observables : @xmath25 , where turbulent kinetic energy is @xmath26 and the mean dissipation rate is @xmath27 . in this case , in the first order of the chapman - enskog ( ce ) expansion one obtains the well - known @xmath3 model with the non - linear @xmath3 model appearing in the next order ( chen et al ( 2004 ) ) . in addition to speed and simplicity , the most attractive feature of this type of turbulence modeling is that it _ does not require theory of pressure - velocity correlations _ which , in complex flows , is an extremely difficult , still unsolved , problem . typically , derivation of hydrodynamic navier - stokes equations from kinetic theory is based on a few assumptions . 1 . if spatial gradients of mean velocity , density and temperature are equal to zero , the gas is assumed in thermodynamic equilibrium ; 2 . expansion in powers of dimensionless rate of strain @xmath28 where @xmath29 leads to transport equations describing various physical phenomena . the equations , derived this way can be used to fully describe large reynolds number turbulent flows requiring at least @xmath30 number of degrees of freedom . in the flows of practical importance this number is huge and therefore , the value of a coarse graining procedure leading to turbulence models can hardly be overestimated . it has been shown ( yakhot et al ( 1992 ) ) that the systematic derivation of turbulence models directly from the navier - stokes equations , can be formulated in terms of wyld s ( 1961 ) diagramamtic expansion in _ two _ dimensionless parameters : the reynolds number @xmath31 and , familiar from kinetic theory , dimensionless rate of strain @xmath32 . the first expansion is responsible for description of isotropic and homogeneous turbulence with the non - zero energy flux across the scales and the second - for the non - zero _ spatial _ fluxes and flow structures . due to proliferation of tensorial indices in the high - order contributions , the resummation of the expansion is an extremely difficult task and it has been conjectured by polyakov ( 2001 ) in the middle of seventies that it may result in a kinetic equation containing all terms of wyld s series . if this is true , a turbulence model of an arbitrary non - linearity and complexity may be contained in a relatively simple kinetic equation . _ the most important and interesting feature of this system is that , unlike equilibrium , the @xmath33 state of isotropic and homogeneous turbulence is not flux- free but involves a large @xmath34 energy flux across the scales . thus , to derive turbulence models , it is desirable to develop a kinetic approach not based on the equilibrium , flux - free , zero - order state . _ in this paper we , building upon a remarkable work of h. chen et al ( 2004 ) , consider both turbulent ( hydrodynamic ) and thermal ( microscopic ) velocity fluctuations as governed by the boltzmann -bgk kinetic equation . since microscopic and hydrodynamic fluctuations occupy their respective fractions of the phase - space , this equation is formally defined on a somewhat enlarged phase - space . the coarse - graining , eliminates the small - scale fast microscopic fluctuations and restores the six - dimensional space , leading to the kinetic equation for the _ large - scale ( turbulent ) velocity component only_. in this case , the zero - order , zero - mean - spatial - gradient state is not an equilibrium but that of isotropic and homogeneous turbulence characterized by a finite energy flux across the scales . the chapman - enskog expansion applied to this state leads to a novel set of turbulence models of arbitrary nonlinearity . this paper is organized as follows . in the next section we , for the sake of clarity and continuity , describe traditional derivation of hydrodynamic approximations from kinetic equation and briefly outline the way turbulence models appear from the ns equations . in section 3 , which is most important for the present development , we introduce the non - equilibrium kinetic equation and the zero - order pdf for isotropic and homogeneous turbulence , both defined on an enlarged phase - space . in section 4 , the finite energy flux is defined as a dynamic constraint and it is shown that the introduced in section 3 non - equilibrium pdf is indeed a solution to non - equilibrium equation which conserves the total kinetic energy of a system . finally , in section 5 , the coarse - grained kinetic equation for hydrodynamic modes is derived . it is shown in sections 6 and 7 , that the ce expansion , leads to transport equations for both turbulent velocity field and kinetic energy . in section 8 , choosing a proper relaxation time , we show that our kinetic equation contains the well known k - e , reynolds stress and non - linear models . in addition , this simple equation contains turbulent models of an arbitrary non - linearity and complexity . some new effects , originating from interaction of spatial fluxes with the energy flux ( cascade ) are identified . summary and conclusions are presented in section 9 . in equilibrium , with all temporal and spatial gradients equal to zero , the left side of kinetic equation ( 1.1),(1.2 ) is equal to zero and the remaining equation @xmath35 has a solution : @xmath36 where the energy @xmath37 is conserved in the process of collision , i.e. @xmath38 and the temperature @xmath39 . for a gas moving with the constant velocity @xmath40 , the solution is : @xmath41 this solution is clear if we introduce internal energy @xmath42 and write @xmath43 in a frame of reference moving with velocity @xmath40 , we have @xmath44 in deriving transport approximation , one expands the probability density powers the large - scale field ( ll ) @xmath45 . the constant component @xmath46 can be removed by the galileo transformation and only derivatives of the large scale field @xmath40 are relevant . in this section we , for the sake of clarity and continuity , briefly describe the main steps leading from the boltzmann equation to hydrodynamic approximations . classic derivation of transport approximations starts with the boltzmann equation ( 1.1 ) , ( 1.2 ) or ( 1.3 ) and equilibrium pdf @xmath47 in this section we denote velocity of the frame of reference @xmath48 which is a physical ( not phase - space variable ) . in what follows we restore the notation @xmath49 used above . this should not lead to any difficulty . the equilibrium pdf is obtained from equation @xmath50 , which is valid locally , so that the hydrodynamic field @xmath51 can be a slow varying function of time and space coordinates . in addition , the conservation laws and the symmetries of collision integral give : @xmath52 introducing the fluid density @xmath53 and integrating ( 1.1)-(1.2 ) over @xmath54 leads to the continuity equation : @xmath55 multiplying ( 1.1)-(1.2 ) by @xmath54 with subsequent integration over the phase space yields : @xmath56 where the stress tensor is : @xmath57 . in thermodymanic equilibrium , the stress - tensor is evaluated readily with the distribution function @xmath23 from ( 2.1 ) giving @xmath58 , so that combining this with the ideal gas equation , gives : @xmath59 in what follows @xmath60 . defining @xmath61 , we obtain : @xmath62 on the equilibrium pdf @xmath23 , @xmath63 and the third - order in fluctuations contribution to ( 2.6 ) is equal to zero . the nature of the large - scale hydrodynamic field @xmath64 in ( 2.4 ) has not been specified . the problem of turbulence modeling is formulated as follows : let the velocity field @xmath65 where @xmath66 and @xmath67 . our goal is to derive the equation for the large - scale slow field @xmath40 . the above decomposition gives : @xmath68 or @xmath69 where @xmath70 now we need a theory or at least set of rules to express both stress @xmath71 and the reynolds stress @xmath72 in terms of the large - scale field @xmath40 and its derivatives . this program has been systematically developed in yakhot et al ( 1992 ) , smith et al ( 1992 ) . the method consists of a few steps : 1 . in the first order , the stress tensor @xmath73 this leads to the navier - stokes equations ( ns ) . 2 . add a statistically isotropic and homogeneous stirring force to the right side of the ns . this force generates an @xmath34 energy flux @xmath74 across scales . in the case of a free turbulence with the total kinetic energy @xmath8 , prepared as an initial condition , the energy balance reads @xmath75 . 3 . use wyld s ( 1961 ) expansion to eliminate small - scale velocity fluctuations . it has been shown ( yakhot et al ( 1992 ) ) that development of turbulent models involves a double expansion in powers of two dimensionless parameters : the reynolds number @xmath76 and rate of strain @xmath77 where @xmath78 is the relaxation time of turbulent fluctuation . the expansion in re leads to the fixed point @xmath79 , corresponding to the zero- spatial -gradient state of isotropic and homogeneous turbulence which serves a a ground state for the additional expansion in powers @xmath80 . the first expansion accounts for the non - zero energy flux of isotropic and homogeneous turbulence while the second - gives rise to the not zero mean velocity field and various fluxes . in the first order of the @xmath80-expansion one derives a variant of @xmath3 model , in the second -the reynolds stress model and non - linear models etc . it is important that the second order @xmath81 terms of the non - linear turbulent models , are well known from kinetic theory as next - order in the chapman - enskog expansion corrections to the navier - stokes equations ( landau - lifshitz ( 1981 ) ) . this fact establishes similarities between wyld s expansion of the ns equations and the ce expansion of kinetic theory . while , one can formally write each term of the @xmath80-expansion , resummation of an infinite series is totally out of reach . in turbulent flows , the velocity field is characterized by three correlation lengths ( landau and lifshitz ( 1959 ) , frisch ( 1996 ) ) : 1 . microscopic mean -free path @xmath82 , where @xmath83 is the speed of sound and @xmath24 is the microscopic relaxation time ; 2 . hydrodynamic inertial interval @xmath84 , where both large - scale forcing and viscous effects can be neglected and 3 . nonuniversal , geometry - dependent , energy - production range @xmath85 . we denote three velocity fields corresponding to these intervals as @xmath86 , respectively . ( see fig.1 ) . locally , at the scales @xmath87 , the system is in thermodynamic equilibrium so that the @xmath88-field obeys maxwell ( gibbs ) statistics . hydrodynamic approximations follow elimination ( small - scale averaging ) of modes from the interval @xmath89 , where @xmath24 is the relaxation time characterizing dynamics of small perturbations from thermodynamic equilibrium . after that , the turbulence models are derived by averaging the navier - stokes equations in the interval @xmath90 . we do not doubt the ability of the navier - stokes equations fully describe turbulent flow of arbitrarily large reynolds number . in this case , since the number of excited modes is at least @xmath91 , the computational cost of these direct simulations ( dns ) is very high . our concern here is with derivation of the coarse - grained equations ( turbulence models ) for partial representation of turbulence in the interval @xmath92 , which is hard to obtain from hydrodynamics . in other words , _ in this work we are interested in the mode - elimination from the interval @xmath93 applied directly to the boltzmann equation . if successful , this way we will obtain the coarse - grained kinetic equation , containing , as a hydrodynamic approximation , all possible turbulence models , which are very hard to derive from the navier - stokes equation . _ in particular , we are interested in equations of motion for the coarse - grained or filtered velocity field which includes only a small fraction of hydrodynamic modes corresponding to the top of inertial range , i.e. @xmath92 . whatever the filtering procedure , the coarse - grained fields are obtained from the full fields by averaging over flow volumes of linear dimension @xmath94 , where @xmath95 is the filtering scale or over the modes @xmath96 with @xmath97 in a fourier space . also , one can define a coarse - grained field as : @xmath98 where the integration is carried out over the volume @xmath99 centered about the point @xmath100 or @xmath101 where @xmath102 is a filter function . we see that the coarse - grained velocity field is a multipoint construction and evaluation of high - order moments is not a simple task . moreover , spatial velocity derivatives are defined on a cut - off as : @xmath103 and , if @xmath104 , their moments , called structure functions , obey multifractal statistics . the inertial range multifractality ( anomalous scaling ) means that the moments of different orders are independent of each other . as a result , the full statistical description of velocity field involves a large , infinite in the limit @xmath105 , number of independent parameters . this makes an accurate coarse - graining with the filter ( mesh ) size @xmath106 basically impossible . the situation changes dramatically in the limit @xmath107 . it is a well - established , though not well - understood , fact that at the scales @xmath108 , both the turbulent velocity fields @xmath2 and @xmath109 obey close- to- gaussian statistics . moreover , it was demonstrated by schumacher et al ( 2007 ) that in the low - reynolds number turbulence , the velocity derivatives , too , are gaussian random variables . if this is so , the entire large - scale field can be accurately described in terms of the second - order moment only . our goal is to derive a coarse - grained equation for the modes @xmath110 fluctuating on the scales @xmath111 . in the standard turbulence modeling , this is operationally achieved by introducing a large `` effective viscosity '' @xmath112 , where @xmath113 , into the navier - stokes equations . due to a particular choice of viscosity @xmath114 , the small - scale modes ( @xmath115 ) are _ overdamped _ and the resulting hydrodynamic equations describe large - scale properties ( @xmath92 ) of the flow only . the defined this way equation is called `` transport model '' as as opposed to the les models based on an inertial range cut -off @xmath116 . since the dominant contribution to turbulent kinetic energy is contained in the large -scale fluctuations , we write the zero - order nonequillibrium probability density function in the laboratory reference frame : @xmath117 in this expression , both @xmath54 and @xmath64- fields are phase - space variables independent of time and space . thus , we have temporarily enlarged the phase space from six to nine dimensions . the flow described by the probability density function ( 3.1 ) can be perceived as an ensemble of laminar flows each in thermodynamic equilibrium with the _ fluctuating equilibrium maxwell - gibbs probability density . _ it is clear that : @xmath118 and @xmath119 and @xmath120 . in the laminar limit @xmath121 @xmath122 and integrating over @xmath64 , we recover the expression ( 2.1 ) . the pdf ( 3.1 ) , resembling that of a mixture of gases with different temperatures @xmath6 and @xmath123 , describes an essentially non - equilibrium state . _ it will be shown below that ( 3.1 ) is a solution to a `` nonequilibrium '' kinetic equation and this temperature difference is related to the energy flux from hydrodynamic modes to the microscopic ones . _ from the above definitions : @xmath124 in what follows , we assume the probability density @xmath125 so that locally @xmath126 and @xmath127 the pdf of hydrodynamic velocity field @xmath128 is obtained from ( 3.3 ) by integrating out the microscopic variable @xmath54 , i.e. @xmath129 _ our goal is to derive a kinetic equation for @xmath128 describing turbulent fluctuations only . to achieve this goal , we first increase the original six - dimensional phase space of the previous section ( @xmath130 ) to the nine - dimensional one ( @xmath131 ) and , eliminating thermodynamic fluctuations , project the nine - dimensional problem to the six - dimensional one @xmath132 . the advantage of this procedure from the turbulence modeling viewpoint is that while original problem involved @xmath133 number of particles , the coarse - grained one - only @xmath134 , where the integral scale @xmath135 is the large scale correlation length of turbulence . _ since the fast @xmath136 -field varies on the time scale @xmath137 , where @xmath83 stands for the speed of sound , @xmath138 , we can consider these components statistically ortogonal and introduce the continuity equation in the enlarged 9d - phase- space . in this case the botzmann equation ( landau / lifshitz ( 1981 ) ) : @xmath139 where @xmath140 is not yet specified ` non - equilibrium collision integral ' with @xmath141 obtained from equation : @xmath142 with @xmath143 ( @xmath144 . _ thus , in the theory developed below , we abandon the equilibrium pdf ( 2.1 ) for the the zero - spatial - gradient state of a gas in favor of the nonequilibrium expression ( 3.1 ) . _ the detailed discussion of this point will be given below . here we just mention that due to the non - zero energy flux across the scales , the kinetic energy of the zero - spatial - gradient ( @xmath33 ) turbulent flow decays leading to increase of the gas temperature @xmath6 . if a turbulent flow is supported by a large - scale energy input per second @xmath145 , by nonlinearity , this energy is transfered to the smallest scales where it is dissipated . it is usually assumed that the nature of the dissipation mechanism is unimportant . on the inertial- range scales , where both energy pumping and dissipation are negligibly small , the dynamics are characterized by the mean energy flux @xmath146 , where @xmath9 is the mean dissipation rate . in case of isotropic and homogeneous turbulence , prepared at initial instant of time with kinetic energy @xmath147 , the energy @xmath148 decays to zero . since in this case , all spatial gradients are equal to zero , no _ mean spatial fluxes _ are involved in the relaxation process . _ unlike equilibrium gas , this strongly non - equilibrium system is characterized by a non - zero constant energy flux _ across the scales _ _ . it has been shown in a series of extraordinary works by zakharov et al ( 1975 ) , ( 1992 ) , ( 1984 ) that , in case of turbulence , the euler equations , with velocity field written in terms of clebsch variables , can be represented as a kinetic equation for waves ( particles ) with the collision integral @xmath149 describing the non - linear interactions . it has been shown that , in addition to equilibrium solutions , the equation @xmath150 for the occupation numbers @xmath151 , playing the role of the pdf in the @xmath152-space , has solutions corresponding to the non - zero constant fluxes in the wave - number space . the direction of these fluxes is exclusively determined by physics of the problem and peculiarities of non - linear interactions . for example , the mean field approximation developed in yakhot ( 1992 ) , led to the kolomogorov energy spectrum as one of zakharov s constant - flux solutions . thus , unlike thermodynamic equilibrium , the kinetic equation for a zero -mean - gradient turbulent flow must satisfy the flux ( energy balance ) constraints : @xmath153 valid in the statistically steady state or @xmath154 - in case of decaying turbulence . the relations ( 4.2 ) and ( 4.3 ) , reflecting the energy balance in a flow , are independent upon the nature of the dissipation mechanism . we split collision integral @xmath140 into two components : @xmath155 responsible for the non - zero energy flux of isotropic and homogeneous turbulence and @xmath156 - for the spatial fluxes of a general turbulent flow . since , as was pointed out above , the mean energy flux is a large - scale property , no differential operator can enter the expression . thus we have : @xmath157 and @xmath158\frac{{\cal e}}{{\cal k}}f({\bf c , v})\ ] ] now we will show that the zero - order pdf ( 3.1 ) is a solution to kinetic equation describing isotropic and homogeneous turbulence . since in this case , all spatial derivatives are equal to zero , the kinetic equation is : @xmath159 and @xmath160 , @xmath161 and @xmath162 . substituting ( 3.1 ) into ( 3.5)-(3.6 ) gives : @xmath163 this equation , valid for all values of independent parameters @xmath164 and @xmath165 , can be correct only if : @xmath166 @xmath167 _ thus , we conclude that the zero - order pdf ( 3.1 ) is indeed a solution to the kinetic equation ( 3.5 ) , ( 3.6 ) for isotropic and homogeneous turbulence , subject to the energy constraints ( 4.1),(4.2 ) . one can also see that according to ( 4.6 ) , ( 4.7 ) : @xmath168 , which reflects conservation of the total kinetic energy in decaying isotropic and homogeneous turbulence . _ introducing conditional probability density @xmath169 we can write : @xmath170 so that @xmath171 @xmath172 and @xmath173 below we develop an approach resembling the one proposed for the probability densities of the passive scalar in a random velocity field in sinai , yakhot ( 1989 ) . integrating the nine - dimensional boltzmann equation ( 3.5 ) , ( 4.3 ) over the relatively fast- field @xmath54 gives : @xmath174 where @xmath175 and the collision integral @xmath176 is formally written with the yet unspecified relaxation time @xmath78 . the coarse -grained zero - order pdf is : @xmath177 the expression for conditional expectation value @xmath178 is a subject to four dynamic constraints . 1 . it must be invariant under galileo transformation and not violate the galileo invariance of equation ( 5.5 ) . this is the reason for the @xmath179 term in the right side of ( 5.3 ) . 2 . integrating ( 5.5 ) over @xmath64 must lead to the continuity equation . this gives : @xmath180 3 . multiplying ( 5.5 ) by @xmath64 and comparing the result with ( 2.6 ) yields : @xmath181 4 . in the zeroth order of expansion in powers of dimensionless spatial gradients ( chapman - enskog expansion ) , the equation ( 5.5 ) must generate ( 2.5 ) with the pressure term containing information about the integrated out microscopic variables . in a simplest case of ideal gas , @xmath182 . a relation satisfying the above constraints is : @xmath183 where @xmath184 thus , the course - grained kinetic equation , correctly describing isotropic and homogeneous turbulence ( zero spatial gradients ) and satisfying the dynamic constraints presented above , is : @xmath185 the equation ( 5.11 ) is the main result of this paper . integrating ( 5.11 ) over @xmath64 gives continuity equation ( 2.3 ) . in addition , multiplying it by @xmath64 and integrating leads to : @xmath186+\frac{\overline{(v_{i}-v_{i})(v_{j}-v_{j})}}{{\cal k}}\frac{\partial p'}{\partial x_{i } } + \frac { \overline{\rho({\bf v - v})({\bf v - v})^{2}}}{{\cal k}}\frac{{\cal e}}{d{\cal k}}=0\ ] ] or @xmath187+\frac{\overline{(v_{i}-v_{i})(v_{j}-v_{j})}}{\rho{\cal k}}\frac{\partial p'}{\partial x_{i}}+ \frac{\overline{\rho({\bf v - v})({\bf v - v})^{2}}}{\rho { \cal k}}\frac{{\cal e}}{d{\cal k}}=0\ ] ] in the zeroth order , averaged with with the pdf ( 5.7 ) , the equation ( 5.14 ) gives the euler equation : @xmath188 where @xmath189 . it follows from ( 5.7 ) and ( 5.11 ) that in the case of strongly nonequilibrium flow as an unperturbed state , the equation of state @xmath190 in ( 5.11 ) is decoupled from the `` temperature '' @xmath123 entering the problem as a parameter in the zero -order pdf ( 5.7 ) . indeed , the coarse grained equation ( 5.11 ) gives for @xmath191 : @xmath192 in the zeroth order ( @xmath33 ) , all gradients equal to zero , the equation ( 5.15 ) reads : @xmath154 we also define a vector @xmath193 to find the solution to the kinetic equation ( 5.11 ) we following chen et al ( 2004 ) write : @xmath194 and @xmath195 and @xmath196 . the mean of any flow property is then @xmath197 where @xmath198 the zero - order pdf @xmath199 is given by expression ( 5.7 ) . the first and second -order corrections are calculated readily with @xmath200 . the result is : @xmath201\equiv [ \frac{\partial } { \partial t}+{\bf v}\cdot \nabla + \frac{{\bf v - v}}{{\cal k}}\cdot(\nabla p')+\frac{({\bf v - v})^{2}-d{\cal k}}{{\cal k}}\frac{{\cal e}}{d{\cal k}}]f^{(0)}=-\frac{f^{(1)}}{\tau_{t}}\ ] ] and @xmath202f^{(1)}-\frac{\partial f^{(0)}}{\partial t_{1}}=-\frac{f^{(2)}}{\tau_{t}}\end{aligned}\ ] ] this gives : and using the pdf given by ( 6.8 ) , we can write : @xmath205 and the normal stress ( pressure ) @xmath206 . setting @xmath207 gives ; @xmath208 and : @xmath209 or introducing turbulence kinetic energy @xmath210 : @xmath211 where @xmath189 . two remarks are in order . firsrt , an additional dimensionless expansion parameter , @xmath212 , which is not present in the ordinary chapman - enskog expansion , appears in the present nonequilibrium theory . in strong turbulence , where molecular relaxation can be neglected , the characteristic time is @xmath213 and @xmath214 , is nothing but the renormalized reynolds number @xmath215 , corresponding to the fixed point of isotropic and homogeneous turbulence . second , the derived pressure @xmath216 leads to renormalization of the sound speed by the turbulent velocity fluctuations , derived in staroselsky et al ( 1990 ) and predicted earlier by candrasekhar ( 1995 ) in the context of gravitational collapse of interstellar gas . combining the above results we , writing for the sake of simplicity of notation @xmath217 , have : @xmath218 in what follows the last term in ( 6.15 ) , responsible for the so called dilatation effects in compressible flows , will be neglected . multiplying this equation by @xmath219 we summarize the above results : _ in the first order of the chapman -enskog expansion , the coarse grained kinetic equation ( 5.11 ) gives _ : @xmath220 and @xmath221 according to ( 6.6 ) @xmath222 where the operator @xmath223 is defined in ( 6.5 ) . the symbol @xmath224 stands for the time - derivative in first - order contributions to the hydrodynamic equations . for example : @xmath225 @xmath226 and @xmath227 it is clear that @xmath228 is a time derivative of all @xmath229 contributions to expansion and , according to ( 7.1 ) , @xmath230 . thus , now , assuming @xmath232 , @xmath39 and @xmath233 , we rewrite ( 6.8 ) as @xmath234 where @xmath235\ ] ] and @xmath236 thus : @xmath237^{2 } + \hspace{1 in } \nonumber \\ 2{\cal k}\tau_{t}[-2[(v - v)_{\alpha}(v - v)_{\beta})- ( v - v)_{\beta}\nabla_{\alpha}p- ( v - v)_{\alpha}\nabla_{\beta } p]s_{\alpha,\beta})+ \frac{1}{\tau_{t}}(v - v)_{\alpha}(v - v)_{\beta}({\cal d}\tau_{t}s_{\alpha\beta } ) ] \}\nonumber \\ -f^{(0)}\frac{\tau_{t}g}{2{\cal k}^{2 } } ( { \bf v - v)\cdot \nabla}{\cal k}\hspace{3in}\end{aligned}\ ] ] where @xmath238 . collecting all contributions up to @xmath239 simple calculation with @xmath240 gives : @xmath241 so that for @xmath242 : @xmath243 and finally : @xmath244 + o(\nabla { \cal k})\ ] ] the remaining @xmath245 contributions can be easily calculated . for example : @xmath246 we see that @xmath247 the usually neglected @xmath248 contributions to the turbulent models , may not be too small in the high - gradient regions of the flow . introducing effective viscosity @xmath249 with @xmath250 gives : @xmath251 + o(\nabla { \cal k})\ ] ] we can summarize the results of this section : in the second order of expansion if @xmath207 and @xmath252 , the turbulence model reads : @xmath253 and @xmath254 in the high reynolds number regions of the flow , @xmath255 and @xmath256 the expressions ( 8.4 ) and ( 8.3 ) have a few features not present in familiar turbulence models . first , consider a simplest possible example of the fully developed statistically steady flow in an infinite channel . in this case , @xmath257 , @xmath258 and the equation ( 8.3 ) reads : @xmath259 the last contribution to the right side of equation ( 8.6 ) describing interaction of the energy flux in the wave - number space and momentum flux in the physical space , is not small in the proximity to the walls . this effect does not appear in familiar turbulence models . the equations ( 8.1 ) - ( 8.5 ) are quite involved and , as in the theories based on renormalized wyld s expansions of the navier - stokes equations , accounting for the higher - order contributions to the hydrodynamic approximation is impossible . _ however , the main difference between the approach developed in this paper and the theories based on the renormalized perturbation expansions of the navier - stokes equations is that here the problem of resummaion of an infinite series does arise : the equations ( 8.1)-(8.5 ) and all high - order models , we even can not write down , are contained in a simple kinetic equation ( 5.11 ) . _ the boltzmann - equation ( 1.1)-(1.2 ) contains the navier - stokes equations capable of accurate description of turbulent flows . as was demonstrated in the nineties ( yakhot et al 1992 ) , rubinstein ( 1990 ) , ) the coarse -graining or small - scale elimination , applied to the ns equations , leads to the low - order turbulent models . we believe that , in principle , _ the coarse - grained boltzmann equation ( 1.1)-(1.2 ) _ contains coarse - grained hydrodynamic approximations and _ all _ possible turbulence models ( both les and transport ) of an arbitrary nonlinearity and complexity . at the present time , this coarse - graining procedure does not exist . one can attempt to develop the wyld diagrammatic expansion applied directly on the non - linear equation ( 1.1)-(1.2 ) , eliminate fast modes and , as a result , derive turbulence models . this well - defined program is extremely complex and we were unable to achieve much progress . instead , the theory proposed in this paper is based on a simple qualitative relaxation time approximation ( 1.1),(1.3 ) combined with the ` turbulent relaxation time ' @xmath213 which is an exact consequence of the energy balance ( 4.6 ) . since kinetic energy is the large - scale property of turbulence , this relaxation time is the largest time - scale in a flow leading to the largest ` turbulent viscosity ' @xmath260 . this way , the fast modes are overdamped and the resulting equations correspond to transport turbulent modeling describing velocity fluctuations on the scales @xmath92 . transport modeling vs les . _ the choice of the often - used maxwell - gibbs distribution function ( 2.1 ) is an important step in derivation of the navier - stokes equations from kinetic theory . given these equations , one defines a spatial cut- off @xmath95 and , averaging over the small - scale ( @xmath261 ) fluctuations , derives the coarse - grained ns equations , called turbulent models ( yakhot ( 1992 ) , smith et al ( 1992 ) , rubistein et al ( 1990 ) , yoshizawa ( 1987 ) ) . by construction , the resulting equations for the `` resolved scales '' involve spatial derivatives evaluated on the mesh ( cut -off ) , i.e. all gradients in these equations are : as was shown above , the high - order modeling involves high powers of velocity gradients ( velocity differences ) , which , due to intermittency ( @xmath104 ) , can not be expressed in terms of small deviations from the gaussian pdf . therefore , to describe a flow , including the inertial range ( @xmath263 ) dynamics , the model must satisfy an infinite number of non - trivial dynamic constraints involving , among other terms , pressure - velocity correlations ( yakhot and sreenivasan ( 2006 ) ) . with increase of the difference @xmath264 , more and more constraints become relevant and , at the scales @xmath115 , the quality of all existing les models rapidly deteriorate : with increase of the moment order , the difference between full ( dns ) and model ( les ) simulations of high - order moments grows . as @xmath107 , the probability density of velocity difference is extremely close to gaussian , the moments @xmath265 , @xmath266 and the problem of anomalous scaling does not appear . ( see fig . ) . this huge simplification was the main and only reason for our choice of the non - equilibrium probability density ( 3.1 ) leading to the gaussian pdf ( 5.7 ) of the coarse - grained hydrodynamic field . thus , the procedure developed here is valid for transport modeling only . to work out a similar approach to the les ( @xmath104 ) , one has to use the expressions for the multifractal pdfs of velocity difference , which is not a simple task . + _ 2 . the models . _ at the large scales @xmath92 , the zero - order ( @xmath33 ) flow can be described in terms of only _ two parameters _ : kinetic energy @xmath8 and mean dissipation rate @xmath9 , so that @xmath5 . despite superficial similarity of relaxation times , the coarse - grained kinetic equation , derived in this paper , is very different from the well -known @xmath267 model , widely used in engineering : + 1 . no explicitly written momentum equation is involved in our model . no separate equation for kinetic energy @xmath8 is needed : all information is contained in the kinetic equation ; + 3 . the simple equation ( 5.11 ) contains @xmath267 , reynolds stress and all non - linear models etc . thus , for example , it can describe the secondary vortices in a square duct flow , which no @xmath267 model is capable of doing . no modeling of pressure - velocity correlations is needed : it is hidden in ( 5.11 ) . this fact can be of extreme importance for computing complex flows . the rapid distortion and memory effects are accounted for in ( 5.11 ) . _ sound speed renormalization . _ the kinetic equation ( 5.11 ) contains an equation of state @xmath268 , which is a remnant of the integrated out microscopic , close - to- equilibrium , modes . the equilibrium speed of sound is defined as usulal : @xmath269 . as follows from ( 6.12 ) , the speed of sound obtained from hydrodynamic approximation , derived in the lowest order of the non - equilibrium ce expansion , proposed in this work , is renormalized : @xmath270 . this effect , appearing in the higher order of the renormalization group procedure applied to the ns equation , was derived and numerically tested in staroselsky et al ( 1990 ) and , even earlier , was suggested by chandrasekhar ( 1955 ) who was interested in the role of turbulence in dynamics of the interstellar gas collapse . this feature may prove important for simulations of turbulent flows . during last two hundred years , the navier - stokes equations enjoyed remarkable success in describing low- knudsen and low- weisenberg number fluid flows . this success was habitually attributed to the large scale- separation between hydrodynamic and microscopic modes : @xmath271 . there is one caveat in this argument , though . one must bear in mind that the ns equations are _ closed _ by imposing proper equations of state relating pressure to density and temperature . invariably , the equilibrium equations of state or equilibrium thermodynamic relations are chosen for this purpose . the problem is that the ce procedure , leading to the ns equations , is an expansion around the zero - spatial- gradient equilibrium state of fluid and both the pressure gradient and the gradient of turbulent kinetic energy are large -scale properties defined on the same hydrodynamic scales . thus , in this state of fluid , the only relevant dimensionless parameter is mach number @xmath272 . therefore , to describe the stresses , in addition to thermodynamic considerations based on an estimate @xmath273 , one has to add the turbulent contribution @xmath274 , which appears naturally in the procedure developed above . thus , strictly speaking , the pressure term in the ns equation , which is the functional of the solution , must be determined self - consistently . given the relaxation time @xmath5 and effective viscosity @xmath275 , the kinetic equation ( 5.11 ) generates various fields such as @xmath40 and @xmath8 . however , to evaluate the relaxation time , the magnitude of the local dissipation rate is needed . this can be done selfconsistently , if , in accord with kolmogorv theory , we assume where @xmath278 denotes position ( coordinate ) of a lattice site and the averaging is carried out selfconsistently on a non - equilibrium pdf , which is a solution to the kinetic equation ( 5.11 ) . another possibility is to use the dissipation rate directly from the energy balance ( 8.5 ) . + the proportionality constants in ( 8 . 2 ) are reasonably close to those quoted in speziale ( 1987 ) , rubinstein et al ( 1990 ) and yoshizawa ( 1987 ) . it should be pointed out that the value of calculations based the low - order trancations is not clear . indeed , recasting the results of sections 7 and 8 in terms of dimensionless rate of strain @xmath279 , we find that @xmath280 and the next order gives : @xmath281 . taking into account that in the flows of engineering importance the parameter @xmath80 is not small , often reaching values @xmath282 , the perturbation expansion does not converge and no trancation is , in general , possible . _ we would like to reiterate that the entire series is contained in a relatively simple kinetic equation ( 5.11)_. the low - order hydrodynamic approximations and turbulence models , contained in ( 5.11 ) , are similar to the well - known ones , which have been widely tested in both scientific and engineering environments ( see chen ( 2004 ) , for example ) . the future simulations of strongly nonlinear , rapidly distorted or oscillating flows , will provide a decisive test of the limits of validity and accuracy of calculations based on the kinetic equation derived in this paper . benzi r. , succi s. , & vergassola m. , 1992 , _ the lattice boltzmann equation : theory and applications _ , phys . rep . * 222 * , 145 . bhatnagar p.l , gross e & m. krook , 1954 _ a model for collisions in gases i. small amplitude processes in charged and neutral one - component systems_. phys . 94 * , 511525 . chandrasekhar s. 1951 , _ london a * 210 * , 26 . chen h. , kandasamy s. , orszag s.a . , shock r. , succi s. , & yakhot v. 2003 , _ extended boltzmann kinetic equation for turbulent flows_. science * 301 * , 633636 . chen h , orszag s.a . , staroselsky i. , & succi s. 2004 , _ expanded analogy between boltzmann kinetic theory of fluids and turbulence _ , j. fluid mechanics , * 519 * 301 - 313 . , & doolen g. 1998 _ lattice boltzmann method for fluid flows _ , ann . rev . fluid mech . * 30 * , 329 . frisch , u. 1996 _ turbulence _ cambridge university press . kolmogorov , a.n . 1942 , _ the equation of turbulent motion in an incompressible viscous fluid _ , izv . nauk sssr , ser . fiz * vi * , 56 - 58 . landau l.d . & lifshitz e.m . _ physical kinetics _ , butterworth - heinenann , oxford ( 1981 ) . launder , b. & spaulding , d. 1974 , _ the numerical computations of turbulent flows _ , comput . meth . appl . mech * 3 * , 269 - 289 . polyakov a.m. 2001 , private communication . prandl , l. 1925 _ bericht ber untersuchungen zur ausgebildeten turbulenz_. z. angew . math . mech . * 5 * , 136 - 139 . rubinstein , r.b . & barton , j.m . 1990 nonlinear reynolds stress models and the renormalization group , phys . fluids a*2 * , 1472 - 1476 . sinai , ya . g. & yakhot , v. 1989 _ limiting probability distributions of a passive scalar in a random velocity field _ , * 63 * 1962 - 1965 . schumacher , j. , sreenivasan , k.r . & yakhot , v. 2007 _ asymptotic exponents from low re flows _ new journ phys . * 9 * , 1 - 19 . speziale , c. 1987 _ on the nonlinear @xmath283 and @xmath267 models of turbulence _ . mech . * 78 * , 459 - 467 . staroselsky , i , yakhot , v. , kida , s. & orszag , s 1988 _ long - time , large - scale properties of randomly stirred compressible fluid _ , phys.rev.lett . , * 65 * , 171 - 174 . succi , s. _ the lattice boltzmann equation for fluid dynamics and beyond _ oxford university press . wyld , _ formulation of the theory of turbulence in incompressible fluid . _ , annals of physics * 14 * , 143 - 165 ( 1961 ) . yakhot , v. , orszag . s. , thangam , s. , gatski , t. & speziale , c. 1992 _ development of turbulence models for shear flows by a double expansion technique_. phys . fluids a*4 * , 1510 - 1520 . yakhot , v. & smith l.m . , 1992 _ the renormalization group , the @xmath284-expansion and derivation of turbulence models _ , j. sci . * 7 * , 35 - 61 . yakhot , v. & sreenivasan , k.r . anomalous scaling of structure function and dynamic constraints on turbulence simulation . . phys . * 121 * , 823 - 841 . yoshizawa , a. 1987 _ statistical modeling of transport - equation for the kinetic energy dissipation rate . fluids * 30 * , 628 - 631 . yakhot , v. & zakharov , v.e . , 1993 _ hidden conservation laws in hydrodynamics ; energy and dissipation rate spectra in strong turbulence . _ physica d * 64 * , 380 - 394 . zakharov , v.e . , 1984 _ kolmogorov spectra in weak turbulence _ in basic plasma physica , galeev a.a . & sudan r.m . , editors , ( north holland , amsterdam ) zakharov , v.e . , falkovich , g. & lvov , v.s . , 1992 _ kolmogorov spectra in weak turbulence . _ springer , heidelberg zakharov v.e . & lvov , v.s . , 1975 _ statistical description of nonlinear wave fields . _ , radiophysics and electronics , springer , heidelberg .

the galaxy luminosity function ( @xmath10 ) remains at the core of both galaxy evolution and cosmology . by integrating @xmath10 over space and time , various observable distributions can be obtained . in particular , differential number counts of galaxies as a function of apparent magnitude , @xmath0 , is obtained by integrating @xmath10 over all redshifts and morphological types . the reliability of the predicted @xmath0 values depends directly on the cosmological model and on how well the evolution of galaxies is modelled from their formation to the present . the study of number counts can therefore be used to test world models or to search for evolution during the look - back time . the function @xmath0 provides one of the most fundamental observables and has been studied by several authors ( for reviews see ellis 1997 ; koo & kron 1992 ; sandage 1988 ) . recently , a major effort has been made to reach very faint magnitudes using the _ hubble space telescope _ ( _ hst _ ) . although the _ hst _ can not compete with ground - based observations in terms of collecting area , it does provide an unprecedent view of the optical sky at small angular scales and faint flux levels . a number of authors have studied faint galaxy counts based on the hubble deep field ( hdf ) @xcite and using different photometry packages @xcite . in this paper , we use the notation @xmath7 , @xmath8 , and @xmath9 to denote magnitudes in the _ hst _ passbands on the @xmath2 system @xcite . @xcite has compared the hdf catalogues available in the literature . they generally show good agreement , although the effects of different isophotal thresholds and different splitting algorithms are apparent . however , @xcite showed that , at @xmath11 , the different galaxy counts agree to within @xmath12 in all the catalogues , whereas at @xmath13 there is a factor of 1.7 difference among them . this emphasizes the fact that galaxy counting is not a precise science . in this paper , the hdf will be used to test the validity of previous number count studies and to produce a new determination of the faint end of @xmath0 for magnitudes fainter than @xmath1 . surface - brightness fluctuation ( sbf ) measurements are used . this allows us to explore @xmath0 beyond the limiting magnitude of the hdf and to overcome most of the limitations arising from incompleteness , providing new , stronger constraints on the faint end of @xmath0 . this work is based on data from the hdf north . observations were made in 1995 december 1830 , and both raw and reduced data have been put into public domain as a community service @xcite . version 2 f450w , f606w and f814w images , released into public domain on 1996 february 29 , have been used here . the final version 2 images were combined using the drizzle algorithm . drizzling causes the noise in one pixel to be correlated with the noise in the adjacent one @xcite . the sbf technique is based on the spectral analysis of an image . because the power spectrum of the image is modified by the drizzling process , drizzled images were not considered in this study . `` weighted and cosmic - ray cleaned '' images have been used instead . these are flat - fielded , cosmic - ray - rejected and sky - subtracted stacked images at each dither position . only dark exposures have been selected . each of the the three wide - field and the planetary camera wfpc2 chips were analyzed separately . the labels of all images considered , the corresponding filter and the total exposure times are listed in table [ t - data ] . the sbf concept was introduced by @xcite , who noted that , in the surface photometry of a galaxy far enough away to remain unresolved , a pixel - to - pixel fluctuation is observed because of the poisson statistics of the spatial distribution of stars , globular clusters , background galaxies , etc . this technique was introduced with the aim of measuring distances . comparing sbfs produced by the stellar population of a galaxy with those of nearby galaxies for which externally calibrated distances are available , accurate estimates of distances can be obtained up to @xmath1440 mpc @xcite . other authors have used sbf studies to determine the age and metallicity of unresolved stellar populations of nearby galaxies @xcite ; however , the sbf signal can provide information about other kinds of undetected and unresolved objects in an image . this is the case , for example , of globular cluster populations @xcite . in this paper , sbfs have been used to characterize the faint end of @xmath0 using undetected galaxies in the hdf - n images . next in this section , the theoretical background of sbfs and the hdf - n signal measurement are described in detail . the sbf technique involves spectral analysis of the pixel - to - pixel fluctuation signal . this provides the total point spread function psf - convolved variance ( @xmath15 ) produced by all point objects whose spatial flux distribution is convolved with the psf , and the total non - psf - convolved variance ( @xmath16 ) . a convolution in the real space transforms into a product in fourier space . for this reason , the power spectrum of an image , @xmath17 , has the form : @xmath18 where @xmath19 is the power spectrum of the psf convolved with some window function . the @xmath19 can be approximated by the power spectrum of the psf alone , @xmath20 , with a negligible error @xcite . so eq . [ pow1 ] transforms into @xmath21 once the power spectrum of an image has been computed , the variances @xmath15 and @xmath16 can be obtained fitting eq . [ pow2 ] to data . in hdf images , @xmath15 is mainly produced by faint galaxies , and , as we will see later , by cosmic rays . so @xmath15 must be equal to the sum of the variances produced by faint galaxies ( @xmath22 ) and by cosmic rays ( @xmath23 ) : @xmath24 on the other hand , @xmath16 is the sum of read - out noise ( @xmath25 ) , photon shot noise ( @xmath26 ) , and dark current ( @xmath27 ) variances : @xmath28 the pixel - to - pixel variance produced by a class of objects can be evaluated as the second moment of the differential number counts of that object population @xcite . in the present study , the target population is composed by faint undetected galaxies . the brightest individuals , which are detected , must be masked out before the sbf analysis . if all sources brighter than a limiting flux ( @xmath29 ) are masked , then the variance produced by the remaining non - masked population is : @xmath30 we can put this equation in terms of magnitudes via the relationship @xmath31 @xmath32 being the flux ( @xmath33 s@xmath34 pix@xmath34 ) , and @xmath35 the magnitude of an object yielding one unit of flux per unit time ; that is , the photometric zero point listed in table [ t - zero ] . if @xmath0 is known , then the theoretical variance produced by faint galaxies can be estimated . assuming the following pure power - law form for @xmath0 : @xmath36 where @xmath37 is a normalizing constant and @xmath38 is the slope of the magnitude distribution , the variance produced by the non - masked galaxy population is then @xmath39 where @xmath40 is the limiting magnitude ; i.e. the magnitude corresponding to the limiting flux , @xmath29 . we will refer to this variance as the @xmath0-estimated @xmath41 from here on . as we have shown , in order to compute it , an initial @xmath0 , obtained from a given photometric catalogue , must be assumed . a number of authors have studied @xmath0 in the hdf - n . the main discrepancies between previous number count studies occur between @xcite and @xcite . these coincide for `` bright '' magnitudes , but large differences arise for fainter galaxies , the @xcite number counts being larger than those of @xcite . as pointed out by @xcite , there are two reasons for this discrepancy . first , as we go fainter the @xcite magnitudes become systematically brighter than those of @xcite . @xcite claimed that the effect of this on the counts is actually quite small , generally @xmath42 . second , @xcite find objects that @xcite do not detect at all . this appears to account for the majority of the differences between the data sets . @xcite argued that virtually all of these objects are merged in the reductions of @xcite but not in their data . in figure [ glf ] , differential number counts results from @xcite ( filled circles ) and @xcite ( open circles ) are plotted for the f450w , f606w , and f814w filters , respectively . these data have been obtained from tables 9 and 10 in @xcite , and from tables 8 , 10 and 12 in @xcite , all expressed in the total magnitude scale . following eq . [ bglf ] , @xmath0 has been fitted to the data of @xcite and @xcite , and the fitted functions ( solid lines ) are also plotted . as @xcite found a change in the @xmath0 slope at a magnitude of around 26 , @xmath0 has been fitted to their data in the two magnitude intervals [ 23 , 26 ] and [ 26 , 29 ] . the fitted @xmath0 functions have been used in eq . [ sigma_bg ] to compute the @xmath0-estimated @xmath41 . results are listed in table [ t - sigma ] . these @xmath0-estimated @xmath41 values will be later on compared with those directly derived from the sbf measurements ( which we will call the sbf - measured @xmath41 ) . this comparison will allow us to evaluate the validity of the differential number counts of both @xcite and @xcite and , as a result , a final @xmath0 will be proposed . before describing the details of the sbf measurements in hdf images , note that the sbf technique is valid only if faint galaxies have a stellar appearance . in @xcite the radius magnitude relation for galaxies in the @xcite hdf catalogue has been analyzed . for magnitudes fainter than @xmath43 , all the galaxies have a radius smaller that @xmath44 , close to the fwhm of the wf chips . so very faint galaxies , in the magnitude range where sbf will be measured , can be assumed to have a stellar appearance . here , the practical procedure for obtaining the sbf signal in hdf - n images is described in detail . first of all , it should be noted that cosmic rays are difficult to discriminate from stars in _ hst _ frames . in order to estimate and eliminate the cosmic - ray contribution to the sbf signal , @xmath15 , a procedure based on the random nature of cosmic - ray events has been used for each filter and chip . the sbf signal has be measured not only on the final combined images , but on all the individual images listed in table [ t - data ] as well . before computing the power spectrum of an image , objects brighter than @xmath40 must be masked out . in this study , the window functions have been created using the @xcite photometric catalogue , which is the only one available to us . as isophotal magnitudes were considered while creating the window functions , in order to convert them to the total magnitude scale an isophotal - to - total magnitude correction of 0.2 mag @xcite was applied . the sbf analysis have been be performed considering two different values of @xmath40 : 27.8 and 28.8 . all objects brighter than @xmath40 have been masked out using a window function whose pixel values are zero in a circle centered on the location of the bright objects and unity in the rest of the image . the window function has been created using a patch radius large enough to completely mask bright galaxies , including their external haloes and therefore merged galaxies where these exist . the procedure creating the mask has been the following : first , the brightest galaxies have been masked one by one manually . in order to avoid residual light beyond the masked regions , very generous patch sizes have been adopted . the shape of the used patchs for these very bright galaxies depends on the shape of the particular masked galaxy . once bright galaxies have been masked , the rest of galaxies brighter than the chosen @xmath40 have been masked using circular patches . the radius of these patches has been chosen to be the same for all galaxies , and its size is again very generous : the adopted patch radius is 15 pixels . in order to test if all the residual light beyond the masked regions has been eliminated , the sbf analysis has been repeated for one image varying the radius of the patches . we have considered patch radii of 20 , 25 and 30 pixels . note that if a radius larger than 30 pixels would be used , the image would be completely masked due to the superposition of adjacent patches . the considered image has been the f450w average image of wf2 with @xmath45 . the sbf results are listed in table [ patches ] . it can be seen that the sbf results are independent of the patch radius . as a conclusion , it can be seen that the adopted patch radius avoid residual light beyond the masked regions , as required . if the differences between the @xcite and @xcite data sets relies on the objects that are merged in the former and not in the latter , as argued by @xcite , then the created window function is virtually the same as one would obtain using the @xcite photometric catalogue , and also masks objects that are in this catalogue and not in that of @xcite . next , multiplying the window function by the image , the residual masked image is obtained . it is this image which is used to compute the power spectrum . as has been said , two sets of masked images have been computed , one for each @xmath46 value . the power spectrum of the masked image is two - dimensional . it is radially averaged in order to obtain the one - dimensional power spectrum . fitting the power spectrum of the images with eq . [ pow2 ] , the quantities @xmath15 and @xmath16 can be obtained for each image . as we could not construct a psf from hdf images , since there were not enough stars in any of the four chips , the used psfs were the high - s / n psfs extracted by p. b. stetson from a large set of uncrowded and unsaturated wfpc2 images . spatial variations of the psf along the ccd mosaic could introduce a significant uncertainty in @xmath15 . to limit this effect , we have used a psf template for each one of the four chips of the wfpc2 . minor psf variations inside each particular chip have not been considered . for each chip , both psf and the computed power spectrum represent mean values across the complete field of the chip , so any spatial variation of the psf across the chip would affect the sbf fitting procedure , introducing an uncertainty in the @xmath15 measurement . this uncertainty is small , and in any case , it is included in the obtained @xmath15 uncertainty . in fig . [ @xmath7.average.29 ] we show an example of the sbf fitting procedure in an hdf image . the observed discrepancy at low wave numbers between the obtained power spectrum and the fit is due to large scale fluctuations in the background brightness of the images . this wave number region , where the discrepancy occurs , is not taken into account when fitting eq . the power - spectrum fitting procedure is the following : eq . [ pow2 ] is used to fit the power spectrum for wave numbers in the range [ @xmath47 , @xmath48 , @xmath49 being the highest wave number of the computed power spectrum , and @xmath47 a number which we vary from 0 to @xmath49 . as a result , two functions , @xmath50 and @xmath51 , are obtained . the function @xmath50 is also shown in fig . [ @xmath7.average.29 ] ( small boxes ) . it can be seen that this function exhibits a `` plateau '' region . the final adopted result and its uncertainty for @xmath15 are obtained computing the average and standard deviation of @xmath50 in the plateau interval . for the @xmath16 measurement , the procedure is exactly the same as for @xmath15 . @xmath15 and @xmath16 results corresponding to @xmath45 and @xmath52 are listed in tables [ t - results28 ] and [ t - results29 ] , respectively . results are listed for all images , corresponding to all chips and all filters . in this section , results from tables [ t - results28 ] and [ t - results29 ] will first be carefully analyzed , and will be used to test the probable effect of flat - fielding errors on the measured @xmath15 . then , the obtained @xmath15 values in different images will be used to estimate the contribution of cosmic rays and to deduce the desired sbf - measured @xmath41 . finally , in order to check the sbf results , two consistency tests will be performed . the possibility of flat - fielding errors affecting the measured @xmath15 must considered . if flat - fielding would contribute to @xmath15 , then its effect should be larger in images with high sky background . in this context , considering same filter and chip images , if flat - fielding is contributing to @xmath15 , a relation between @xmath15 and the sky level should appear . to test if the flat - fielding errors have an influence on the measured @xmath15 , filter f606w has been considered because it provides a larger number of images covering a wider range of sky levels . sky levels are provided in the weighted , cosmic - ray cleaned image headers . as an example , fig . [ flatfielding ] shows the measured @xmath15 as a function of these sky levels for the wf2 f606w images with @xmath45 . it can be seen that no relation is apparent , so flat - fielding errors are insignificant in the @xmath15 measurements . the _ hst _ provides images of exceptional resolution . as a consequence , discriminating stars from cosmic rays is a difficult task . this also has implications for sbf measurements . if a cosmic - ray event alters only one pixel , then it is handled as white noise , and consequently , it contributes to the @xmath16 signal . but if two or more pixels are affected , then a cosmic ray can be confused with a point source and it will therefore contribute to the @xmath15 signal . in the wfpc2 , the majority of cosmic ray events affect 23 pixels , so if undetected cosmic rays exist in an image , the sbf results will be affected by an increase of both @xmath16 and @xmath15 . but there is noise also due to detected cosmic rays . the pixels containing cosmic rays have been masked during the data reduction of the images , and therefore , the noise in these pixels is higher than it is in pixels where none of the input images were masked . as we will see later , the contribution of cosmic rays to @xmath15 is not negligible , and is necessary to control this effect in order to obtain the sbf - measured @xmath41 . it is for this reason that a strategy aimed at estimating and eliminating the contribution of cosmic rays from @xmath15 must be developed . such a strategy is described next . in an individual weighted and cosmic - ray cleaned image , @xmath15 has two contributions , faint galaxies and cosmic rays : @xmath53 if such individual images averaged , the contributions from galaxies and cosmic rays will be different , because of their different natures . measured in the averaged image , @xmath15 becomes @xmath54 where @xmath55 is the number of individual images used for the average . in tables [ t - results28 ] and [ t - results29 ] , results for @xmath15 are listed for all the individual weighted and cosmic - ray cleaned images and for the averaged images as well . if the influence of cosmic rays were negligible , @xmath15 would be the same in all the individual images and in the averaged one . it can be seen that this is not the case , from which we may conclude that the contribution of cosmic rays to @xmath15 is non - negligible . equations [ cr1 ] and [ cr2 ] can now be used to obtain @xmath41 and @xmath56 . for @xmath15 , the mean of the values of the individual images for each chip and filter have been used . results are listed in table [ t - results - bg ] for the two considered values of @xmath46 . the final sbf - measured @xmath41 values for each filter can now be obtained from the averages of the single - chip results . they are given in the last row of each @xmath46 set in table [ t - results - bg ] and will be compared in [ discussion ] with the @xmath0-estimated @xmath41 values . in this section , a couple of consistency tests have been performed in order to check the reliability of our sbf - measured @xmath41 results . together with the psf - convolved variance , @xmath15 , sbfs provide the value of @xmath16 , the non - psf - convolved variance . this can be compared with its expected value , directly obtained from the read - out noise , the dark current , and the sky brightness analysis of each image . for hdf version 2 , each of the weighted and cosmic - ray cleaned images is the result of combining several exposures with the same dither position . the different exposures are combined with weights proportional to the inverse variance ( @xmath57 ) at the mean background level . the variance @xmath16 , in electrons , is computed from the following noise model @xcite : @xmath58 where @xmath59 is the exposure time , @xmath60 is the sky background rate , @xmath61 is the dark current , and @xmath62 is the read - out noise . the inverse variances , @xmath57 , of each exposure are provided in the header of the resulting weighted and cosmic - ray cleaned image . from this information , the value of @xmath16 corresponding to the latter can be computed . as an example , these values are listed in table [ t - p1 ] ( column 2 ) for the wf2 images and @xmath7 . the @xmath16 values obtained directly from the sbf analysis of the images , using @xmath63 , are listed in column 3 . both computed and observed values of @xmath16 are equivalent in all cases . only a slight excess in the observed @xmath16 is noticeable . this excess is produced by cosmic rays , which also contribute to the measured @xmath15 values , as we have shown . this test shows that , with this technique , the white noise ( @xmath16 ) is determined with high precision , thereby reinforcing the correctness of @xmath15 measurements . [ interval ] this second consistency test is based on the comparison of the parameter @xmath64 $ ] computed in two different ways . we call @xmath64 $ ] to the variance produced by galaxies with magnitudes within a given interval : @xmath65(m_{a},m_{b } ) \equiv \int_{f_{b}}^{f_{a } } n(f ) f^2 df,\ ] ] where @xmath66 is the magnitude corresponding to a flux @xmath67 , and @xmath68 to @xmath69 . defined in this way , @xmath64 $ ] is the difference between the variances computed with two values of @xmath40 , namely @xmath70 and @xmath71 : @xmath72(m_{a},m_{b})=\sigma_{\rm bg}^2(m_{\rm c}=m_{a})-\sigma_{\rm bg}^2(m_{\rm c}=m_{b } ) \nonumber\\\end{aligned}\ ] ] the @xmath0-estimated and sbf - measured @xmath64 $ ] can be now compared . to do this , the magnitude interval [ 27.8 , 28.8 ] has been considered . as the @xcite photometric catalogue has been used to create the window functions , only objects found by them in the magnitude interval [ 27.8 , 28.8 ] contribute to the sbf - measured @xmath64 $ ] . in particular , some of the objects within the interval [ 27.8 , 28.8 ] in the @xcite catalogue will remain masked , namely those which , following @xcite , are merged in the @xcite catalogue . this implies that the sbf - measured @xmath64 $ ] is not expected to coincide with the @xmath0-estimated @xmath64 $ ] computed using the @xcite data . on the other hand , the @xmath0-estimated @xmath64 $ ] computed using the @xcite data should , for rigor , be similar to the sbf - measured @xmath64 $ ] only if all the objects of the @xcite catalogue in the interval [ 27.8 , 28.8 ] are unmerged however , if merged galaxies exist in other intervals @xcite , there is no reason why they should not be present here also . the effect of mergers on @xmath64 $ ] can be tested considering that a fraction of @xcite objects in the interval [ 27.8 , 28.8 ] are the result of a merger between two fainter galaxies with integrated fluxes @xmath73 and @xmath74 . we have considered three simple situations : i ) @xmath75 ; ii ) @xmath76 ; and iii ) @xmath77 . it should be noted that since sbfs are a measure of the second moment of the brightness function , the more similar @xmath73 and @xmath74 are , the larger is the effect introduced in @xmath64 $ ] . so case i ) is the most pessimistic and , although it is unrealistic , will give the maximum expected effect on @xmath64 $ ] for a given fraction of mergers . in figure [ mergers ] we show the results of @xmath0-estimated @xmath64 $ ] for the three cases and the f606w filter , considering different values of the percentage of merged objects . it can be seen that , even in the very pessimistic case where @xmath78 of the @xcite objects are mergers of two identical galaxies , their influence on @xmath64 $ ] is less than @xmath12 . for example , in a perhaps more realistic situation in which about 2025% of the objects are mergers of two galaxies with different magnitudes , the effect on @xmath64 $ ] would smaller than @xmath1410% ( see figure [ mergers ] ) . in conclusion , the effect of mergers on @xmath64 $ ] computed using the @xcite number counts is small and can be expected to remain below about 15% for any reasonable scenario . we can hence proceed with our test on @xmath64 $ ] for the [ 27.8 , 28.8 ] interval using the @xcite data . the results for sbf - measured and @xmath0-estimated @xmath64 $ ] are given in table [ t - interval ] . the value of @xmath0-estimated @xmath64 $ ] has been computed using @xcite and @xcite data . the results are given in columns 2 and 3 , while column 4 lists the @xmath0-measured @xmath64 $ ] values . it can be seen that they are not compatible with the values obtained from the @xcite data , as expected . on the other hand , the @xmath0-estimated @xmath64 $ ] obtained from the @xcite number counts and the sbf - measured @xmath64 $ ] are very close . in order to compare them with more detail , let analyze figure [ mergers ] again . the @xmath0-measured @xmath64 $ ] value for the f606w filter and its error interval have also been plotted in the figure ( shadowed region ) . it can be seen now that they fully coincide if a number of mergers about 2030% is assumed . this situation is realistic and compatible with @xcite s claims . summarizing , the test has been successful and shows that the sbf measurements are well - calibrated . in this section , sbf - measured @xmath41 results ( listed in table [ t - results - bg ] ) and the @xmath0-estimated @xmath41 values obtained from both the @xcite and @xcite data ( listed in table [ t - sigma ] ) will be compared . in the following , two possibilities will be considered and their consequences discussed : i ) that the @xcite data represent the right differential number counts , and ii ) that the @xcite number counts are correct . the obtained sbf measurements will be used to test the validity of these possibilities and , as a result , a final @xmath0 will be proposed . here we assume that the @xcite number counts are correct . comparing the @xmath0-estimated @xmath41 obtained using @xcite data and sbf - measured @xmath41 , it can be seen that the former are much larger . there are only two possible sources to account for this discrepancy : first , a faint unresolved stellar population , belonging to the milky way halo , could be responsible of the excess in the sbf signal ; and second , the faint end of @xmath0 is different from the fitted one used here to evaluate the @xmath0-estimated @xmath41 . in the first case , sbf results may be used to characterize such a halo population . in the second case , sbf results may be used to deduce a new faint end of @xmath0 able to account for the sbf - measured @xmath41 . we shall analyze both possibilities in detail and discuss the feasibility of each one and its compatibility with the observations ; finally , we shall deduce its implications for @xmath0 . here , we will consider a milky way halo population of faint stars be responsible of the observed excess in the sbf signal . we will firstly deduce the halo population necessary to cause this sbf signal excess . in order to check the feasibility of this hypothesis , the obtained halo population will then be compared with observations in the hdf . lets consider a simple population of objects with absolute magnitude @xmath79 following the standard spatial distribution used by @xcite : @xmath80 where @xmath81 is the number of objects per pc@xmath82 at a distance from the milky way center between @xmath62 and @xmath83 ; @xmath84 is the core radius , and @xmath85 is the object density in the milky way center . for simplicity we take @xmath86 . to derive the sbf signal from the former population , we must first express the equations in terms of distance from the sun ( @xmath87 ) . this can be done using : @xmath88 where @xmath89 is the galactocentric radius , i.e. , the distance from the sun to the milky way center , and @xmath90 are galactic coordinates . the spatial distribution of objects expressed in spherical coordinates is then : @xmath91 integrating for the hdf - n and considering its coordinates @xmath92 and @xmath93 , the former expression is reduced to : @xmath94 now , @xmath95 can be written in terms of magnitudes by means of the distance modulus to obtain : @xmath96 the number of resolved objects with @xmath97 that should appear in the hdf - n , ( @xmath98 ) can be now deduced from eq . [ eq2 ] , as well as the variance that the faint part of the population ( @xmath99 ) would produce ( @xmath100 ) : @xmath101 @xmath102 in order to compare these predictions with the hdf data , the @xmath8 filter results have been used . a value must be assumed for the core radius , @xmath84 . realistic values are around @xmath103 pc @xcite , but @xmath104 pc and @xmath105 pc have been also used to check a wide interval of possibilities . making @xmath100 equal to the @xmath106 excess observed in the filter @xmath8 for @xmath52 ( that is , @xmath107 @xmath108 ^ 2 $ ] , see tables [ t - sigma ] and [ t - results - bg ] ) and introducing the value in eq . [ eq4 ] , the central density @xmath85 can be obtained and used in eqs [ eq1 ] and [ eq2 ] to derive @xmath109 and @xmath110 . results are plotted in fig . [ mod_halo ] , where @xmath109 and @xmath98 are shown as functions of @xmath111 , the absolute magnitude of the halo population objects . this figure implies the existence of a large number of halo objects that should be present in the hdf images . note that @xmath98 is greater than @xmath14250 in all cases . this result is not compatible with the hdf - n observations , where no obvious stars are present , except for a few 20th magnitude ones , @xcite . in conclusion , the observed excess in sbf - measured @xmath106 can not be produced by objects belonging to the milky way halo . otherwise a large number of resolved objects from this halo population would show up in the hdf - n images , which is not the case . if the observed @xmath112 excess can not be produced by milky way halo objects , the only possibility is that it is caused by faint galaxies . the large excess obtained in the sbf - measured @xmath112 with respect to the @xmath0-estimated @xmath112 would imply an increase in the slope of @xmath0 at some magnitude fainter than @xmath52 . this slope can be computed by fitting the @xmath0-estimated @xmath112 to our sbf - measured @xmath112 and taking the slope as a free parameter . if it is assumed that the slope change occurs at @xmath113 for all filters , the resulting slopes for the fainter range are @xmath114 , @xmath115 , and @xmath116 for @xmath7 , @xmath8 , and @xmath9 , respectively . these slopes would be valid up to @xmath117 , @xmath118 , and @xmath119 at least since the contribution of fainter magnitudes to @xmath112 becomes smaller than the uncertainties in the sbf - measured @xmath112 results . if the slope change were to occur at a magnitude fainter than 28.8 , it would result in a steeper @xmath0 . in any case , such big changes in the slope of @xmath0 seem unrealistic . in our opinion , this possibility should be rejected . as a consequence , it must be concluded that the @xcite data are incomplete . assuming that the @xcite differential number counts are correct , the sbf - measured @xmath41 results listed in table [ t - results - bg ] and the @xmath0-estimated @xmath41 values obtained using the @xcite data , listed in table [ t - sigma ] , can be compared . it can be seen that the sbf - measured and @xmath0-estimated @xmath41 coincide within the error bars for the f814w filter , and is very similar for the f450w filter . only in the filter f606w some differences arise . this implies that extrapolation of the @xcite @xmath0 function to magnitudes fainter than 28.8 accounts almost entirely for the measured sbf signal , thus indicating a high level of precision in the @xcite data . however , as sbf - measured and @xmath0-estimated @xmath41 present slight differences in the f450w and f606w filters , the most likely @xmath0 function that completely fits our sbf measurements can be determined . this will be done in the next subsection . in this section , the most likely @xmath0 function will be obtained for magnitudes fainter than 28.8 . we consider that the most likely @xmath0 function is that which fits our sbf measurements . an @xmath0 function is completely determined by giving the slope ( @xmath38 ) and the number of galaxies in a given area and magnitude interval ( @xmath120 ) . in particular we consider @xmath121 , computed for the magnitude interval [ 28.75 , 29.25 ] and the area of the hdf - n . note that there are infinite sets of ( @xmath38 , @xmath121 ) that can produce the same sbf signal . in figure [ sbf ] , the ( @xmath38 , @xmath121 ) pairs that can account for our measured sbf signal have been plotted ( solid lines ) for the f450w , f606w , and f814w filters . with short - dashed lines we represent the ( @xmath38 , @xmath121 ) pairs that would produce the sbf measurements @xmath122 . in this figure , the ( @xmath38 , @xmath121 ) sets corresponding to the extrapolation of @xmath0 obtained from @xcite ( solid circle ) and @xcite data ( open circle ) have also been plotted . it can be seen in figure [ sbf ] that the @xcite ( @xmath38 , @xmath121 ) reproduce our sbf results for the f814w filter , as previously shown . for f450w , the @xcite ( @xmath38 , @xmath121 ) is very close to our sbf measurements . in all cases ( except for f606w ) , the @xcite ( @xmath38 , @xmath121 ) results are far from our sbf results , as previously argued . the most likely @xmath0 function can be obtained from figure [ sbf ] . we assume that , for each filter , the best @xmath0 estimate for magnitudes fainter than 28.8 is given by the nearest point of the solid lines to the @xcite point . in this case , results for the slopes are @xmath4 , @xmath5 , and @xmath6 for @xmath7 , @xmath8 , and @xmath9 , respectively . the results are listed in table [ t - glf ] and plotted in figure [ glf2 ] ( solid lines ) . the slopes obtained are valid down to magnitude @xmath3 at least . the contribution to the sbf signal by objects of fainter magnitudes is less than the uncertainty in the sbf measurements . this value represents an extension of more than two magnitudes beyond the limits of the previous photometric studies by @xcite and @xcite . it should also be mentioned that , within the uncertainties and except for the objects that could be merged into brighter ones in @xcite , it is free from incompleteness . in this paper , the faint end of the differential galaxy number counts , @xmath0 , has been studied by means of sbf measurements . once the contribution from cosmic rays has been evaluated and eliminated from the sbf signal , the background psf - convolved variance originating from faint objects has been carefully analyzed . our conclusions can be summarized as follows : * comparing the sbf - measured @xmath41 with the @xmath0-estimated @xmath41 predicted by the extrapolation of @xcite number counts a clear excess has been found in the measured signal . the possibility that the excess might be produced by milky way halo stars is ruled out because it would be totally incompatible with the resolved stellar population present in the hdf . on the other hand , if this excess is caused by a faint galaxy population modifying the faint end of @xmath0 , then the required slopes for magnitudes fainter than 28.8 are @xmath114 , @xmath115 , and @xmath116 for @xmath7 , @xmath8 , and @xmath9 , respectively . such big changes in the @xmath0 slope seem unrealistic . in our opinion , this possibility should be rejected . in conclusion , the @xcite number counts are not compatible with our sbf measurements , probably owing to the incompleteness in their data . * comparing the sbf - measured @xmath41 with the @xmath0-estimated @xmath41 , predicted by the extrapolation of @xcite number counts , we find that they coincide within the error bars for the f814w and f450w filters and are similar for f606w . this implies that the extrapolation of the @xcite @xmath0 function to magnitudes fainter than 28.8 nearly accounts for the measured sbf signal , indicating a high level of precision in the @xcite results . * the most likely @xmath0 function has been obtained fitting our sbf results . results for the @xmath0 slope for magnitudes fainter than @xmath1 are @xmath4 , @xmath5 , and @xmath6 for @xmath7 , @xmath8 , and @xmath9 , respectively . the obtained slopes are valid down to magnitude @xmath3 at least . this value represents an extension of more than two magnitudes beyond the limits of the previous photometric studies by @xcite and @xcite . this work is based on observations with the nasa / esa _ hubble space telescope _ , obtained in the space telescope science institute , which is operated by the association of universities for research in astronomy , inc . ( aura ) , under nasa contract nas5 - 26555 . this research has been supported by the instituto de astrofsica de canarias ( grant p3/94 ) , the dgesic of the kingdom of spain ( grant pi97 - 1438-c02 - 01 ) , and the dgui of the autonomous government of the canary islands ( grant pi1999/008 ) . lcc f450w.d1.dark & f450w & 17200 + f450w.d2.dark & f450w & 7900 + f450w.d3.dark & f450w & 11500 + f450w.d4.dark & f450w & 12400 + f450w.d5.dark & f450w & 11800 + f450w.d6.dark & f450w & 12900 + f450w.d8.dark & f450w & 13100 + f450w.d9.dark & f450w & 10700 + + f606w.d1.dark & f606w & 6700 + f606w.d2.dark & f606w & 4800 + f606w.d3.dark & f606w & 6450 + f606w.d4.dark & f606w & 10300 + f606w.d5.dark & f606w & 15600 + f606w.d6.dark & f606w & 17300 + f606w.d7.dark & f606w & 14600 + f606w.d8.dark & f606w & 10100 + f606w.d9.dark & f606w & 7100 + f606w.d10.dark & f606w & 8300 + f606w.d11.dark & f606w & 7800 + + f814w.d1.dark & f814w & 10800 + f814w.d2.dark & f814w & 12200 + f814w.d3.dark & f814w & 14200 + f814w.d4.dark & f814w & 12000 + f814w.d5.dark & f814w & 13000 + f814w.d6.dark & f814w & 13100 + f814w.d8.dark & f814w & 5800 + f814w.d9.dark & f814w & 12900 + lcc f450w & pc1 & 21.92 + f450w & wf2 & 21.93 + f450w & wf3 & 21.93 + f450w & wf4 & 21.90 + f606w & pc1 & 23.02 + f606w & wf2 & 23.02 + f606w & wf3 & 23.03 + f606w & wf4 & 23.00 + f814w & pc1 & 22.08 + f814w & wf2 & 22.09 + f814w & wf3 & 22.09 + f814w & wf4 & 22.07 + + + 27.8 & 2.82 @xmath123 & 2.74 @xmath124 & 5.50 @xmath123 + 28.8 & 6.45 @xmath125 & 6.43 @xmath123 & 1.32 @xmath123 + + + + 27.8 & 7.20 @xmath123 & 5.76 @xmath124 & 1.05 @xmath124 + 28.8 & 2.23 @xmath123 & 1.62 @xmath124 & 3.04 @xmath123 + f450w.d1.dark & 1.90 @xmath126 0.16 & 3.86 @xmath126 0.01 & 1.05 @xmath126 0.11 & 5.4 @xmath126 0.1 & 0.86 @xmath126 0.05 & 5.4 @xmath126 0.1 & 1.09 @xmath126 0.05 & 4.6 @xmath126 0.1 + f450w.d2.dark & 3.04 @xmath126 0.10 & 7.21 @xmath126 0.01 & 1.89 @xmath126 0.07 & 11.4 @xmath126 0.1 & 1.70 @xmath126 0.23 & 10.8@xmath126 0.1 & 1.13 @xmath126 0.17 & 10.0 @xmath126 0.1 + f450w.d3.dark & 1.48 @xmath126 0.06 & 5.25 @xmath126 0.01 & 1.06 @xmath126 0.10 & 8.0 @xmath126 0.1 & 1.01 @xmath126 0.07 & 7.8 @xmath126 0.1 & 1.33 @xmath126 0.10 & 6.9 @xmath126 0.1 + f450w.d4.dark & 1.39 @xmath126 0.09 & 4.45 @xmath126 0.01 & 1.34 @xmath126 0.06 & 6.9 @xmath126 0.1 & 1.25 @xmath126 0.09 & 6.7 @xmath126 0.1 & 1.42 @xmath126 0.13 & 6.2 @xmath126 0.1 + f450w.d5.dark & 1.73 @xmath126 0.09 & 5.18 @xmath126 0.01 & 1.21 @xmath126 0.05 & 7.5 @xmath126 0.1 & 1.36 @xmath126 0.04 & 7.2 @xmath126 0.1 & 1.89 @xmath126 0.07 & 6.3 @xmath126 0.1 + f450w.d6.dark & 1.47 @xmath126 0.06 & 4.31 @xmath126 0.01 & 1.04 @xmath126 0.12 & 6.9 @xmath126 0.1 & 1.22 @xmath126 0.03 & 6.7 @xmath126 0.1 & 1.00 @xmath126 0.08 & 6.1 @xmath126 0.1 + f450w.d8.dark & 1.24 @xmath126 0.05 & 4.32 @xmath126 0.01 & 0.77 @xmath126 0.08 & 6.9 @xmath126 0.1 & 1.40 @xmath126 0.05 & 6.7 @xmath126 0.1 & 0.76 @xmath126 0.11 & 6.1 @xmath126 0.1 + f450w.d9.dark & 2.10 @xmath126 0.08 & 5.95 @xmath126 0.01 & 1.13 @xmath126 0.11 & 8.9 @xmath126 0.1 & 0.98 @xmath126 0.21 & 8.5 @xmath126 0.1 & 1.70 @xmath126 0.06 & 7.7 @xmath126 0.1 + average image & 0.29 @xmath126 0.03 & 0.636 @xmath126 0.001 & 0.452@xmath126 0.006 & 0.901@xmath126 0.002 & 0.563 @xmath126 0.014 & 0.847@xmath126 0.004 & 0.441 @xmath126 0.006 & 0.762 @xmath126 0.002 + + f606w.d1.dark & 4.50 @xmath126 0.24 & 17.6 @xmath126 0.1 & 7.7 @xmath126 0.3 & 36.8 @xmath126 0.1 & 6.8 @xmath126 0.9 & 37.4 @xmath126 0.2 & 2.5 @xmath126 0.2 & 35.2 @xmath126 0.1 + f606w.d2.dark & 6.1 @xmath126 0.5 & 27.2 @xmath126 0.1 & 8.2 @xmath126 0.2 & 54.9 @xmath126 0.4 & 5.5 @xmath126 0.6 & 56.5 @xmath126 0.1 & 1.4 @xmath126 0.9 & 52.5 @xmath126 0.3 + f606w.d3.dark & 4.7 @xmath126 0.6 & 22.2 @xmath126 0.1 & 5.5 @xmath126 0.5 & 42.7 @xmath126 0.2 & 6.2 @xmath126 1.7 & 42.1 @xmath126 0.3 & 8.8 @xmath126 0.6 & 42.6 @xmath126 0.3 + f606w.d4.dark & 2.92 @xmath126 0.21 & 13.3 @xmath126 0.1 & 4.2 @xmath126 0.4 & 26.1 @xmath126 0.1 & 6.2 @xmath126 0.5 & 25.7 @xmath126 0.1 & 5.2 @xmath126 0.4 & 23.9 @xmath126 0.1 + f606w.d5.dark & 4.71 @xmath126 0.18 & 9.3 @xmath126 0.1 & 6.1 @xmath126 0.2 & 17.6 @xmath126 0.1 & 6.8 @xmath126 0.6 & 19.1 @xmath126 0.2 & 4.5 @xmath126 0.6 & 16.8 @xmath126 0.2 + f606w.d6.dark & 2.48 @xmath126 0.09 & 7.5 @xmath126 0.1 & 6.5 @xmath126 0.3 & 16.2 @xmath126 0.1 & 6.0 @xmath126 0.2 & 17.4 @xmath126 0.1 & 5.2 @xmath126 0.4 & 14.8 @xmath126 0.1 + f606w.d7.dark & 3.52 @xmath126 0.16 & 9.0 @xmath126 0.1 & 6.3 @xmath126 0.3 & 18.2 @xmath126 0.1 & 7.1 @xmath126 0.4 & 18.6 @xmath126 0.1 & 4.3 @xmath126 0.2 & 16.7 @xmath126 0.1 + f606w.d8.dark & 3.6 @xmath126 0.4 & 12.6 @xmath126 0.1 & 7.4 @xmath126 0.5 & 25.2 @xmath126 0.1 & 6.7 @xmath126 0.6 & 25.7 @xmath126 0.1 & 5.6 @xmath126 0.4 & 24.2 @xmath126 0.2 + f606w.d9.dark & 5.17 @xmath126 0.24 & 20.3 @xmath126 0.1 & 6.6 @xmath126 0.4 & 37.6 @xmath126 0.1 & 7.8 @xmath126 0.3 & 38.4 @xmath126 0.2 & 6.5 @xmath126 0.4 & 33.7 @xmath126 0.2 + f606w.d10.dark & 4.50 @xmath126 0.13 & 15.1 @xmath126 0.1 & 7.3 @xmath126 0.4 & 30.2 @xmath126 0.1 & 7.6 @xmath126 0.6 & 31.3 @xmath126 0.2 & 4.7 @xmath126 0.7 & 28.2 @xmath126 0.2 + f606w.d11.dark & 5.37 @xmath126 0.16 & 15.5 @xmath126 0.1 & 5.5 @xmath126 0.8 & 31.3 @xmath126 0.2 & 7.4 @xmath126 0.8 & 32.3 @xmath126 0.2 & 4.4 @xmath126 0.7 & 28.9 @xmath126 0.2 + average image & 0.96 @xmath126 0.18 & 1.38 @xmath126 0.01 & 3.19 @xmath126 0.13 & 2.68 @xmath126 0.02 & 3.08 @xmath126 0.20 & 2.52 @xmath126 0.04 & 3.2 @xmath126 0.5 & 1.95 @xmath126 0.14 + + f814w.d1.dark & 2.24 @xmath126 0.10 & 7.94 @xmath126 0.01 & 0.83 @xmath126 0.16 & 14.9 @xmath126 0.1 & 1.5 @xmath126 0.3 & 15.7 @xmath126 0.1 & 1.8 @xmath126 0.3 & 14.7 @xmath126 0.1 + f814w.d2.dark & 1.89 @xmath126 0.11 & 7.59 @xmath126 0.01 & 1.99 @xmath126 0.15 & 13.3 @xmath126 0.1 & 3.23 @xmath126 0.09 & 24.4 @xmath126 0.1 & 1.70 @xmath126 0.18 & 13.3 @xmath126 0.1 + f814w.d3.dark & 2.04 @xmath126 0.04 & 6.13 @xmath126 0.01 & 1.34 @xmath126 0.17 & 11.8 @xmath126 0.1 & 1.68 @xmath126 0.11 & 11.9 @xmath126 0.1 & 1.56 @xmath126 0.07 & 10.5 @xmath126 0.1 + f814w.d4.dark & 2.03 @xmath126 0.05 & 6.92 @xmath126 0.01 & 2.51 @xmath126 0.23 & 13.0 @xmath126 0.1 & 1.94 @xmath126 0.20 & 13.8 @xmath126 0.1 & 0.75 @xmath126 0.10 & 13.6 @xmath126 0.1 + f814w.d5.dark & 2.1 @xmath126 0.3 & 7.00 @xmath126 0.01 & 2.6 @xmath126 0.3 & 12.9 @xmath126 0.1 & 0.92 @xmath126 0.14 & 14.6 @xmath126 0.1 & 2.07 @xmath126 0.09 & 13.0 @xmath126 0.1 + f814w.d6.dark & 2.04 @xmath126 0.11 & 6.62 @xmath126 0.01 & 2.7 @xmath126 0.3 & 12.1 @xmath126 0.1 & 3.09 @xmath126 0.17 & 13.6 @xmath126 0.1 & 1.29 @xmath126 0.16 & 12.8 @xmath126 0.1 + f814w.d8.dark & 5.98 @xmath126 0.14 & 20.26 @xmath126 0.01 & 4.1 @xmath126 0.3 & 27.5 @xmath126 0.1 & 3.5 @xmath126 0.3 & 28.6 @xmath126 0.1 & 3.0 @xmath126 0.6 & 25.3 @xmath126 0.1 + f814w.d9.dark & 2.55 @xmath126 0.19 & 6.77 @xmath126 0.01 & 2.0 @xmath126 0.4 & 12.6 @xmath126 0.1 & 2.3 @xmath126 0.3 & 13.0 @xmath126 0.1 & 2.60 @xmath126 0.07 & 12.0 @xmath126 0.1 + 0 average image & 0.413 @xmath126 0.019 & 1.074 @xmath126 0.002 & 1.011 @xmath126 0.024 & 1.669 @xmath126 0.007 & 0.47 @xmath126 0.05 & 2.057 @xmath126 0.012 & 0.88 @xmath126 0.03 & 1.592@xmath126 0.011 + f450w.d1.dark & 1.53 @xmath126 0.04 & 3.66 @xmath126 0.01 & 0.75 @xmath126 0.08 & 5.4 @xmath126 0.1 & 0.62 @xmath126 0.06 & 5.4 @xmath126 0.1 & 0.83 @xmath126 0.04 & 4.7 @xmath126 0.1 + f450w.d2.dark & 3.00 @xmath126 0.09 & 7.14 @xmath126 0.01 & 2.01 @xmath126 0.06 & 11.2 @xmath126 0.1 & 1.48 @xmath126 0.23 & 10.8 @xmath126 0.1 & 1.46 @xmath126 0.17 & 9.9 @xmath126 0.1 + f450w.d3.dark & 1.43 @xmath126 0.04 & 5.24 @xmath126 0.01 & 1.04 @xmath126 0.11 & 8.0 @xmath126 0.1 & 0.81 @xmath126 0.07 & 7.8 @xmath126 0.1 & 1.22 @xmath126 0.05 & 6.9 @xmath126 0.1 + f450w.d4.dark & 1.30 @xmath126 0.06 & 4.45 @xmath126 0.01 & 1.09 @xmath126 0.06 & 6.9 @xmath126 0.1 & 1.14 @xmath126 0.05 & 6.7 @xmath126 0.1 & 1.23 @xmath126 0.10 & 6.1 @xmath126 0.1 + f450w.d5.dark & 1.68 @xmath126 0.08 & 5.09 @xmath126 0.01 & 0.76 @xmath126 0.06 & 7.6 @xmath126 0.1 & 1.2 @xmath126 0.3 & 7.2 @xmath126 0.1 & 1.49 @xmath126 0.04 & 6.4 @xmath126 0.1 + f450w.d6.dark & 1.46 @xmath126 0.05 & 4.31 @xmath126 0.01 & 0.59 @xmath126 0.04 & 6.9 @xmath126 0.1 & 0.67 @xmath126 0.18 & 6.8 @xmath126 0.1 & 0.77 @xmath126 0.06 & 6.1 @xmath126 0.1 + f450w.d8.dark & 1.55 @xmath126 0.04 & 4.29 @xmath126 0.01 & 0.71 @xmath126 0.07 & 6.9 @xmath126 0.1 & 1.16 @xmath126 0.07 & 6.8 @xmath126 0.1 & 0.81 @xmath126 0.07 & 6.0 @xmath126 0.1 + f450w.d9.dark & 1.97 @xmath126 0.06 & 5.96 @xmath126 0.01 & 1.32 @xmath126 0.03 & 8.8 @xmath126 0.1 & 1.2 @xmath126 0.4 & 8.4 @xmath126 0.1 & 1.57 @xmath126 0.05 & 7.7 @xmath126 0.1 + average image & 0.265 @xmath126 0.009 & 0.626 @xmath126 0.001 & 0.317@xmath126 0.009 & 0.929 @xmath126 0.003 & 0.341@xmath126 0.015 & 0.896@xmath126 0.004 & 0.247 @xmath126 0.011 & 0.822 @xmath126 0.004 + + f606w.d1.dark & 4.75 @xmath126 0.17 & 17.7 @xmath126 0.1 & 5.8 @xmath126 0.3 & 37.2 @xmath126 0.1 & 6.0 @xmath126 0.6 & 37.7 @xmath126 0.2 & 2.4 @xmath126 0.7 & 35.0 @xmath126 0.3 + f606w.d2.dark & 5.9 @xmath126 0.8 & 27.3 @xmath126 0.1 & 8.9 @xmath126 0.8 & 55.2 @xmath126 0.3 & 4.2 @xmath126 0.7 & 56.8 @xmath126 0.2 & 3.6 @xmath126 0.7 & 51.9 @xmath126 0.3 + f606w.d3.dark & 4.9 @xmath126 0.4 & 21.2 @xmath126 0.1 & 3.8 @xmath126 0.4 & 43.0 @xmath126 0.1 & 4.7 @xmath126 1.7 & 42.5 @xmath126 0.5 & 6.0 @xmath126 0.8 & 44.8 @xmath126 0.3 + f606w.d4.dark & 2.49 @xmath126 0.22 & 13.4 @xmath126 0.1 & 1.9 @xmath126 0.4 & 26.5 @xmath126 0.1 & 4.4 @xmath126 0.5 & 26.1 @xmath126 0.2 & 4.2 @xmath126 0.4 & 24.2 @xmath126 0.1 + f606w.d5.dark & 4.69 @xmath126 0.17 & 9.3 @xmath126 0.1 & 4.0 @xmath126 0.2 & 18.0 @xmath126 0.1 & 4.4 @xmath126 0.3 & 19.1 @xmath126 0.1 & 3.2 @xmath126 0.4 & 16.9 @xmath126 0.2 + f606w.d6.dark & 2.41 @xmath126 0.12 & 7.6 @xmath126 0.1 & 4.2 @xmath126 0.2 & 16.6 @xmath126 0.1 & 4.1 @xmath126 0.2 & 17.6 @xmath126 0.4 & 3.8 @xmath126 0.4 & 15.1 @xmath126 0.2 + f606w.d7.dark & 2.9 @xmath126 0.4 & 9.1 @xmath126 0.1 & 4.4 @xmath126 0.3 & 18.6 @xmath126 0.1 & 4.5 @xmath126 0.3 & 19.2 @xmath126 0.1 & 3.8 @xmath126 0.7 & 16.8 @xmath126 0.3 + f606w.d8.dark & 3.48 @xmath126 0.16 & 12.6 @xmath126 0.1 & 4.9 @xmath126 0.5 & 25.8 @xmath126 0.2 & 5.5 @xmath126 1.0 & 26.1 @xmath126 0.3 & 2.5 @xmath126 1.3 & 24.3 @xmath126 0.4 + f606w.d9.dark & 5.3 @xmath126 0.3 & 20.6 @xmath126 0.1 & 4.9 @xmath126 0.6 & 37.6 @xmath126 0.2 & 4.4 @xmath126 0.4 & 39.2 @xmath126 0.1 & 4.8 @xmath126 0.4 & 34.1 @xmath126 0.2 + f606w.d10.dark & 4.41 @xmath126 0.20 & 15.0 @xmath126 0.1 & 6.2 @xmath126 0.3 & 30.5 @xmath126 0.1 & 6.4 @xmath126 0.8 & 31.7 @xmath126 0.2 & 2.6 @xmath126 0.7 & 28.7 @xmath126 0.3 + f606w.d11.dark & 5.6 @xmath126 0.4 & 15.4 @xmath126 0.1 & 3.4 @xmath126 1.0 & 31.8 @xmath126 0.3 & 5.3 @xmath126 0.4 & 33.1 @xmath126 0.1 & 2.3 @xmath126 0.5 & 29.2 @xmath126 0.2 + average image & 0.557@xmath126 0.021 & 1.41 @xmath126 0.01 & 1.38 @xmath126 0.15 & 2.81 @xmath126 0.02 & 1.72 @xmath126 0.10 & 2.71 @xmath126 0.02 & 1.4 @xmath126 0.3 & 2.43 @xmath126 0.09 + + f814w.d1.dark & 2.34 @xmath126 0.08 & 7.99 @xmath126 0.01 & 0.28 @xmath126 0.12 & 15.1 @xmath126 0.1 & 1.3 @xmath126 0.3 & 15.9 @xmath126 0.1 & 2.02 @xmath126 0.23 & 14.8 @xmath126 0.1 + f814w.d2.dark & 1.51 @xmath126 0.09 & 7.64 @xmath126 0.01 & 1.26 @xmath126 0.13 & 13.3 @xmath126 0.1 & 3.6 @xmath126 0.5 & 26.2 @xmath126 0.1 & 2.09 @xmath126 0.17 & 13.2 @xmath126 0.1 + f814w.d3.dark & 2.04 @xmath126 0.04 & 6.10 @xmath126 0.01 & 1.43 @xmath126 0.09 & 11.9 @xmath126 0.1 & 1.54 @xmath126 0.22 & 12.0 @xmath126 0.1 & 1.00 @xmath126 0.09 & 10.6 @xmath126 0.1 + f814w.d4.dark & 2.14 @xmath126 0.05 & 6.91 @xmath126 0.01 & 1.48 @xmath126 0.08 & 13.3 @xmath126 0.1 & 1.54 @xmath126 0.11 & 13.9 @xmath126 0.1 & 0.9 @xmath126 0.4 & 13.9 @xmath126 0.1 + f814w.d5.dark & 2.1 @xmath126 0.3 & 7.02 @xmath126 0.01 & 1.4 @xmath126 0.5 & 13.1 @xmath126 0.1 & 1.37 @xmath126 0.16 & 14.9 @xmath126 0.1 & 1.6 @xmath126 0.3 & 13.2 @xmath126 0.1 + f814w.d6.dark & 1.79 @xmath126 0.06 & 6.67 @xmath126 0.01 & 2.12 @xmath126 0.12 & 12.5 @xmath126 0.1 & 2.83 @xmath126 0.15 & 13.2 @xmath126 0.1 & 1.95 @xmath126 0.12 & 12.9 @xmath126 0.1 + f814w.d8.dark & 5.2 @xmath126 0.5 & 20.81 @xmath126 0.01 & 4.27 @xmath126 0.15 & 27.7 @xmath126 0.1 & 3.2 @xmath126 0.3 & 28.6 @xmath126 0.1 & 2.3 @xmath126 0.3 & 25.4 @xmath126 0.1 + f814w.d9.dark & 2.61 @xmath126 0.18 & 6.78 @xmath126 0.01 & 2.13 @xmath126 0.12 & 12.3 @xmath126 0.1 & 2.50 @xmath126 0.06 & 13.0 @xmath126 0.1 & 2.04 @xmath126 0.07 & 12.3 @xmath126 0.1 + average image & 0.397 @xmath126 0.022 & 1.084 @xmath126 0.002 & 0.54 @xmath126 0.03 & 1.787 @xmath126 0.007 & 0.484@xmath1260.022 & 2.140 @xmath1260.006 & 0.43 @xmath126 0.03 & 1.763@xmath126 0.010 + + pc1 & 0.35 @xmath126 0.16 & 0.215 @xmath126 0.007 & 3.1 @xmath126 0.9 & 0.336 @xmath126 0.020 & 0.48 @xmath126 0.10 & 0.314 @xmath126 0.006 + wf2 & 0.347 @xmath126 0.008 & 0.105 @xmath126 0.005 & 2.26 @xmath126 0.13 & 0.331 @xmath126 0.018 & 0.83 @xmath126 0.03 & 0.176 @xmath126 0.012 + wf3 & 0.468 @xmath126 0.015 & 0.094 @xmath126 0.006 & 2.72 @xmath126 0.20 & 0.36 @xmath126 0.03 & 0.52 @xmath126 0.05 & 0.218 @xmath126 0.013 + wf4 & 0.320 @xmath126 0.008 & 0.121 @xmath126 0.005 & 3.0 @xmath126 0.5 & 0.16 @xmath126 0.05 & 0.74 @xmath126 0.04 & 0.139 @xmath126 0.021 + sbf - measured @xmath127 & 0.37 @xmath126 0.04 & & 2.8 @xmath126 0.3 & & 0.64 @xmath126 0.03 & + + pc1 & 0.26 @xmath126 0.05 & 0.211 @xmath126 0.006 & 0.91 @xmath126 0.11 & 0.370 @xmath126 0.011 & 0.42 @xmath126 0.12 & 0.297 @xmath126 0.012 + wf2 & 0.214 @xmath126 0.010 & 0.102 @xmath126 0.004 & 1.04 @xmath126 0.15 & 0.342 @xmath126 0.021 & 0.36 @xmath126 0.03 & 0.179 @xmath126 0.011 + wf3 & 0.242 @xmath126 0.019 & 0.099 @xmath126 0.011 & 1.40 @xmath126 0.10 & 0.32 @xmath126 0.03 & 0.23 @xmath126 0.03 & 0.251 @xmath126 0.013 + wf4 & 0.115 @xmath126 0.012 & 0.132 @xmath126 0.004 & 1.2 @xmath126 0.3 & 0.22 @xmath126 0.04 & 0.24 @xmath126 0.03 & 0.187 @xmath126 0.013 + sbf - measured @xmath127 & 0.208 @xmath126 0.014 & & 1.13 @xmath126 0.09 & & 0.31 @xmath126 0.03 & + lcc f450w.d1.dark & 5.2 & 5.48 @xmath126 0.02 + f450w.d2.dark & 11.5 & 11.25 @xmath126 0.01 + f450w.d3.dark & 7.8 & 8.02 @xmath126 0.03 + f450w.d4.dark & 6.9 & 6.96 @xmath126 0.02 + f450w.d5.dark & 7.3 & 7.62 @xmath126 0.02 + f450w.d6.dark & 6.7 & 6.99 @xmath126 0.01 + f450w.d8.dark & 6.6 & 6.91 @xmath126 0.02 + f450w.d9.dark & 8.3 & 8.81 @xmath126 0.01 +

the physical mechanisms to set the mass ratios of binary stars is of fundamental astrophysical importance , since more than half of main - sequence , pre - main sequence , and even protostars having masses comparable to the sun are members of binary systems @xcite . binary stars with the primary masses comparable to the solar mass exhibit a wide , essentially flat distribution of the mass ratios ( @xmath12 ) from @xmath130.1 to 1 @xcite . the physical processes to reproduce such a flat distribution of the binary mass ratios are , however , still controversial . protostellar binary systems , precursors of main - sequence binary systems , are often surrounded by disks of molecular gas and dusts , `` circumbinary disks '' ( hereafter cbds ) @xcite . the radii of these cbds range from @xmath9100 to @xmath7500 au , and these cbds often show inner emission depressions and ringlike structures . mass accretion from these cbds onto the central protostellar binaries has been considered as one of the key physical processes to set the binary mass ratios . recent high - resolution observations of the cbds have indeed detected gas flows from cbds to the inner protostellar binaries , while the spatial resolutions are not high enough to disentangle the accretions onto the primaries and secondaries @xcite . theoretically , bate & bonnell ( 1997 ) and bate ( 2000 ) have shown that from their smoothed particle hydrodynamic ( sph ) simulations the majority of circumbinary materials accrete onto the secondary . high - resolution grid - based simulations by ochi et al . ( 2005 ) and hanawa et al . ( 2010 ) have shown the opposite results , that is , the primary accretes more than the secondary . ochi et al . ( 2005 ) and hanawa et al . ( 2010 ) argued that high - resolution simulations to sufficiently resolve the gas motions near the l2 and l3 lagrangian points are essential to properly trace the mass accretion , since accreting gas flows into the binary through these lagrangian points . on the other hand , latest sph simulations have demonstrated that mass accretion in cbds is a function of gas temperatures , and in the case of the high temperature the primary accretes more than the secondary , and vice versa @xcite . young et al . ( 2015 ) argued that the difference between the above results can simply be explained by the difference of the adopted gas temperatures in the models . it has been still controversial to construct a consistent , unified model of mass accretion from cbds onto protostellar binaries . cbds in protostellar binary systems are considered to be embedded in larger - scale protostellar envelopes , which can replenish the central cbds with fresh materials @xcite . since the observed masses of cbds around protostellar binaries ( @xmath90.1 @xmath14 ) are much smaller than the masses of the central protostellar binaries @xcite , further supply of materials from the protostellar envelopes to the cbds is essential to significantly change the mass and mass ratios of the protostellar binaries . if a protostellar envelope keeps supplying materials to the central cbd , further growth of the mass and the mass ratio of the protostellar binary will be expected . on the other hand , if the most of the envelope materials are being dissipated , the present mass and mass ratio of the protostellar binary will be close to the final values . such dissipations of protostellar envelopes through the interactions with the associated outflows or stellar winds have been investigated observationally @xcite . thus , in addition to physics of accretion from cbds onto the protostellar binaries , mass replenishment from the surrounding protostellar envelopes , and the connection between protostellar envelopes and cbds , must be taken into account . in the present paper , we focus on a protostellar envelope harboring the protostellar binary system l1551 ne . l1551 ne is a young class i protostellar binary ( @xmath15=91 k , @xmath16=4.2 @xmath17 ; froebrich 2005 ) located to @xmath72@xmath185 northeast of another , brightest protostellar binary l1551 irs 5 in the l1551 region @xcite . the protostellar binary consists of the northwestern source named `` source a '' , and the southeastern source `` source b '' , with the projected separation of @xmath770 au at a position angle of 300@xmath19 @xcite . our previous sma observations of l1551 ne have identified a @xmath6300 au scale cbd in keplerian rotation with a central stellar mass of 0.8 @xmath14 @xcite , plus a possible outer infalling component @xcite . our subsequent alma cycle 0 observation of l1551 ne @xcite , at a spatial resolution that is @xmath71.6 times higher ( in beam area ) and a sensitivity that is @xmath76 times better ( in brightness temperature ) than those attained in our previous sma observations , unveiled substructures of the cbd in the 0.9-mm dust - continuum emission . the revealed substructures are consistent with our 3-d adaptive mesh refinement ( amr ) hydrodynamic simulation @xcite for the presence of two spiral arms driven by gravitational torques from the central binary system . the alma data of l1551 ne in the c@xmath0o ( 32 ) line also exhibits the deviations from the keplerian motion in the cbd , consistent with our amr simulation that gravitational torques impart angular momentum along the spiral arms ( driving material outwards ) and extract angular momentum between the spiral arms ( driving infall ) . our theoretical model which reproduces the observed features of the cbd around l1551 ne predicts that secondary accretes more than the primary . since our series of the sma and alma observations of l1551 ne have revealed the cbd + protostellar binary system , the next question is the connection from the protostellar envelope to the cbd , and the mass replenishment from the envelope to the cbd . we conducted single - dish mapping observations of the protostellar envelope surrounding the protostellar binary system of l1551 ne in the c@xmath0o ( @xmath1=32 ) , @xmath2co ( @xmath1=32 ) , cs ( @xmath1=76 ) , and the so ( @xmath3=7@xmath46@xmath5 ) lines with the atacama submillimeter telescope experiment ( aste ) 10-m telescope . the aste mapping data in the c@xmath0o line are combined with our previous sma data in the same line , which have a larger field of view and a shorter spacing than those of the alma data and thus more suitable to study the connection between the extended protostellar envelope and the cbd . in this paper we report these aste+sma results of the protostellar envelope in l1551 ne . the structure of this paper is as follows . in section 2 , our new aste observations , as well as the process to combine the aste mapping data with the sma data in the c@xmath0o line , are described . in section 3 , the results of the aste mapping observations and the combination of the aste + sma data are presented . in section 4 , gas motions and physical properties of the identified gas components are estimated . in section 5 , we discuss implications of the observed features in the context of growth of protostellar binaries , and concise summary is written in section 6 . we performed observations of l1551 ne in the c@xmath0o ( @xmath1=32 ; 329.3305453 ghz ) , @xmath2co ( @xmath1=32 ; 330.587965 ghz ) , so ( @xmath3=7@xmath46@xmath5 ; 340.71416 ghz ) , and the cs ( @xmath1=76 ; 342.882857 ghz ) lines with the aste 10-m telescope on 2013 september 23 - 26 and october 4 . remote observations were performed from the aste operation room of naoj at mitaka , japan , using the network observation system n - cosmos3 developed by naoj @xcite . a cartridge - type , double - sideband 350 ghz receiver with an if frequency range of 4.5 to 7 ghz ( cats345 ) mounted on aste @xcite was used , and all the four lines were observed simultaneously , except that on september 23 only the so and cs lines were observed due to the mistake of the instrumental setting . we used only the data with the dsb system noise temperature ranging @xmath7200 - 800 k , and the data with the higher noise temperatures are excluded . the telescope pointing was checked every @xmath71.5 - 2 hours by five - point co ( @xmath1 = 32 ) observations of a late - type star , nml - tau , and was found to be better than @xmath72@xmath11 . as a standard source we also observed orion kl @xcite and l1551 irs 5 @xcite , and confirmed that the relative intensity was consistent within @xmath730@xmath20 with a main beam efficiency of 0.6 . no further correction for the sideband ratio was performed . hereafter we show the observed line intensities in the unit of @xmath21 . the observations consisted of two parts ; one - point , deep observations toward the center of l1551 ne ( @xmath22,@xmath23)= ( 04@xmath2431@xmath2544@xmath2647 , 180832@xmath272 ) , which is the field center of our previous sma observations and matches approximately the positions of the protostellar binary @xcite , and mapping observations around l1551 ne . both observations were conducted in the position - switching mode . the so and cs line data toward the center of l1551 ne taken on september 23 , when the other line data were not taken due to the mistake of the instrumental setting , are included in the deep integrations . the resultant total on - source integration times of the so and cs data toward the center of l1551 ne is 1620 sec , while those of the @xmath2co and c@xmath0o data 1220 sec . the rms noise levels of the @xmath2co , c@xmath0o , so , and the cs spectra toward the center of l1551 ne are 0.055 k , 0.062 k , 0.025 k , and 0.029 k , respectively . the mapping observations were conducted with a grid spacing of 10@xmath11 ( nyquist sampling ) , and a typical on - source integration time per point of @xmath780 seconds . the map center is set to be the position of the deep integrations . the spatial and spectral resolutions and the noise level of the mapping observations in the c@xmath0o line are summarized in table 1 . we combined the aste image cube in the c@xmath0o line with the interferometric c@xmath0o data of l1551 ne taken with the sma in its subcompact and compact configurations @xcite , adopting the method described by takakuwa et al . ( 2007b ) . details of the sma observations are described by takakuwa et al . the conversion factor from @xmath21 ( k ) to s ( jy beam@xmath8 ) was derived to be 48.5 as @xmath28 where @xmath29 is the boltzmann constant , @xmath30 is the wavelength , and @xmath31 is the solid angle of the aste beam (= 23@xmath11 ) . the spatial and spectral resolutions and the noise level of the combined sma+aste image cube in the c@xmath0o line are summarized in table 1 . figure [ astesp ] shows the observed aste spectra of the c@xmath0o ( @xmath1=32 ) , @xmath2co ( @xmath1=32 ) , cs ( @xmath1=76 ) , and the so ( @xmath3=7@xmath46@xmath5 ) lines toward the protobinary of l1551 ne . the c@xmath0o spectrum appears to consist of narrow and wide components . two - component gaussian fitting to the c@xmath0o spectrum shows that the peak brightness temperature , line width , and the centroid velocity of the narrow component are 2.4 k , 0.72 km s@xmath8 , and 6.68 km s@xmath8 , and those of the wide component 1.1 k , 2.24 km s@xmath8 , and 6.99 km s@xmath8 , respectively ( light lines in figure [ astesp ] ) . previous cso observations also found that the c@xmath0o ( 32 ) spectrum toward l1551 ne can be decomposed of two gaussian components with similar line widths and centroid velocities @xcite , in particular , the centroid velocity of the broad component is similar to the centroid velocity of the cbd derived from our previous sma observations (= 6.9 km s@xmath8 ) @xcite . as shown in the next subsection , the broad c@xmath0o component is present only toward the protobinary position . these results indicate that the broader component in the aste c@xmath0o spectrum likely originates from the central cbd . on the other hand , the narrower c@xmath0o component traces a distinct component at a slightly blueshifted ( @xmath70.2 km s@xmath8 ) centroid velocity . as shown below , this narrow component traces the extended envelope component around l1551 ne . hereafter in this paper , @xmath32 = 6.7 km s@xmath8 is adopted as the systemic velocity of the extended envelope component , while @xmath32 = 6.9 km s@xmath8 as the systemic velocity of the cbd in l1551 ne . the @xmath2co ( @xmath1=32 ) spectrum shows an absorption dip at around the centroid velocity of the narrow component of the c@xmath0o spectrum , and a broad redshifted wing up to @xmath32 @xmath713 km s@xmath8 . while a part of the redshifted wing in the @xmath2co spectrum likely originates from the cbd as in the case of the c@xmath0o spectrum , mapping observations of the @xmath2co line demonstrate that the redshifted @xmath2co emission is most likely arising from the outflow component driven from l1551 irs 5 , as will be discussed below . with a deeper integration a much better cs ( @xmath1=76 ) spectrum of l1551 ne than our previous one @xcite is obtained . the cs spectrum shows a non - gaussian , flat - top spectral shape . whereas the so ( @xmath3=7@xmath46@xmath5 ) line is weak compared to the other lines , the peak intensity ( @xmath70.14 k ) is detected above 5.5@xmath33 . figure [ astemom0 ] shows total integrated intensity maps of the four molecular lines observed with aste . the c@xmath0o ( 32 ) emission exhibits condensed and diffuse components , the former of which is centered on the protobinary position . two - component , two - dimensional gaussian fitting to the c@xmath0o total integrated map shows that the c@xmath0o emission can be decomposed of a central component with a size of @xmath75500 au @xmath34 2600 au ( p.a.=-46@xmath19 ) and a diffuse component with a size of @xmath721000 au . emission peaks toward the protostellar position are also seen in the cs ( 76 ) and so ( 7@xmath46@xmath5 ) lines . on the other hand , the @xmath2co ( 32 ) peak emission does not coincide with the protostellar position but is offset at @xmath720@xmath11 west from the protostar . the @xmath2co emission shows an elongated feature toward the southwest direction , which is also evident in the c@xmath0o and cs emission . figure [ aste13co ] shows the aste velocity channel maps of the @xmath2co line at a velocity interval of 0.455 km s@xmath8 . in the blueshifted velocity range ( @xmath32=4.62 5.99 km s@xmath8 ) , an emission component to the northwest of the protostar and that centered on the protostellar position are seen . around the systemic velocity ( 6.45 6.90 km s@xmath8 ) the @xmath2co line shows less intensity contrast , suggesting that the @xmath2co emission traces the overall cloud component . in the redshifted velocity of 7.36 9.18 km s@xmath8 , two emission components , one located to @xmath720@xmath11 west of the protostar , and the other to @xmath740@xmath11 southwest , while the southwestern component diminishes at velocities higher than 9.63 km s@xmath8 , the western component is seen until @xmath712.36 km s@xmath8 . the location and velocity of this high - velocity redshifted component is consistent with those of a redshifted outflow component seen in the @xmath35co ( 10 ; 21 ; 32 ) lines @xcite and the cs ( 21 ; 32 ) lines @xcite driven from the neighboring protostellar source , l1551 irs 5 . thus , the high - velocity redshifted component seen in the @xmath2co line most likely traces the redshifted outflow component from l1551 irs 5 . figure [ astec18o ] shows the aste velocity channel maps of the c@xmath0o line at the same velocity bin as that of the @xmath2co velocity channel maps . the c@xmath0o emission component associated with the protostar is seen from @xmath32=5.53 km s@xmath8 . the c@xmath0o emission component exhibits an elongated feature at 6.44 6.90 km s@xmath8 along the northwest to the southeast direction , which is approximately consistent with the major axis of the central disk . while the c@xmath0o emission peak at 6.44 km s@xmath8 is located to the northeast of the protostar , the emission location is systematically shifted from northeast to southwest from the blueshifted to redshifted velocities ( 6.44 km s@xmath8 to 7.81 km s@xmath8 ) . these results indicate that there is a velocity gradient in the c@xmath0o emission along the direction of the minor axis of the central disk . these c@xmath0o emission components most likely trace the protostellar envelope surrounding l1551 ne ( hereafter `` env '' ) . in the redshifted velocities from 7.35 km s@xmath8 , another emission component to the southwest of the protostar becomes evident . this component appears to trace the same southwestern gas component seen in the @xmath2co emission . in even higher redshifted velocities ( 8.72 10.54 km s@xmath8 ) , the c@xmath0o counterpart of the western redshifted component found in the @xmath2co emission is also seen , although the highest velocity end of the c@xmath0o counterpart is lower than that of the @xmath2co emission . figure [ astecs ] shows the same velocity channel maps as those of figures [ aste13co ] and [ astec18o ] but for the cs ( 76 ) line . while the cs emission traces a gas component associated with the protostar in the velocity range of 5.54 7.81 km s@xmath8 , the most prominent cs emission component is the redshifted ( 8.27 9.18 km s@xmath8 ) component to the southwest of the protostar . this southwestern redshifted component is also seen in the @xmath2co and c@xmath0o emission . previous bima and nma observations of the cs ( 21 ; 32 ) lines have also found the same cs component , where the cs abundance is enhanced by a factor of a few @xcite . this cs component is considered to be shock - excited molecular gas of the outflow driven from l1551 irs 5 @xcite . the detection of the submillimeter cs ( 76 ) emission with the present aste observations , which traces warm ( @xmath10 60 k ) molecular gas @xcite , further supports this interpretation . hereafter we call this redshifted component to the southwest of the protostar `` red1 '' , and the highest - velocity redshifted component seen in the @xmath2co and c@xmath0o lines to the west of the protostar `` red2 '' . in figure [ colli ] , we compare spatial distributions of env as seen in the c@xmath0o emission , red1 in the cs emission , and red2 in the @xmath2co emission . line profiles in the peak positions of red1 and red2 , as well as those 20@xmath11 east of the protostar , are also shown ( note that the line profiles of env are shown in figure [ astesp ] . ) . the two outflow components of red1 and red2 are located to the western side of env . toward red1 , the @xmath2co line profile shows an intense redshifted peak plus a red wing up to @xmath32 @xmath712 km s@xmath8 . the c@xmath0o line profile exhibits redshifted emission `` plateau '' up to @xmath32 @xmath710 km s@xmath8 , and the cs line profile shows the emission only in the redshifted velocity range . similar @xmath2co redshifted wing and c@xmath0o emission plateau are also seen toward red2 , where a weak redshifted cs emission is also present . on the other hand , toward the 20@xmath11 east of the protostar , such a c@xmath0o emission plateau is not seen and the @xmath2co redshifted wing is much less prominent , and the submillimeter cs emission is not detected . these results imply that the outflow streams of red1 and red2 are terminated at the western side of env . one possible interpretation for this termination is the collision between the outflow components and env . in figure [ collipv ] , we show the position - velocity ( p - v ) diagrams of the @xmath2co , c@xmath0o , and the cs lines along the cuts passing through the protostellar position and the peak positions of red1 ( p.a . = 45@xmath19 ) and red2 ( p.a . = 90@xmath19 ) . in the c@xmath0o p - vs the envelope component is seen as labeled . to the west of env the @xmath2co , c@xmath0o , and the cs emission components arising from red1 , and the @xmath2co and c@xmath0o emission components from red2 , can be identified . there appears velocity gradients both in red1 and red2 as delineated by dashed lines , and the velocities of red1 and red2 become closer to the systemic velocity as the positions become closer to env . these results suggest that the flow velocities of red1 and red2 are decelerated as they become close to env , presumably due to the collision with env . figure [ astepvs ] shows aste p - v diagrams of the c@xmath0o and @xmath2co lines along the major and minor axes of the central cbd passing through the protostellar position . in the c@xmath0o p - vs , the line broadening at the protostellar position arising from the cbd is clearly seen , and the velocity gradient originated from the keplerian rotation ( blue curves in figure [ astepvs ] ) can be identified along the major axis . as discussed in the c@xmath0o line profile toward the protostellar position ( figure [ astesp ] ) , the centroid velocity of the cbd component and that of the envelope component are offset by @xmath70.2 km s@xmath8 ( two vertical dashed lines in figure [ astepvs ] ) , and the c@xmath0o p - vs exhibit this offset clearly . furthermore , along the major axis the envelope component as seen in the c@xmath0o emission does not show any noticeable velocity gradient , and the keplerian rotation of the central disk is not continuous to the outer envelope . on the other hand , along the minor axis the c@xmath0o envelope shows a clear velocity gradient along the northeast ( blue ) to southwest ( red ) direction ( tilted black dashed line in figure [ astepvs ] ) . since the associated nir jets are blueshifted to the southwest and redshifted to the northeast @xcite , the southwestern side of the disk and envelope must be far - side and the northeastern side near - side , assuming that the midplane of the disk and envelope is perpendicular to the axis of the jets . correspondingly , the southwestern redshifted component would then be located on the far side and the northeastern blueshifted component on the near side , suggesting the outward radial motion on the plane . in the @xmath2co p - v diagram along the minor axis , both red1 and red2 and their decelerated velocity structures ( red and pink dashed lines ) are identified . these results indicate that the envelope is being dissipated due to the interaction with the redshifted outflowing gas driven from l1551 irs 5 . we will further discuss this dispersing motion of the envelope with the combined sma+aste images of the c@xmath0o emission below . figure [ mom0s ] compares moment 0 maps of the aste , sma , and the combined sma+aste image cubes of the c@xmath0o ( 32 ) line in l1551 ne . while the aste image exhibits an extended envelope structure as described in the last subsection , the sma image shows the central smaller - scale ( @xmath71000 au @xmath34 800 au ) structure with its position angle ( @xmath7167@xmath19 ) consistent with that of the keplerian cbd @xcite . combining these two images shows that the central structure found by the sma observations is embedded in the more extended envelope . figure [ combch ] shows velocity channel maps of the sma+aste image cube of the c@xmath0o ( 32 ) line . in the highly blueshifted ( @xmath32 = 4.2 6.1 km s@xmath8 ) and redshifted ( 7.5 9.2 km s@xmath8 ) velocities , compact c@xmath0o emission located to the north and south of the protostellar binary are seen , respectively . these compact components trace the keplerian cbd as we already discussed in our previous papers @xcite . the extended envelope component appears in the lower velocity range ( 6.1 8.2 km s@xmath8 ) . in the blueshifted velocity range of 6.3 6.6 km s@xmath8 , the extended c@xmath0o emission is located predominantly to the eastern side of the protostellar binary , whereas in the redshifted velocity range of 7.5 8.2 km s@xmath8 to the western side . thus , there is a velocity gradient along the east ( blue ) to west ( red ) direction in the envelope , which is also identified in the aste only c@xmath0o image cube as discussed in the last subsection . in the lower velocity region ( 6.6 7.4 km s@xmath8 ) the envelope emission extends over the entire region . to highlight these velocity structures in the sma+aste image cube , we integrated the velocity channel maps over the high - velocity , middle - velocity , and the low - velocity blueshifted and redshifted regions , and the resultant images are shown in figure [ comb3 ] . in the high - velocity region , compact blueshifted and redshifted c@xmath0o emission located to the north and south of the protostellar binary , which originates from the keplerian cbd , are seen . in the middle - velocity region , as well as the central disk component , an extended envelope component , with the blueshifted emission located predominantly to the east and the redshifted emission to the west , are seen . as we discussed in the last subsection , one interpretation of this velocity structure is the dispersing gas motion in the flattened envelope . in the low - velocity region both the blueshifted and redshifted emission spread over almost the entire field of view . we note , however , that the blueshifted and redshifted emission peaks in the central compact emission are located slightly to the west and east of the protostellar binary , respectively . the sense of this velocity structure is opposite to that of the extended envelope seen in the middle velocity range . this velocity structure is also seen in the sma only image cube and the centroid velocity of this velocity feature matches with that of the cbd . as discussed in our previous paper @xcite , one interpretation for this velocity structure is an infalling gas motion in the inner envelope toward the central cbd . such coexistence of inner infall and outer dispersing motions has also been identified in an envelope surrounding a t - tauri star dg tauri @xcite . figure [ combpv ] ( left ) shows the p - v diagrams of the combined sma+aste image cube of the c@xmath0o line along the major ( upper panel ) and minor axes ( lower ) of the cbd . for comparison , keplerian rotation curves derived from the sma data @xcite are drawn in black lines , and the centroid velocity of the cbd and the envelope in vertical dashed lines . in the p - v along the major axis , the central compact component , cbd , exhibits a velocity gradient consistent with the keplerian rotation . on the other hand , the extended envelope component does not show any clear velocity gradient along the major axis , and the offset of the centroid velocity of the envelope from that of the cbd is clearly seen . these results indicate that the keplerian rotation of the central cbd is discontinuous to the outer envelope . in the p - v diagram along the minor axis , the central component exhibits a wide velocity width but no velocity gradient , as expected from the keplerian rotation . on the other hand , along the minor axis the emission ridges in the envelope component are located predominantly to the northeast in the blueshifted velocities , and to the southwest to the redshifted velocities , as seen in the middle velocity ranges of figure [ comb3 ] . this velocity structure can be interpreted as the dispersing gas motion in the flattened envelope . we will interpret these detected velocity structures of the combined sma+aste c@xmath0o image cube with our toy model below . the combined sma+aste c@xmath0o image cube reveals a sign of an outer dispersing envelope motion , possible inner infalling motion , and keplerian rotation in the central cbd . to interpret these results quantitatively , we constructed a toy model of the envelope and disk assuming that the disk and envelope are geometrically thin and co - planar . our model is essentially the same as that by takakuwa et al . ( 2013 ) , which reproduces the velocity structures of the sma only image cube , but now the outer dispersing envelope component is added to reproduce the velocity structure of the extended envelope component traced by aste . as a model moment 0 map , we adopted a combination of two 2-dimensional gaussians , which are derived from the gaussian fitting to the observed sma+aste moment 0 map ( figure [ mom0s ] bottom - right ) . the velocity structures in the inner component traced by the sma are described as , @xmath36 where @xmath37 is the radius , @xmath38 the gravitational constant , @xmath39 the mass of the protostellar binary , @xmath40 and @xmath41 are the rotational and radial velocities , respectively , @xmath42 the outermost radius of the keplerian rotation , and @xmath43 an adopted constant infall velocity beyond @xmath42 . as discussed by takakuwa et al . ( 2013 ) , we adopt @xmath39 = 0.8 @xmath14 , @xmath42 = 300 au , and @xmath43 = -0.6 km s@xmath8 , and the disk inclination angle @xmath44 . the velocity structure of the outer envelope traced by aste is expressed as , @xmath45 where @xmath46 and @xmath47 are the innermost radius of the dispersing envelope and the dispersing velocity , respectively . @xmath46 = 700 au is adopted , inferred from the comparison between the sma only and sma+aste moment 0 maps . @xmath47 can be roughly estimated from the velocity gradient seen in the combined p - v diagram along the minor axis as shown in figure [ combpv ] . the velocity shift of the emission ridge of the envelope across the minor axis is @xmath70.5 km s@xmath8 , centered on the centroid velocity of 6.7 km s@xmath8 , and at an inclination of @xmath48 @xmath7 62@xmath19 the dispersing velocity along the flattened envelope is @xmath47 @xmath7(0.5 km s@xmath8 / 2 ) / sin 62@xmath19 @xmath70.3 km s@xmath8 . note that the centroid velocity of the envelope component is offset by @xmath7 - 0.2 km s@xmath8 from that of the cbd , and in our model this offset of the centroid velocity is incorporated . the internal velocity dispersion @xmath49 = 0.4 km s@xmath8 , which reproduces the velocity channel maps of the central cbd @xcite , is adopted throughout the entire components . figure [ modelch ] shows the velocity channel maps of our toy model . in the high - velocity blueshifted ( 4.6 5.7 km s@xmath8 ) and redshifted ranges ( 7.7 9.2 km s@xmath8 ) the features of the keplerian cbd are seen . in the blueshifted range of 5.9 6.3 km s@xmath8 the envelope component is located to the east of the protostellar binary while in the redshifted range of 7.2 7.5 km s@xmath8 to the west , reflecting the dispersing motion in the envelope . in the lower - velocity range ( 6.4 7.0 km s@xmath8 ) the envelope component is extended both in the eastern and western sides , but the east ( red ) - west ( blue ) velocity gradient arising from the infalling motion in the central component is evident . these characteristics reproduces primary features of the observed sma+aste c@xmath0o velocity channel maps in l1551 ne ( figure [ combch ] , [ comb3 ] ) . figure [ combpv ] ( right ) shows the p - v diagrams of our toy model along the major and minor axes . the model p - vs can reproduce primary features of the observed p - vs , such as the high - velocity component arising from the keplerian disk , and the extended envelope component which shows a velocity gradient along the minor axis reflecting the dispersing motion , but no rotation along the major axis . these results show that the observed sma+aste c@xmath0o image cube in l1551 ne can be understood with three distinct velocity components ; a central keplerian cbd at a radius of @xmath9300 au , a possible infalling region at 300 au @xmath50 @xmath37 @xmath50 500 au , and an outer dispersing envelope with a dispersing velocity of @xmath70.3 km s@xmath8 at @xmath37 @xmath10500 au . our aste mapping observations of the l1551 ne region have revealed three distinct gas components ; @xmath51 , env , red1 , and red2 . env is the component of the protostellar envelope surrounding the protostellar binary system l1551 ne , and exhibits the dispersing gas motion as discussed above . red1 and 2 are most likely the redshifted outflow components driven from l1551 irs 5 , which are interacting with env . to discuss the interaction between env and red1 and 2 quantitatively , physical parameters of these components were estimated as follows . first , the moment 0 map of each component was made in the relevant tracers and the velocity ranges listed in table [ comphys ] . as shown above , env is most clearly identified in the c@xmath0o ( 32 ) emission , and red2 in the @xmath2co ( 32 ) emission . we adopt the cs ( 76 ) emission to deduce the physical condition of red1 , because in the cs emission red1 can be identified as a distinct gas component most easily , and the submillimeter cs emission must trace higher - temperature gas associated with outflows . we also note that it is not straightforward to separate red2 in the 3-dimensional space from the other lower - velocity components . we here identified red2 as the @xmath2co emission component in the highest redshifted velocity range of 9.47 12.53 km s@xmath8 , since in this velocity range red2 is a distinct , highest redshifted component seen primarily in the @xmath2co emission . after making the moment 0 maps of these distinct components , two - dimensional gaussian fittings to the moment 0 maps were performed to derive the central positions , beam - deconvolved sizes along the major and minor axes , and the total flux . for env a single two - dimensional gaussian fitting to the moment 0 map did not provide a satisfactory fit . thus , for env two - component , two - dimensional gaussian fitting was performed , and the position was defined as the peak position of the central gaussian component , while the size was defined as the size of the outer gaussian component . the line widths of these components ( @xmath52 ) were derived from the fwhm values of the relevant spectra toward the central positions . projected flow velocities of red1 and red2 , @xmath53 , were defined as the mean velocity at the center of the components with respect to the velocity of env (= 6.7 km s@xmath8 ) . masses of the components ( @xmath54 ) were derived from the total fluxes , on the assumption of the lte condition , optically - thin emission , and the molecular abundances of @xmath55 = 1.7 @xmath34 10@xmath56 @xcite , @xmath57 = 7.7 @xcite , and @xmath58 = 6.8 @xmath34 10@xmath59 @xcite . since the peak brightness temperature ( @xmath60 ) of the c@xmath0o emission toward the protobinary position exceeds @xmath613.4 k , the excitation temperature ( @xmath62 ) must exceed @xmath109.0 k as @xmath63 must be higher than @xmath64 . on the other hand , to excite the submillimeter cs ( 76 ) emission a gas temperature as high as @xmath760 k is required @xcite . thus , @xmath65 = 10 - 60 k is assumed as a probable range of @xmath65 . virial masses of the components ( @xmath66 ) were derived as ; @xmath67 where @xmath68 in the above expressions @xmath69 denotes the geometrical mean of the sizes along the major and minor axes of the components , and @xmath70 the effective sound speed of gas . the isotropic velocity dispersion inside the components is assumed . for red1 and 2 , the sizes ( @xmath74000 au ) are as small as the beam size ( @xmath73200 au ) , and thus the line widths derived from the central spectra represent the average velocity dispersions . for env , there is no discernible spatial variation of the c@xmath0o line width as shown in the p - v diagrams ( figures [ collipv ] , [ astepvs ] ) , except for that arising from the central disk . thus , the assumption of the isotropic velocity dispersion inside these components is probably valid . momenta ( @xmath71 ) , internal gas energies ( @xmath72 ) , and kinetic energies ( @xmath73 ) of the components were derived as ; @xmath74 @xmath75 @xmath76 table [ comphys ] summarizes these derived physical parameters . the size ( @xmath720000 au ) and lte mass ( @xmath71 - 3 @xmath14 ) of env are typical of protostellar envelopes @xcite . the virial mass of env is comparable to or slightly larger than the lte mass plus the central protobinary mass . from the comparison between the virial and lte masses it is not straightforward to ascertain whether env is gravitational bound or not . on the other hand , the identification of the dispersing gas motion in env implies that env is gravitationally unbound . the outflow components of red1 and 2 are characterized with their compact sizes ( @xmath94000 au ) and wide line widths ( @xmath712 km s@xmath8 ) . the lte masses of the outflow components are likely lower than the corresponding virial masses , although for red1 only the upper limit of the virial mass is obtained . these results indicate that red1 and 2 are not gravitationally - bound gas condensations , consistent with our interpretation that red1 and 2 are outflowing gas components driven from l1551 irs 5 . the summed flow momentum of red1 and 2 is comparable to that of the momentum of the dispersing motion in env . furthermore , the net internal+kinetic energy of these outflow components is also comparable to the internal+kinetic gas energy of env . these results imply that red1 and 2 can exert significant dynamical impacts on env , and can induce the observed dispersing gas motion of env through the interactions . our new aste observations and combining the aste data with our previous sma data in l1551 ne have revealed that the protostellar envelope around the keplerian cbd is being dispersed . the dispersing envelope does not show any rotating gas motion in contrast with the keplerian rotation in the cbd , and the centroid velocity of the envelope is @xmath70.2 km s@xmath8 blueshifted with respect to that of cbd . these results indicate that the protostellar envelope is kinematically distinct from the central cbd . the outflow components from l1551 irs 5 , as seen most prominently in the submillimeter cs ( 76 ) line ( red1 ) and the high - velocity redshifted @xmath2co ( 32 ) line ( red2 ) , appear to collide and interact with the protostellar envelope . the net momentum , and kinetic and internal energy of these outflow components are comparable to those of env , suggesting that the outflow collisions can enforce significant dynamical impacts on the envelope . a natural interpretation of these results is that the protostellar envelope in l1551 ne is being disrupted through the interaction with the outflows driven from l1551 irs 5 . figure [ scheme ] ( bottom ) shows a schematic , inferred configuration of the l1551 region viewed from the north , in parallel with the scuba 850 @xmath77 dust - continuum image @xcite as a representative projected image onto the sky plane ( top ) . in the scuba image a torus - like continuum feature connecting the envelopes around l1551 irs 5 and ne is seen as delineated with a dashed green curve . this feature is drawn schematically in the bottom as the northernmost feature . since the envelope around l1551 ne is being collided with the redshifted outflow components driven from l1551 irs 5 ( red1 and 2 ) , the location of l1551 ne must be backward with respect to the location of l1551 irs 5 along the line of sight . as already described the western side of the flattened envelope around l1551 ne is far - side and the eastern side near - side . thus the blueshifted emission to the east and the redshifted emission to the west are interpreted as the dispersing gas motion in the flattened envelope . our aste single - dish observations of the @xmath2co ( 32 ) , c@xmath0o ( 32 ) , cs ( 76 ) , and so ( 7@xmath46@xmath5 ) lines , as well as sma and alma observations of the @xmath2co ( 32 ) and c@xmath0o ( 32 ) lines @xcite , do not find any evidence for molecular outflows driven from l1551 ne itself . on the other hand , previous nir observations of l1551 ne found collimated [ fe ii ] jets driven from source a as well as a reflection nebula with its apex close to the position of source b @xcite . molecular outflows driven from l1551 ne have also been reported in the @xmath35co ( 32 ; 21 ; 10 ) lines @xcite , although it is not straightforward to disentangle the outflows from l1551 irs 5 and ne because of the close overlaps of the two outflows . our aste observations show that interactions with the powerful l1551 irs 5 outflows have greater impact on the disruption of the envelope than the outflow driven from l1551 ne itself . this is in contrast with previous findings of interactions between protostellar envelopes and outflows or stellar winds driven from the sources embedded in the envelopes @xcite . yokogawa et al . ( 2003 ) have suggested that the interactions with the outflows driven from l1551 irs 5 triggered the star formation in l1551 ne . we consider , however , that the formation of the protostellar binary system l1551 ne has initiated before the arrival of the irs 5 outflows , because at present the protostellar envelope is kinematically distinct and thus physically detached from the cbd and the protostellar binary system . the similar binary masses ( @xmath70.8 @xmath14 in l1551 ne and @xmath70.5 @xmath14 in irs 5 ) , envelope masses ( @xmath71 @xmath14 ) , and the bolometric temperatures ( @xmath15 @xmath790 k ) indicate that both l1551 irs 5 and ne were born at a similar age . after their birth both l1551 ne and irs 5 grow simultaneously . the angular separation between l1551 irs 5 and ne is @xmath72@xmath185 , and assuming the transverse velocity of the outflows driven from l1551 irs 5 is @xmath710 km s@xmath8 @xcite , the transverse time for the irs 5 outflow to reach the l1551 ne location is @xmath710@xmath78 yr . this time scale is approximately consistent with the estimated age of l1551 ne of 0.65.0@xmath3410@xmath78 yr @xcite . after the outflows from l1551 irs 5 have arrived at the protostellar envelope around l1551 ne , the outflows start disrupting the envelope through the interactions . the possible infalling motion in the inner envelope found with the sma may be a remnant gas motion before the initiation of the interactions with the irs 5 outflows . such a remnant infalling gas motion embedded in the dispersing envelope has also been identified in a t - tauri star dg tauri @xcite . the envelope dispersion via the interactions with the irs 5 outflows will impact the final fates of the protostellar binary system of l1551 ne significantly , as will be discussed in the next subsection . binary stars with the primary masses comparable to that of the sun show a wide range of the binary mass ratios , and the distribution of the mass ratio ( @xmath12 ) is essentially flat from @xmath130.1 to 1 @xcite . physical mechanisms to reproduce such a wide range of binary mass ratios have been controversial , and gas accretion in the cbds onto the protostellar binaries has been discussed as a mechanism to set the final binary mass ratios . smoothed particle hydrodynamic ( sph ) simulations of circumbinary accretion show that the majority of materials accrete onto the secondary , because the secondary orbits further from the center of mass of the binary and thus sweeps more materials in the cbd than the primary @xcite . our 3-d adaptive mesh refinement ( amr ) hydrodynamic simulation @xcite to reproduce the alma results of the cbd in l1551 ne also shows that the secondary ( source b ) accretes more than the primary ( source a ) @xcite . on the other hand , high - resolution grid - based simulations by ochi et al . ( 2005 ) and hanawa et al . ( 2010 ) show that primaries accrete more than secondaries . latest sph simulations demonstrate that flows from the secondary to the primary within the roche lobes are a sensitive function of gas temperatures , and in the case of the high temperature the primary accretes more than the secondary , and vice versa @xcite . our aste observations of an archetypal protostellar binary l1551 ne imply that dispersion of protostellar envelopes , which replenish cbds with fresh materials , also needs to be taken into account as a physical mechanism to set the binary masses and the ratios . the present masses of sources a and b in l1551 ne are estimated to be @xmath70.67 @xmath14 and @xmath70.13 @xmath14 , respectively , and the mass of the cbd @xmath70.026 @xmath14 @xcite . the mass of the possible infalling component around the cbd is even smaller ( @xmath70.0023 @xmath14 ) @xcite . thus , even if all the amount of the material in the cbd plus the infalling component is accreted onto source b , the mass ratio does not change much ( from 0.19 to 0.24 ) . replenishment of the material to the cbd from the surrounding envelope is required to significantly change the mass and the mass ratio of the protostellar binary system . our aste observations show that the envelope is being dispersed , however , and that further replenishment of materials from the envelope to the cbd will not be expected . therefore , even though l1551 ne is a class i protostellar binary associated with the cbd and the protostellar envelope , the mass and the mass ratio have been already close to the final values . physical processes before the start of the envelope dispersion through the interaction with the outflow driven from l1551 irs 5 , such as the fragmentation of the initial pseudo - disk and subsequent accretion from the cbd and the envelope , must be responsible to set the mass and mass ratio of the binary system . our aste observations of l1551 ne indicate that dispersion of natal envelopes surrounding the protostellar binary systems is one of the key physical phenomena to stop the growth of the binary and to determine the final masses and the mass ratios of the binary . as described in the last subsection , interactions between protostellar envelopes and outflows , which causes the envelope dissipation , are ubiquitous . systematic studies of envelope dispersion surrounding the protostellar binaries with various masses and mass ratios will be intriguing to understand the physical mechanism to set the final binary mass ratios and the origin of the wide range of the binary mass ratios . we have conducted single - dish mapping observations of dense gas around an archetypal protostellar binary l1551 ne in the c@xmath0o ( @xmath1=32 ) , @xmath2co ( @xmath1=32 ) , cs ( @xmath1=76 ) , and the so ( @xmath3=7@xmath46@xmath5 ) lines with aste . we have also combined the aste mapping data in the c@xmath0o line with the interferometric data taken with the sma . the main results are summarized below : * all the four molecular lines are detected toward the protostellar position , with multiple and/or non - gaussian spectral shapes . in particular , the c@xmath0o spectrum consists of a narrow ( @xmath790.7 km s@xmath8 ) component with a central velocity of @xmath806.7 km s@xmath8 and a wide ( @xmath792.2 km s@xmath8 ) component with a central velocity of @xmath806.9 km s@xmath8 . the wide component originates from the compact circumbinary disk ( cbd ) , while the narrow component from the extended protostellar envelope . * the c@xmath0o map in l1551 ne primarily traces a @xmath720000-au scale protostellar envelope ( env ) , elongated along the northwest to southeast direction , approximately parallel to the major axis of the central cbd . velocity channel maps of the c@xmath0o line show that the blueshifted ( @xmath806.26.7 km s@xmath8 ) emission is located to the northeast while the redshifted emission ( @xmath806.7 8.0 km s@xmath8 ) to the southwest ; @xmath51 , a velocity gradient of env along the minor axis , while there is no detectable velocity gradient along the major axis . the observed velocity gradient in env can be interpreted as a dispersing gas motion . velocity channel maps of the high - velocity redshifted ( @xmath818.0 km s@xmath8 ) @xmath2co emission show two @xmath94000-au scale redshifted components to the @xmath740@xmath11 southwest of the central protostar ( red1 ) and @xmath720@xmath11 west ( red2 ) . the submillimeter cs ( 76 ) emission selectively traces red1 , suggesting high - temperature in this high - velocity component . both red1 and red2 exhibit similar velocity gradients such that the velocities become closer to the systemic velocity of env as the positions become closer to that of env . these high - velocity redshifted components most likely trace the outflow components driven from the neighboring protostellar source , l1551 irs 5 , which are colliding with env surrounding l1551 ne . * the combined aste+sma map in the c@xmath0o emission shows that the @xmath6300 au scale cbd and @xmath6700 au scale infalling component found in our previous sma observations are surrounded by env with a slightly blueshifted velocity ( @xmath70.2 km s@xmath8 ) . env does not show any rotational motion but dispersing gas motion with an outward velocity of @xmath70.3 km s@xmath8 in contrast to the inner gas components . these results indicate that env is kinematically distinct from the inner gas components . * the net momentum , kinetic and internal energies of the redshifted outflow components driven from l1551 irs 5 are comparable to those of env , suggesting that the outflow collisions can enforce significant dynamical impacts on env . thus , the dispersing motion in env is likely caused by the interactions with the outflows driven from l1551 irs 5 . * since env is being dispersed and the inner gas components do not have sufficient materials to alter the mass and mass ratio of the protostellar binary of l1551 ne , the current mass ( @xmath70.8 @xmath14 ) and the mass ratio ( @xmath70.19 ) are close to the final values , even though l1551 ne is a young class i protostellar binary . our aste+sma observations of l1551 ne suggest that dispersion of natal protostellar envelopes which can replenish the cbds with new materials is one of the important physical mechanisms to set the final binary mass ratios . the interactions with the outflows from l1551 irs 5 are unlikely to trigger the binary formation in l1551 ne as suggested by previous studies , but suppress the further growth of the protostellar binary system . we would like to thank t. matsumoto , t. hanawa , j. lim , n. ohashi , and p. t. p. ho for their fruitful discussions . we are grateful to s. ohashi for his support during our aste observations . s.t . acknowledges a grant from the ministry of science and technology ( most ) of taiwan ( most 102 - 2119-m-001 - 012-my3 ) in support of this work . andr , p. , ward - thompson , d. , & barsony , m. 2000 , in protostars and planets iv , ed . mannings , v. , boss , a. p. , & russell , s. s. ( tucson : univ . arizona press ) , 59 arce , h. g. & sargent , a. i. 2006 , , 646 , 1070 bate , m. r. 1997 , , 285 , 16 bate , m. r. , & bonnell , i. a. 1997 , , 285 , 33 bate , m. r. 2000 , , 314 , 33 chen , x. , arce , h. g. , zhang , q. @xmath82 2013 , , 768 , 110 chou , t .- l . , takakuwa , s. , yen , h .- w . , ohashi , n. , & ho , p. t. p. 2014 , , 796 , 70 crapsi , a. , caselli , p. , walmsley , c. m. , @xmath82 2004 , , 420 , 957 dutrey , a. , di folco , e. , guilloteau , s. , @xmath82 2014 , , 514 , 600 froebrich , d. 2005 , , 156 , 169 fuente , a. , martn - pintado , j. , bachiller , r. , rodrguez - franco , a. , & palla , f. 2002 , , 387 , 977 fuller , g. a. , & ladd , e. f. 2002 , , 573 , 699 goodwin , s. p. 2013 , , 430 , l6 hanawa , t. , ochi , y. , & ando , k. 2010 , , 708 , 485 hayashi , m. & pyo , t .- s . 2009 , , 694 , 582 jrgensen , j. k. , schier , f. l. , & van dishoeck , e. f. 2004 , , 416 , 603 kamazaki , t. , ezawa , h. , tatematsu , k. , @xmath82 2005 , astronomical data analysis software and systems xiv , asp conf . ser . , 347 , ed . p. shopbell , m. britton , & r. ebert , ( san francisco : asp ) , 533 kitamura , y. , kawabe , r. , & saito , m. 1996 , , 457 , 277 kohno , k. 2005 , the cool universe : observing cosmic dawn , asp conf . 344 , ed . c. lidman & d. alloin , ( san francisco : asp ) , 242 matsumoto , t. 2007 , , 59 , 905 momose , m. , ohashi , n. , kawabe , r. , hayashi , m. , & nakano , t. 1996 , , 470 , 1001 momose , m. , ohashi , n. , kawabe , r. , nakano , t. , & hayashi , m. 1998 , , 504 , 314 moriarty - schieven , g. h. , butner , h. m. & wannier , p. g. 1995 , , 445 , l55 moriarty - schieven , g. h. , johnstone , d. , bally , j. , & jenness , t. 2006 , , 645 , 357 motte , f. , & andr . , , 365 , 440 myers , p. c. , evans , n. j. , ii & ohashi , n. 2000 , in protostars and planets iv , ed . mannings , v. , boss , a. p. , & russell , s. s. ( tucson : univ . arizona press ) , 217 ochi , y. , sugimoto , k. , & hanawa , t. 2005 , , 623 , 922 plambeck , r. l. , & snell , r. l. 1995 , , 446 , 234 raghavan , d. , mcalister , h. a. , henry , t. j. @xmath82 2010 , , 190 , 1 reipurth , b. , yu , k. c. , heathcote , s. , bally , j. , & rodrguez , l. f. 2000 , , 120 , 1449 reipurth , b. , rodrguez , l. f. , anglada , g. , & bally , j. 2002 , , 124 , 1045 reipurth , b. , clarke , c. j. , boss , a. p. @xmath82 2014 , protostars and planets vi , 267 saito , m. , kawabe , r. , kitamura , y. , & sunada , k. 2001 , , 547 , 840 schilke , p. , groesbeck , t. d. , blake , g. a. , & phillips , t. g. 1997 , , 108 , 301 stojimirovi , i. , narayanan , g. , snell , r. l. , & bally , j. 2006 , , 649 , 280 takahashi , s. , saito , m. , takakuwa , s. , & kawabe , r. 2006 , , 651 , 933 takakuwa , s. , ohashi , n. , & hirano , n. 2003 , , 590 , 932 takakuwa , s. , ohashi , n. , ho , p. t. p. , @xmath82 2004 , , 616 , l15 takakuwa , s. , kamazaki , t. , saito , m. , yamaguchi , n. , & kohno , k. 2007 , , 59 , 1 takakuwa , s. , ohashi , n. , bourke , t. l. , @xmath83 @xmath84 2007 , , 662 , 431 takakuwa , s. , & kamazaki , t. 2011 , , 63 , 921 takakuwa , s. , saito , m. , lim , j. , @xmath82 2012 , , 754 , 52 takakuwa , s. , saito , m. , lim , j. , & saigo , k. 2013 , , 776 , 51 takakuwa , s. , saito , m. , saigo , k. , @xmath82 2014 , , 796 , 1 tang , y .- w . , dutrey , a. , guilloteau , s. , @xmath82 2014 , , 793 , 10 tobin , j. j. , chandler , c. j. , wilner , d. j. , @xmath82 2013 , , 779 , 93 wilson , t. l. , & rood , r. 1994 , , 32 , 191 wu , p .- f . , takakuwa , s. , & lim , j. 2009 , , 698 , 184 yokogawa , s. , kitamura , y. , momose , m. , & kawabe , r. 2003 , , 595 , 266 young , m. d. , baird , j. t. , & clarke , c. j. 2015 , , 447 , 2907 llllccccccccc env & c@xmath0o ( 32 ) & 04 31 44.4 & 18 08 36.4 & 5.31 - 8.15 & 22100@xmath3418200 & 0.72 & & 1.3 - 3.1 & 5.3 & 0.38 - 0.93 & 3.5 - 8.7 & 1.14 - 2.78 + red1 & cs ( 76 ) & 04 31 42.0 & 18 08 06.0 & 8.04 - 9.13 & @xmath504000 & 0.97 & 1.9 & 0.02 - 0.83 & @xmath501.9 & 0.039 - 1.574 & 0.10 - 4.23 & 0.73 - 29.74 + red2 & @xmath2co ( 32 ) & 04 31 42.8 & 18 08 35.0 & 9.47 - 12.53 & 4300@xmath344000 & 2.0 & 3.5 & 0.02 - 0.05 & 8.4 & 0.069 - 0.169 & 0.42 - 1.04 & 2.4 - 5.9 +

interest in the gravitational @xmath1-matrix at transplanckian energies @xcite has revived in the past few - years @xcite , when explicit solutions of the so - called reduced - action model @xcite have been found @xcite . the model is a much simplified version of the acv eikonal approach @xcite to transplanckian scattering in string - gravity , and is valid in the regime in which the gravitational radius @xmath2 is much larger than the string length @xmath3 , so that string - effects are supposed to be small . the reduced - action model ( sec . [ s : rcram ] ) was derived by justifying the eikonal form of the @xmath1-matrix at impact parameter @xmath4 on the basis of string dynamics and by then calculating the eikonal itself ( of order @xmath5 ) in the form of a 2-dimensional action , whose power series in @xmath0 corresponds to an infinite sum of proper irreducible diagrams ( the `` multi - h '' diagrams @xcite ) , evaluated in the high - energy limit . the model admits a quantum generalization @xcite of the @xmath1-matrix in the form of a path - integral with definite boundary conditions of the reduced - action exponential itself . the main feature of the model and of its boundary conditions is the existence of a critical impact parameter @xmath6 such that , for @xmath7 the @xmath1-matrix matches the perturbative series and is unitary , while for @xmath8 the field solutions are complex - valued and the elastic @xmath1-matrix is suppressed exponentially . the suppression exponent is of order @xmath9 ( @xmath10 being the planck length ) or , if we wish , of the same order as the entropy of a black - hole of radius @xmath11 . from various arguments we believe that in the region in which @xmath8 ( that is , @xmath4 is smaller than the gravitational radius ) , a classical gravitational collapse is taking place . a key issue related to the collapse region is the possible existence of information loss at quantum level , which in the acv model shows up as lack of @xmath1-matrix unitarity . according to @xcite such unitarity loss is mostly related to a restrictive boundary condition , which is required for the solutions of the model be uv - safe . if such condition is relaxed , further solutions show up , which could contribute to unitarity , but are irregular i.e. , dominated by planckian distances , region in which the model itself is inadequate . the main motivation of the present paper is to improve and complete some unsatisfactory aspects of the reduced - action model which might be crucial at planckian distances , but show up already at distances of order @xmath12 , the gravitational radius . the new model that we shall present here features two improvements . firstly , the eikonal representation itself is embedded in three dimensions , so as to be able to describe the motion of the breit - frame during the scattering process . by comparison , the usual two - dimensional representation is inadequate outside the regime of very small angle scattering , and in particular in the collapse region . in fact , the energy - momentum of the impinging particles is taken to be an external light - like source without any deflection which generates gravitational fields characterized by aichelburg - sexl - like shock waves @xcite . however , the resulting action predicts the einstein deflection angle @xcite and corrections to it @xcite , at variance with the original assumption of frozen sources . instead , we know from the beginning @xcite that in the classical limit the gravitational equations should predict both particle motion and fields in a self - consistent way . for that to be possible , the improved eikonal representation is needed . the additional important improvement of the present model is the treatment of rescattering corrections of the produced gravitons . such effects were argued in @xcite to be of particular interest for the irregular solutions of the model , perhaps explaining their relationship to collapse and their contribution to unitarity . but , independently of such a feature , we shall argue here that even for regular ( uv - safe ) solutions , rescattering corrections are needed in order to achieve self - consistency of motion and metric at higher orders in the eikonal expansion . to start with , we deal with the self - consistency problem at leading level , in which the eikonal refers to frozen undeflected sources , the metric contains two aichelburg - sexl ( as ) shock waves , while the action predicts the einstein deflection angle @xmath13 , as function of the impact parameter @xmath4 and of the gravitational radius @xmath14 ( @xmath15 being the invariant mass of the system ) . the problem in introducing the particle motion is that the kinematical corrections implementing it are of the same relative order ( @xmath16 or higher ) as the irreducible dynamical corrections to the einstein deflection . how to disentangle the former terms from the latter ones ? a hint about solving the question above was provided in ref . @xcite , where it was shown that given the particular coulombic form of the leading eikonal it is possible to take into account the motion of the breit - frame of the particles without affecting the @xmath1-matrix eigenvalue which is still provided by the naive two - dimensional fourier transform . such observation is the basis for our treatment of motion in sec . [ s : ilepp ] , which suggests writing a modified metric , which is self - consistent at leading level . the main difference with the previous one @xcite lies in the introduction of two - body `` shifts '' of t hooft type @xcite , summarizing the action of the leading @xmath1-matrix at two - body level . that is not enough however . the reduced - action model contains the longitudinal fields @xmath17 ( @xmath18 ) characterized by as shock waves centered on the particles with profile functions @xmath19 ( @xmath20 ) ; furthermore , it generates a gravitational wave also with a field @xmath21 defined by the h - diagram . the longitudinal field is here calculated in the improved eikonal representation in sec . [ s : ssw ] and found to be consistent with as waves @xcite which are delayed in time and rotated in space , as suggested in sec .. [ s : ilepp ] . the transverse field , on the other hand , provides corrections to the leading eikonal profiles in @xmath22 and @xmath23 , and causes modifications of the metric inside the light - cone by providing important rescattering corrections . we then set up a perturbative procedure , in order to deal with the self - consistency problem at higher orders in the eikonal expansion . at next order we find the h - diagram deflection @xcite which again appears as a feature of the action , but is not incorporated in the metric , which would require further modifications of the energy - momentum tensor . surprisingly , we find that the class of multi - h diagrams considered in @xcite is not sufficient in order to provide a satisfactory metric at this level but we need to go one further step , and calculate rescattering corrections , which carry the information due to two - body shifts for the produced gravitons . such shifts occur in an approximate solution to the rescattering equations under consideration since a long time @xcite , which is here worked out to completion in sec . [ s : rcram ] . the results so obtained complete the picture of the improved eikonal representation in sec . [ s : ssw ] for corrections of relative order @xmath0 . finally , in sec . [ s : isrp ] we somewhat change subject , and we address the gross features of the ultraviolet - sensitive solutions of the model , whose importance lies in the fact that they could play a role for the recovery of unitarity and/or the related information loss . we then summarize the essential results of the paper and their consequences for action and metric in sec . [ s : conc ] , by discussing also some suggestions which arise from our preliminary analysis of singular solutions . acv @xcite have shown that the leading contributions to the high - energy elastic scattering amplitude come from the @xmath24-channel iteration of soft - graviton exchanges , which can be represented by effective ladder diagrams as in fig . [ f : ladder ] . the purpose of the present section is to recall the method of resumming the effective ladder contributions to all orders so as to provide the so - called eikonal representation for the elastic @xmath1 matrix . this representation is here `` improved '' in the sense that we do not make a separation of longitudinal and transverse variables by neglecting the leading scattering angle . instead , by following @xcite , we use the exact 3-dimensional phase - space of the on - shell particles at each gravitational eikonal exchange and we prove that due to the coulombic form of the exchange the @xmath1-matrix is nevertheless provided by a 2-dimensional transform . -matrix in the eikonal approximation . solid lines : on - shell external particles ; dashed lines : eikonal gravitational exchanges._,scaledwidth=70.0% ] [ f : ladder ] it is important to note that the exchanged gravitons ( dashed lines ) are associated to a propagator @xmath25 and are coupled to the colliding particles / strings ( solid lines ) with an interaction strength equal to @xmath26 for each pair of vertices . furthermore , the particles lines are on - shell along the effective ladder , as proved in ref @xcite by the sum over inelastic excitations also . the generic ladder is thus built by iteration of the basic rung @xmath27 we start computing the 2-rung ladder ( fig . [ f : ladder].b ) @xmath28 in the center - of - mass ( cm ) frame where @xmath29 , @xmath30 , @xmath31 , @xmath32 , the two mass - shell delta functions in eq . ( [ amp2 ] ) become @xmath33 and , for the same reason , the two external mass - shell deltas in eq . ( [ rung2 ] ) constrain @xmath34 , @xmath35 , @xmath36 . in this way , the 4d integration @xmath37 reduces to a 2d integral is a function of @xmath24 and @xmath38 . ] @xmath39 by taking into account the expression of @xmath40 and @xmath41 in terms of @xmath42 , @xmath43 and @xmath44 and using the identity @xmath45 it is easy to evaluate the azimuthal integral in eq . ( [ a2 ] ) and , more interestingly , to show that it can be rewritten as the azimuthal integral of a pure 2d propagator : @xmath46 where @xmath47 and @xmath48 are fictitious 2d vectors with the same modulus as @xmath49 and @xmath50 respectively , and @xmath51 is the angle between them . with this trick , @xmath52 can be rewritten as a 2d convolution : @xmath53({\boldsymbol{q } } ) \ ; , \quad ( { \boldsymbol{q}}^2=-q^2)\;.\ ] ] it is now straightforward to generalize that procedure to the @xmath54-rung ladder , by deriving the recursion formula @xmath55 \;. \label{ampn}\end{aligned}\ ] ] it should be noted that such simplification is due to the peculiar form of the graviton interaction from which eq . ( [ azimuthal ] ) follows . finally , the eikonal @xmath1 matrix ( with overall momentum conservation factored out ) is given by the sum of the ladder diagrams with the usual combinatorial factor @xmath56 and with the final - state mass - shell deltas removed : [ seik ] @xmath57 \;.\ ] ] by applying a 2d fourier transform to eq . ( [ ssum ] ) , convolutions become standard products and the r.h.s . of eq . ( [ ssum ] ) reduces to an exponential series : @xmath58 where we interpret the 2d fourier transform of the one - graviton amplitude @xmath59 as _ eikonal action _ @xmath60 . actually , the definition of @xmath60 ( and of all convolutions involved in this calculation ) requires the introduction of an infrared cutoff @xmath61 in order to regularize the `` coulomb '' divergence typical of long - range interactions @xmath62 : @xmath63 where @xmath64 and @xmath65 . a couple of remarks are in order : * the `` position '' variable @xmath66 , being conjugate to @xmath47 , is usually interpreted as impact parameter of the collision , and @xmath67 as the associated phase shift . however , @xmath47 is a fictitious 2d vector whose polar angle is actually undefined , therefore @xmath66 can not be thought of as a purely transverse ( to @xmath68 ) vector . * the dependence of @xmath69 on @xmath70 amounts to a @xmath4-independent phase , and therefore it is unimportant in the determination of the scattering angle . however such a phase depends on the total energy @xmath71 . this may cause a cutoff - dependent time evolution of the scattered particles , hopefully without physical consequences . both issues will be discussed in detail in the next section . the purpose of this section is to derive the motion of two quantum particles subject to a scattering amplitude given by the leading eikonal result of @xcite and recalled in sec . [ s : ersa ] . suppose we prepare a state of two well separated light - like free particles in the past . in a first - quantization description the wave function is given by @xmath73 in the far future the wave function is expressed in terms of the @xmath1-matrix as follows : @xmath74 it is convenient to perform the calculation in the cm frame . as usual , we define the cm coordinate @xmath75 and relative coordinate @xmath76 [ cmcoordinates ] @xmath77 for simplicity , we limit ourselves to states which have zero total momentum , by setting [ wavefunc ] @xmath78 with this restriction the cm coordinate is completely delocalized , but we can build a state representing a localized wave packet in the relative coordinate : @xmath79 where @xmath80 is the relative momentum , @xmath81 is its mean value and @xmath82 a measure of its broadening , @xmath83 is a normalization factor , @xmath84 is the total cm energy , while @xmath85 is the relative position at time @xmath86 . in fact , by inserting eqs . ( [ wavefunc ] ) into eq . ( [ d : psiin ] ) we are left with an integral in @xmath87 whose phase @xmath88 is stationary for @xmath89 since the shape of @xmath90 forces @xmath91 . going back to particles coordinates , we confirm that in the past each particle moved at the speed of light in the direction of @xmath92 : @xmath93 on the other hand , by substituting eqs . ( [ wavefunc ] ) in the outgoing wave ( [ d : psiout ] ) together with the energy - momentum conserving eikonal @xmath1-matrix ( [ seik ] ) @xmath94 we find @xmath95 where @xmath47 and @xmath66 are auxiliary 2d vectors , the former being determined by the condition @xmath96 . in this case , the stationarity of @xmath97 with respect to independent variations of @xmath87 , @xmath98 and @xmath66 yields the asymptotic motion in the future , as follows . in the plane of scattering @xmath99 we introduce the polar angles @xmath100 of momenta and coordinates , and write @xmath101 where we used the notation @xmath102 , @xmath64 , @xmath103 and the relations @xmath104 , @xmath105 . we can now determine the five unknowns @xmath106 , @xmath107 , @xmath108 and @xmath66 by finding the stationarity point of the phase with respect to variations of the five integration variables @xmath109 , @xmath107 , @xmath108 and @xmath66 , while @xmath86 and @xmath85 are known parameters . we have @xmath110 where @xmath111 denotes the sign of @xmath112 . the first equation causes @xmath66 to be aligned with @xmath49 ( @xmath113 ) and to have modulus @xmath114 the second equation yields a further relation involving the physical impact parameter @xmath115 : @xmath116 allowing us to determine the ( attractive ) scattering angle @xmath117 @xmath118 the third equation identifies the outgoing trajectory to be the straight line @xmath119 with the same impact parameter of the incoming one , thus implying angular momentum conservation . note that this trajectory intersects the incoming one at the point @xmath120 , as is apparent from fig . [ f : deflection ] . [ f : deflection ] the fourth equation provides the motion of the outgoing relative coordinate @xmath121 meaning that each particle keeps on moving at the speed of light . a very peculiar behaviour of our result is that the motion of the outgoing particles suffers a time delay given by the energy - derivative of the action @xmath122 with respect to a motion with the same deflection but at constant speed . this is easily seen by projecting the incoming trajectory ( [ xin ] ) on @xmath87 @xmath123 and by choosing the origin of time @xmath124 when @xmath125 is ( or would be ) at the intersection point @xmath126 of the trajectories ( [ outtraj],[inmotion ] ) , i.e.@xmath127 , so that eqs . ( [ inmotion],[motion ] ) become @xmath128 showing that @xmath129 at @xmath130 . let us stress that this time delay is peculiar of gravity , since the energy dependence of the amplitude stems from the gravitational coupling @xmath131 . in other theories , like electrodynamics , the coupling is independent of the particles energies , and one obtains analogous equations with @xmath132 . we have seen that the leading eikonal @xmath1-matrix provides a definite scattering angle and wave - packet motion . it also provides a gravitational metric , associated to the gravitaional fields @xmath133 , whose longitudinal terms are expressed in terms of the eikonal amplitude @xmath134 as follows @xcite [ hfields ] @xmath135 where @xmath136 and we actually deal with the situation before collision ( @xmath137 ) . the collision process , according to the reduced action model to be described in sec . [ s : rcram ] , introduces further ( transverse ) components of the metric field , and modifies the expressions ( [ hfields ] ) by corrections of relative order @xmath0 and higher . nevertheless , at leading level , starting from the ( improved ) eikonal representation only , it should be possible to describe the scattering on the basis of an associated metric . here we suggest a simple way to do it , which is based on t hooft s understanding @xcite of the @xmath1-matrix as a coordinate shift . the starting point is to write the leading order metric before collision , provided by the expressions ( [ hfields ] ) as follows : @xmath138 \;.\ ] ] we see that it consists of two aichelburg - sexl ( as ) shock waves @xcite travelling against each other . if only one is present the one at @xmath139 , say , a test particle impinging on it will acquire a shift in the @xmath140 direction : @xmath141 as a consequence , after the shift , the test particle will be deflected . in particular , if the particle was moving at fixed @xmath140 and @xmath142 , the scattering angle @xmath143 is given by @xmath144 , see fig . [ f : deflection ] . the scattering angle formula is derived by either calculating the action @xmath145 of eq . ( [ amp1b ] ) from the shift ( [ vshift ] ) as in ref . @xcite and then using eq . ( [ angle ] ) , or by computing the geodesics by standard methods . denote the transverse direction starting from the center of mass ( cm ) of the two colliding particles . at the point @xmath146 ( @xmath147 ) the test particle leaves the first ( second ) wave front._,scaledwidth=60.0% ] [ f : doppioshift ] the next question is : what happens when both as waves are present ? in order to understand that , consider a test particle at @xmath148 , travelling along with particle 1 ( @xmath149 ) shortly before it . then , at @xmath139 ( @xmath150 ) the test particle will collide with particle 2 and will acquire the shift ( [ vshift ] ) , while the location of particle 1 will be forced to do the same with @xmath151 , evaluated at the `` on shell '' point @xmath152 ( @xmath153 ) , see fig . [ f : doppioshift ] . next , the test particle trajectory is deflected on a trajectory @xmath154 , @xmath155 , therefore , due to the @xmath40 factor , it will be shortly hit by particle 1 , travelling at @xmath156 , and thus will acquire the shift @xmath157 to sum up , the test particle acquires two shifts @xmath158 separated by a short time interval @xmath159 , which vanishes in the @xmath160 limit . as a consequence , it is natural to assume that particle 1 acquires two shifts also : @xmath161 where @xmath162 is actually a possibly singular parameter describing the scattering of particle 1 on itself , which is here taken to be a @xmath115-independent constant , @xmath163 . taking into account both shifts , particle 1 goes from @xmath164 to @xmath165 at time @xmath166 and is then deflected by an angle @xmath167 on the trajectory means `` rotated '' ; see also eqs . ( [ rotation ] ) . ] [ coor ] @xmath168 where we recall that @xmath169 . the total collision time from @xmath164 to @xmath170 is therefore @xmath171 from the stationarity condition of the collision time with respect to @xmath172 , we obtain the scattering angle @xmath173 where @xmath174 is `` on - shell '' . since @xmath175 , eq . ( [ scatang ] ) is consistent with the action determination , at leading level : @xmath176 besides providing the scattering angle , the double - shift picture has a strict analogy with the wave packet motion in sec . [ s : wpm ] . the former can be summarized in 3 steps : * the `` come - back '' motion from @xmath164 back to @xmath177 , taking the time @xmath178 which is just the retardation time with respect to travel at the speed of light ; * the motion along the residual shift @xmath179 , taking time @xmath180 = \xi\tan\frac{\bar\theta}{2}\;,\ ] ] where we have defined the angular shift @xmath181 as in fig . [ f : doppioshift ] ; * the motion along the deflected trajectory , taking time @xmath182 by comparison , in the wave packet motion the delay time is just the same , while steps ( b ) ( travel to point @xmath183 joining incoming and outgoing trajectories ) and ( c ) ( travel to @xmath184 ) both take time @xmath185 . although different , the expressions ( [ timeb ] ) and ( [ timec ] ) correspond to the total time @xmath186\ ] ] whose difference with the former is of relative order @xmath0 , because both @xmath187 and @xmath181 are of order @xmath188 . therefore the two pictures coincide at leading order . similarly , the outgoing impact parameter is @xmath189\ ] ] which differs from the incoming one by relative order @xmath0 also . furthermore both differences vanish in the limit @xmath190 , which is obtained by setting @xmath191 this model , that we could call the `` linear model '' , could be described by the scattering of extended sources , in which the energy of each source inside a disk of radius @xmath106 increases like @xmath106 . we conclude that the double - shift picture provides us with some understanding of the evolution of the metric at collision and after , which is satisfactory at leading level , even if a full account of the collision process is lacking . the basic features argued so far are : * the existence of a delay time @xmath192 which causes a shift of either shock wave around @xmath193 in the form @xmath194 around particle 1 , with a similar one for particle 2 ; * the deflection of the particle trajectories , causing a rotation by the scattering angle @xmath187 for @xmath195 and reaching @xmath184 at the collision time @xmath196 @xmath197 where the rotated coordinates are @xmath198 therefore , the particle motion occurs explicitly in the expression of the metric , that we could call at this point , a self - consistent leading metric . field ( red cross ) . the parametrization of momenta is choosen so as to simplify the the calculations._,scaledwidth=90.0% ] [ f : simpleinsertion ] we have seen before how the improved eikonal model is able to describe particle motion at leading level ( sec . [ s : wpm ] ) , and how this feature can be incorporated by trajectory shifts in the corresponding shock - waves ( sec . [ s : sdts ] ) . we want now to show how the shock - wave fields can be explicitly derived from the sum of diagrams describing eikonal scattering , as defined in sec . [ s : ersa ] . we confirm in this way the delay shifts introduced previously , leading eventually to the scattered energy - momentum . in order to calculate the fields @xmath199 and @xmath200 we should consider the eikonal diagrams ( fig . [ f : ladder ] ) in presence of a linearly coupled external source ( @xmath201 , @xmath202 ) which generates the fields by a functional derivative of the corresponding semiclassical @xmath1-matrix or , in other words , of the corresponding action . we can write @xmath203 and compute the result by inserting the @xmath202 source on the eikonal diagrams , in all possible ways . since the external current diagrams generate insertions of momentum @xmath204 and thus generally do not conserve energy - momentum , we should restore time - ordered integrations in order to evaluate them properly . since @xmath200 is coupled to @xmath205 only , the @xmath206 conservation is not affected , and it is sufficient to consider @xmath207-ordered diagrams as depicted in fig . [ f : simpleinsertion ] . let us first consider the small scattering angle kinematics @xmath208 , in which the longitudinal dynamics is separated from the transverse one . in order to understand the issue , let us start from the simplest diagrams in fig . [ f : simpleinsertion ] . we notice that , while the @xmath209-exchange in the eikonal line has negligible @xmath210 because of two mass - shell conditions @xmath211 , the @xmath204-exchange has only one , @xmath212 , so that some @xmath213 is still allowed , provided @xmath214 . therefore @xmath215 is still negligible , but the @xmath216 leakage should be considered in the @xmath217 evolution . in the example of fig . [ f : simpleinsertion ] we choose to couple the @xmath202 current to initial and final particles of type 1 symmetrically , ) with a wave function having energy centered around @xmath218 , which amounts to setting @xmath219 and @xmath220 . ] so that a @xmath216-independent factor of @xmath221 is factored out . in this way , the lowest - order diagram in fig . [ f : simpleinsertion].a yields the born - level field @xmath222 where @xmath223 is the leading profile function of the shock - wave . on the other hand , the @xmath209-exchange yields the leading eikonal @xmath224 coupled to the charge @xmath225 , where @xmath226 according to eq . ( [ h0++ ] ) . therefore , diagram [ f : simpleinsertion].b predicts a ( subleading ) contribution to @xmath200 [ h1bc ] @xmath227\ ] ] while diagram [ f : simpleinsertion].c yields @xmath228\ ] ] because of the opposite sign in @xmath216 and @xmath207 . we shall interpret the results ( [ h1bc ] ) as the first order expansion in @xmath229 of a shifted field @xmath230 which is retarded ( advanced ) for @xmath231 ( @xmath232 ) . therefore , the result ( [ h++sh ] ) confirms the delay - shift of sec . [ s : sdts ] , in a form which is appropriate for the calculation of the real part of the action @xmath233 , or to the principal - value prescription of the @xmath234 propagator . ( red cross ) . the blobs represent ladders of eikonal exchanges._,scaledwidth=65.0% ] [ f:2bolle ] ( red cross)._,scaledwidth=36.0% ] [ f:2prop1x ] the higher - order contributions to the shifted field are computed from their definition in fig . [ f:2bolle ] . in order to do the explicit calculation , we have to explain in more detail how the @xmath207-ordering just mentioned is generated from the usual @xmath235-ordering in perturbation theory . we shall work off energy - shell , but on mass - shell for the intermediate particles so that , in the example of fig . [ f:2prop1x ] , @xmath236 , @xmath237 , , @xmath238 . therefore , to each eikonal exchange , we assign the energy differences in the center of mass frame @xmath239 , e.g. , @xmath240 and so on.s occurring in the last equation s line is sufficient to our purposes , because we shall need it around @xmath241 . the argument is therefore valid for the finite - angle kinematics too . ] the ordered time integrations , up to time @xmath235 of the field , yield therefore the factor @xmath242}{[{\varepsilon}+{\mathrm{i}}(\delta_1+\delta_2)]\ , [ { \varepsilon}+{\mathrm{i}}\delta_1 ] } \to \frac{\exp[{\mathrm{i}}(\delta_1+\delta_2)t]}{2 ( { \varepsilon}+{\mathrm{i}}\delta_2)({\varepsilon}+{\mathrm{i}}\delta_1 ) } \;,\ ] ] where the last expression is obtained after @xmath243 symmetrization , because of the integrations on @xmath244 , @xmath245 induced by @xmath246 , @xmath247 , with a factor which is symmetrical under @xmath248 permutations . finally the factorized @xmath249 integrations induced by the angular ones on the @xmath248 s are done separately , by using the identity @xmath250 in which the integration contour is closed in the upper ( lower ) half - plane for @xmath251 ( @xmath137 ) by noting that corrections due to the large @xmath252 contour are exponentially small in @xmath253 in our finite angle kinematics . similarly , the time and @xmath252 integrations can be done after the insertion , and yield @xmath254 this means that eikonal echanges occurring both before and after the insertion are suppressed , in the @xmath255-functions so that overlap of subsequent contributions is allowed for a limited time . the uncertainty so introduced in the time of shift is expected to be larger than @xmath256 because of the number of coherent reinteractions , but in any case smaller than @xmath11 , which would correspond to the maximal number @xmath257 of leading eikonal scatterings . ] because of @xmath258 , so that the whole sum reduces to the diagrams in fig . [ f:1bolla1x ] . ._,scaledwidth=80.0% ] [ f:1bolla1x ] in order to compute them we first diagonalize the scattering amplitude in impact parameter @xmath66 space and we then compute the inserted field by fourier transform in the @xmath259 variables , in order to yield its @xmath260 dependence . the non - trivial point about impact parameter transform is that it is 2-dimensional , while the integration variables @xmath261 and @xmath262 ( in , say fig . [ f:2prop1x ] ) are 3-dimensional in the finite angle kinematics . fortunately , the conditions @xmath263 fix the 3-components of both @xmath261 and @xmath49 , by determining the directions of the respective breit frames . we thus have the bessel transform @xmath264 which , by the clever identity ( [ azimuthal ] ) yields the same result as the 2-dimensional azimuthal average @xmath265 which is a 2-dimensional convolution diagonalized in @xmath66-space . we thus obtain that the amplitude in front of the field insertion resums to the exponentiated result @xmath266 and cancels out with the factor @xmath267 in eq . ( [ derfunz ] ) . we are left with the calculation of the @xmath200 field itself . notice that in fig . [ f:1bolla1x].b the insertion is on the incoming leg @xmath87 , while in fig . [ f:1bolla1x].a it occurs on the final leg @xmath268 . since @xmath269 because of the mass - shell plus energy - conservation constraints , the difference is just a rotation @xmath270 by the angle @xmath271 . when inserted in the field propagator it leads to a rotated variable projection should be rotated also with the @xmath272-axis in the direction of the deflected longitudinal direction . ] @xmath273 with a mass - shell condition^{-1}$ ] . however , for @xmath274 the contour can be closed on the pole , with the same result , while for @xmath275 it vanishes . this remark confirms the ordering @xmath276 ( see app . [ a : hdi ] ) . ] @xmath277 , and we get @xmath278 where we have neglected the exponent dependence on @xmath279 , when @xmath280 is a very small variable if we make the natural assumption @xmath281 , insuring that the field insertion does not modify the scattering angle . ( @xmath199 ) generated by particle 1 ( 2 ) , here represented by the blue ( red ) line . _ ] [ f : shiftedshocks ] finally , by taking the value @xmath282 of the initial energy as we did in the beginning , we get for the field a factorized value proportional to @xmath283 which translates @xmath284 into @xmath276 . the complete resummation in fig . [ f:2bolle ] is thus summarized in the exponentiated form @xmath285 \times \nonumber \\ & \qquad\left\{2\pi r\left[\theta(-x^+)\delta(x^-)a_0({\boldsymbol{x}})+\theta(x^+)\delta(x^-_r)a_0({\boldsymbol{x}}_r ) \right]\right\ } \nonumber \\[1 mm ] & = 2\pi r \left [ \delta\big(x^-_r-\pi r a_0(b)\big ) a_0({\boldsymbol{x}}_r)\theta(x^+ ) + \delta\big(x^- + \pi r a_0(b)\big ) a_0({\boldsymbol{x}})\theta(-x^+)\right]\;. \label{hppr}\end{aligned}\ ] ] an identical calculation with the exchange @xmath286 yields an analogous result for the field @xmath287 . the overall situation is depicted in fig . [ f : shiftedshocks ] , where we show the shifted ( and slightly rotated ) shock waves , on which the source particles lie before and after the collision ( solid lines ) . the evolution of the particles during the collision actually depends on the reference frame used to describe their motion . for instance , a time - like observer that crosses first shock 2 ( the one on the bottom - right ) describes the scattering in two steps : firstly the shock 2 drags shock 1 and particle 1 by an amount @xmath288 in the @xmath140 direction ; then shock 1 drags shock 2 and particle 2 in the @xmath207 direction . in this frame particle 1 travels along the dashed blue line in fig . [ f : shiftedshocks ] . a symmetric description is given by an observer crossing first shock 1 , which `` sees '' particle 2 moving along the dashed red line . the transplanckian field equations in the acv proposal @xcite were based on two main groups of results . firstly , the acv investigation of string - gravity showed that at transplanckian energies @xmath289 ( @xmath290 ) the eikonal representation with an eikonal operator which is calculable in principle by expanding in @xmath291 and @xmath188 yields a good representation of the scattering amplitude , and incorporates both string- and strong - gravity effects . secondly , in the regime @xmath292 in which string effects are supposed to be small , the irreducible eikonal diagrams are much in correspondence with the effective action of lipatov and co - workers @xcite , who calculated a regge - graviton emission vertex @xcite which is the building block of the effective lagrangian used in @xcite . finally , the equations of motion of the latter by neglecting rescattering terms were shown to yield a shock - wave solution for the fields which is the basis for the reduced - action model investigated in detail in later years @xcite . here , we have considered so far only the leading graviton - exchange kernel in the eikonal , and we have improved the amplitude representation based on it , so as to include a motion of the breit - frame on which the exchange is defined . next , we want to reconsider the field equations in 4-dimensions including rescattering and we shall provide a solution for the fields which fits very well in our understanding of scattering developed so far , by adding corrections which are of relative order @xmath0 and higher . in the effective action framework , the elastic @xmath1-matrix of the tree diagrams in fig . [ f : subldiagrams ] is given in terms of the classical solutions of the lagrangian equation of motion as @xmath293 where @xmath294 and @xmath295 are just rescaled versions of the longitudinal fields considered so far , @xmath296 is related to the transverse field which is proportional to @xmath297 , and @xmath298 represents ( up to an unconventional but convenient factor of @xmath299 ) the energy - momentum tensor of the colliding particles . the usual metric components are given by @xcite @xmath300 + \frac{\kappa}{4 } \left [ \epsilon_{\mu\nu}^{tt } { \boldsymbol{\nabla}}^2\re\phi - \epsilon_{\mu\nu}^{lt } { \boldsymbol{\nabla}}^2\im\phi\right ] { \mathrm{d}}x^\mu { \mathrm{d}}x^\nu \ ; , \label{hmunudxdx}\end{aligned}\ ] ] where @xmath301 denotes the transverse laplacian and the @xmath302 s denote graviton polarizations to be specified below . the lagrangian is conveniently written in terms of real light - cone variables @xmath303 , @xmath304 and a complex transverse variable @xmath305 , @xmath306 . it consists of a kinetic term @xmath307 where the longitudinal fields have a mostly transverse propagator and the ( complex ) @xmath296 field a mostly longitudinal one , of a graviton emission term @xmath308 incorporating lipatov s vertex @xcite , and finally , of a rescattering term @xmath309 which is supposed to take into account the rescattering diagrams of fig . [ f : subldiagrams].c , d . this term is quadratic in @xmath296 , and is likely to play a role when the latter is large . ; _ ( b ) _ a `` multi - h '' diagram of order @xmath310 involving only lipatov s vertices ( red disks ) ; _ ( c ) _ the first diagram with a rescattering vertex ( green triangle ) is of order @xmath311 , but actually vanishes on - shell ; _ ( d ) _ the first nonvanishing rescattering diagram is of order @xmath310._,scaledwidth=95.0% ] [ f : subldiagrams ] notice in particular in eq . ( [ emissionterm ] ) the current @xmath312 , which describes transverse graviton emission @xcite in the so - called h - diagram @xcite and is non - local because of the inverse laplacian needed to find it . for this reason it has been convenient to introduce the complex field @xmath296 , which corresponds to two graviton polarizations . the detailed computation shows that @xmath313 and , more generally , using a notation with real - valued indices , that @xmath314 where @xmath315 so that we have the polarizations @xmath316 the resulting equations of motion are given by [ eqs ] @xmath317 + \frac12\kappa\sqrt{s}\delta^2({\boldsymbol{x}})\delta(x^- ) \nonumber \\ & \qquad = -\kappa\left({\partial^2}\phi^*{\partial^2}_-{\partial^{*2}}\phi \;+\;(\partial\leftrightarrow\partial^ * ) \right ) \label{eqh } \\ & 4\partial_+\partial_- |\partial|^4\phi -\kappa \left [ { \partial^{*2}}{\tilde{h}}^{++}{\partial^2}{\tilde{h}}^{-- } - { |\partial|^2}{\tilde{h}}^{++}{|\partial|^2}{\tilde{h}}^{--}\right ] \nonumber \\ & \qquad= \kappa \left [ { \partial^{*2}}\left({\tilde{h}}^{++}{\partial^2}_+{\partial^2}\phi\right ) + { \partial^2}\left({\tilde{h}}^{--}{\partial^2}_-{\partial^{*2}}\phi\right ) \;+\;(\partial\leftrightarrow\partial^*)\right ] \ ; , \label{eqphi } \end{aligned}\ ] ] where the rescattering terms are set on the r.h.s . , with similar equations for @xmath318 and @xmath319 . notice that in eqs . ( [ eqs ] ) the impinging particles are represented by the frozen external sources @xmath201 and @xmath202 written in eq . ( [ enmomtens ] ) , which pertain to the collinear kinematics used before . according to this small - angle approach , scattering comes out because of a non - trivial action , but is not taken into account in the energy - momentum of the particles . shortly , we shall take a somewhat different attitude . we start recalling @xcite that , if the rescattering terms are dropped , eqs . ( [ eqs ] ) admit a shock - wave solution of the form [ separation ] @xmath320 where the metric component @xmath321 are of as type @xcite while the transverse field @xmath296 has support inside the whole _ light - wedges _ @xmath322 . the latter propagation , of retarded plus advanced type , corresponds to the principal value part of the feynman propagators , as is appropriate for the real part of the amplitude . the field equations ( [ eqs ] ) induce a set of differential equations on @xmath22 , @xmath23 and @xmath323 , which can be solved on the basis of proper boundary conditions which make them regular , i.e. , ultraviolet safe . furthermore , by replacing the expressions ( [ separation ] ) into ( [ eqs ] ) and omitting the rescattering term , we obtain the acv equations of motion and the corresponding reduced action @xmath324 \big\ } \ ; , \label{reda}\end{aligned}\ ] ] where we have freely performed some integration by parts , assuming a smooth enough ultraviolet behaviour of the solutions , and we have replaced the current @xmath312 by its reduced counterpart @xmath325 : @xmath326 we state below some of the properties of the ram solutions @xcite in the particularly simple case of axisymmetric sources like particle - ring scattering which is relevant for the azimuthal averaged particle - particle case . by assuming the real fields @xmath327 , @xmath328 and the complex field @xmath329 to be functions of @xmath330 only , and by calling @xmath331 , @xmath332 the axisymmetric sources , we soon realize that the action @xmath333 can be recast in one - dimensional form . this is because the current @xmath325 defined by eq . ( [ defhc ] ) is simply obtained in this case by setting @xmath334 where the dot denotes the @xmath335-derivative and is therefore real and axisymmetric also . as a consequence , the interaction term involves @xmath336 only and is proportional to the @xmath337 kinetic term as follows @xmath338 \qquad \left(\cdot \equiv \frac{{\mathrm{d}}}{{\mathrm{d}}r^2 } \right ) \;.\ ] ] here we have replaced @xmath323 by the auxiliary field @xmath339 a sort of renormalized squared distance defined by @xmath340 \ ; , \qquad h \equiv { \boldsymbol{\nabla}}^2\phi = 4\frac{{\mathrm{d}}}{{\mathrm{d}}r^2}(r^2\dot\phi ) = \frac{1}{(\pi r)^2}(1-\dot\rho ) \;,\ ] ] which incorporates the @xmath323-@xmath22-@xmath23 interaction . furthermore , the field @xmath323 is now taken to be real - valued , describing the @xmath341 polarization only . the external axisymmetric sources @xmath24 and @xmath342 are able to approximately describe the particle - particle case by setting @xmath343 , @xmath344 , where the azimuthal acv averaging procedure is assumed . the equations of motion , derived from the reduced - action ( [ 1dimaction ] ) and specialized to the case of particles at impact parameter @xmath4 ( with the axisymmetric sources just quoted ) , provide the profile functions @xmath345 from the analogue of eq . ( [ eqh ] ) , and the @xmath346-field ( or @xmath323-field ) @xmath347 from the analogue of eq . ( [ eqphi ] ) . ( [ eomrho ] ) shows a `` coulomb '' potential in @xmath346-space , which is repulsive for @xmath348 , acts for @xmath349 and plays an important role in the tunneling phenomenon @xcite . by replacing the equation of motion ( [ eoma ] ) into eq . ( [ 1dimaction ] ) , the reduced action can be expressed in terms of the @xmath346 field only , and takes the simple form @xmath350 \equiv -\int_0^\infty { \mathrm{d}}r^2 \ ; { \mathcal{l}}(\rho,\dot\rho , r^2 ) \;,\ ] ] which is the one we shall consider at quantum level in the following . the effective metric generated by the axisymmetric fields @xmath346 , @xmath22 and @xmath23 was calculated @xcite on the basis of the complete form of the shock - wave ( [ separation ] ) and is given by @xmath351 \nonumber \\ & + ( { \mathrm{d}}x^+)^2 \delta(x^{+ } ) \left [ 2\pi r \bar{a}(r^2 ) -{{\textstyle\frac{1}{4}}}(1-\dot\rho ) |x^{-}| \right ] \nonumber \\ & + ( { \mathrm{d}}x^-)^2 \delta(x^{- } ) \left [ 2\pi r a(r^2 ) -{{\textstyle\frac{1}{4}}}(1-\dot\rho ) |x^{+}| \right ] \nonumber \\ & + { \mathrm{d}}r^2\left[1 + 2(\pi r)^2\theta(x^{+}x^{-})\dot{\phi } \right ] + { \mathrm{d}}\theta^2\,r^2 \left[1 + 2(\pi r)^2\theta(x^{+}x^{-})(\dot{\phi } + 2r^2\ddot{\phi})\right ] \;. \label{metr}\end{aligned}\ ] ] this metric is dynamically generated and may be regular or singular at short distances , depending on the behaviour of the field solutions themselves . it is not fully consistent , however , since it does not take into account the longitudinal shifts that the fields cause to each other . in the following sections we shall compute such shifts at subleading level and we shall present the improved expression of the self - consistent metric . the first nontrivial use of the reduced action ( [ reda ] ) is the calculation of the first order correction to the einstein deflection @xcite which is due to the h - diagram contribution to the action ( fig . [ f : subldiagrams].a ) . the latter is obtained by expanding the action ( [ reda ] ) and the corresponding equations of motion ( quoted in eqs . ( [ eoma],[eomrho ] ) for the axisymmetric case ) in the parameter @xmath0 where @xmath352 is the gravitational radius and @xmath4 is the impact parameter conjugated to the transverse momentum @xmath209 ( and related to the true impact parameter @xmath115 by eq . ( [ b0 ] ) ) . by expanding the equations of motion we obtain first order corrections to the profile function @xmath353 and to the transverse field @xmath354 as consequence of the lipatov s vertex @xmath312 in eq . ( [ emissionterm ] ) . it is then straighforward to obtain for the action @xcite @xmath355 \;,\end{aligned}\ ] ] where @xmath356 and the h - diagram contribution is @xmath357 in the following , we specialize the expression ( [ hdiagram ] ) to the axisymmetric case ) ) which is due to the @xmath358 polarization and is to be subtracted out by the exponentiation procedure of ref @xcite . ] in which , according to eq . ( [ rho ] ) , @xmath359 and by the equations of motion ( [ eomrho ] ) we obtain the result @xmath360 as contribution of the @xmath341 polarization only . here we have used the regular solution of eq . ( [ eomrho ] ) @xmath361 where @xmath362 and @xmath363 is the largest real solution of eq . ( [ tbcond ] ) . the complete result , including the @xmath364 polarization would be @xcite @xmath365 and provides by stationarity of the eikonal phase the scattering angle @xmath366 \;.\ ] ] the recollection above suggests to generalize the calculation of the shifted field @xmath200 in sec . [ s : ssw ] to the next order in the @xmath188 expansion , by including the h - diagram in the eikonal . in so doing we find 3 kinds of insertions of the @xmath200 source , illustrated in fig . [ f : hdiagraminsertions ] : _ ( a ) _ insertions on the on - shell propagators , which can be done in a similar way as we did the leading one ; _ ( b ) _ insertions on the particle propagators of the h - dagram itself , and _ ( c ) _ insertions on the emitted transverse field @xmath367 . the latter are new and involve the rescattering vertex , which is omitted in the ram calculation and will be included in the next subsection . ( red cross ) on the h - diagram._,scaledwidth=90.0% ] [ f : hdiagraminsertions ] insertions of type _ ( a ) _ above add up to the leading shift @xmath368 and are done by replacing @xmath217 with @xmath369 in the corresponding energy charge . between past and future . note that in the acv equations framework the profile function @xmath134 and the shift @xmath370 occur together in the product @xmath371 ( eq . ( [ eqd ] ) ) , while in the insertion formalism @xmath372 is probed by the leading profile @xmath373 only , which dominates for @xmath374 . ] however , @xmath375 contains two @xmath217 factors and thus acquires a factor of 2 , as follows @xmath376 \nonumber \\ & = 2\pi gs ( a_0 + a_h ) + 2\pi r(a_0 + 2 a_h){\mathrm{i}}\frac{\partial}{\partial x^- } + \cdots \;.\end{aligned}\ ] ] the outcome is that the overall shift @xmath377 is proportional to @xmath378 , as given by the perturbative formula ( [ h++sh ] ) : @xmath379 and thus it extends to the h - diagram the relationship between shift and energy - derivative of the action suggested in sec . [ s : ilepp ] . consider now the full field equations , starting from ( [ eqh ] ) . while , by replacing in the l.h.s . the shock - wave ( [ separation ] ) , we generate terms proportional to @xmath380 , the rescattering term in the r.h.s . yields instead the structure @xmath381 which suggests an @xmath140 translation , i.e. , a shift similar to those just found at leading and h - diagram level with the eikonal insertions . therefore , we make the ansatz [ shifts ] @xmath382 \ ; , \label{phishift } \end{aligned}\ ] ] where @xmath383 are shift variables to be determined . we take a similar ansatz for energy - momentum too , because we know from last section the latter is really shifted at leading level . by then formally expanding the @xmath384-function ( [ hppshift ] ) in @xmath229 , we get two equations from the zeros of the @xmath384 and @xmath385 contributions . the first equation is simply the reduced - action model ( ram ) equation for @xmath22 , which reads @xmath386 + { { \textstyle\frac{1}{2}}}\delta^2({\boldsymbol{x } } ) = 0\ ] ] and reduces , in the axisymmetric case and for real - valued @xmath323 , to @xmath387 = \frac{{\mathrm{d}}}{{\mathrm{d}}r^2}\big[\dot a\rho ] = -\frac1{2\pi}\delta(r^2 ) \;.\ ] ] the latter is equivalent to eq . ( [ eomrho ] ) and yields the solution @xmath388 where we have introduced the infrared cutoff parameter @xmath70 by setting @xmath389 . since by eqs . ( [ regsol ] ) above @xcite @xmath390 @xmath327 approaches the leading value for @xmath391 @xmath392 and diverges logarithmically for @xmath393 . the second equation ( from the vanishing of the @xmath385-coefficient ) involves the rescattering terms and is supposed to determine the shift parameter @xmath252 , which occurs in the combination @xmath394 , as follows @xmath395 = -\frac{\delta_0}{2}\delta^2({\boldsymbol{x}})+(2\pi r)^2 |{\partial^{*2}}\phi|^2 \;,\ ] ] where we have set the rescattering terms on the r.h.s . and we notice in particular the energy - momentum shift which , according to sec . [ s : ssw ] , has the contribution @xmath396 emerging from the eikonal insertions up to @xmath0 accuracy . in the axisymmetric limit , ( [ eqd ] ) takes the simplified form @xmath397 where we note that the rescattering source @xmath398 vanishes for @xmath399 , as a consequence of the equations of motion ( [ rhoasy ] ) . it is then straightforward to find a solution for @xmath400 which is regular everywhere , in the form @xmath401 \displaystyle \delta_0 - \frac1{a(r^2)}\left[2\pi a(b ) \int_{b^2}^\infty { \mathrm{d}}r'{}^2\,{{j_\textsc{r}}}(r'{}^2 ) - \int_{b^2}^{l^2}\frac{{\mathrm{d}}r'{}^2}{\rho(r'{}^2 ) } \int_{r'{}^2}^\infty{\mathrm{d}}r''{}^2{{j_\textsc{r}}}(r''{}^2 ) \right ] & ( r^2\leq b^2 ) \end{cases}\ ] ] where @xmath402 in other words , the shift parameter of the longitudinal field takes over the ( constant ) leading value @xmath403 from the energy - momentum tensor and adds an @xmath335-dependent term of relative order @xmath0 which , for @xmath404 , is related to the kinetic term of the ram action @xmath405 \equiv 2\pi gs \ , { \mathcal{a}}(b ) \;.\ ] ] this means that , while the ram action is sufficient in order to describe the scattering parameters at relative order @xmath0 ( h diagram ) , the rescattering terms are needed in order to describe the form of the fields , like shock - wave shifts , and thus the metric properties at a comparable level of accuracy . it is amusing to check the result ( [ soldelta ] ) which is based on the rescattering equations and on the energy momentum shift in eq . ( [ eqd ] ) by using the direct insertions _ ( b ) _ and _ ( c ) _ on the h - diagram mentioned before , and depicted in fig . [ f : hdiagraminsertions ] . by keeping track of the flow of energy charges we calculate the above insertions on the imaginary parts ( app . [ a : hdi ] ) and we find that diagram _ ( b ) _ is already counted by lipatov s vertices , while diagram _ ( c ) _ contributes the absorptive part @xmath406 the @xmath216-dependent part , by dividing out the rapidity factor @xmath407 ( as requested in order to obtain the real part of the diagram from the dispersion relations ) and the leading field @xmath408 fits with the displacement form @xmath409 , by thus providing the shift @xmath410 from past to future , where @xmath411 we recognize here the 2-dimensional laplacian green function @xmath412 applied to the h - diagram density . by translating it to the azimuthal - averaged formalism we can use the one - dimensional form @xmath413 and apply it to the rescattering current @xmath398 . the result is just identical to the r.h.s . of eq . ( [ soldelta ] ) with the @xmath414 contribution subtracted out . in particular , for @xmath415 eq . ( [ deltahb ] ) factorizes and yields @xmath416 , thus recovering the form ( [ akin ] ) of the total shift . our overall interpretation of the present findings is that a large - distance observer will see the particle shock - waves to suffer a total time delay @xmath417 , directly related to the action and thus to the scattering angle @xmath418 because of the contribution of the rescattering shift . on the other hand , a test particle parallel to particle 2 , traveling at transverse distance @xmath419 from particle 1 at deflection time , corresponding to the perpendicular distance ( impact parameter ) @xmath420 , according to the discussion of sec . [ s : wpm ] . ] will see a profile function @xmath421 related to its own scattering angle @xmath422 which turns out to be the same in the limit @xmath423.-derivative of @xmath424 has implicit contributions which vanish by the equation of motion . ] furthermore , the test - particle will suffer a time delay that , due to the double - shift picture of sec . [ s : sdts ] , is @xmath425 , close to the short - distance value @xmath426 coming from the eikonal insertions of sec . [ s : ssw ] finally , while the rescattering contribution is needed for the large distance shifts , its contribution to the action starts at order @xmath310 because the diagram in fig . [ f : subldiagrams].c ( formally of order @xmath311 ) vanishes by the property @xmath427 . therefore it does not affect the consistency of the metric and action descriptions at order @xmath428 . was argued @xcite to vanish for light - particle scattering by analiticity and relativity arguments . ] of course , a symmetrical calculation yields the shift @xmath429 for the field @xmath199 , provided one exchanges light - cone indices @xmath430 and the profile functions and shifts @xmath431 , @xmath432 . we now go over to the second basic equation ( [ eqphi ] ) which essentially describes rescattering properties ( fig . [ f : subldiagrams].c , d ) of the transverse field @xmath21 , once emitted by the longitudinal fields . we note that , by replacing the ram fields ( [ shifts ] ) in the r.h.s . of eq . ( [ eqphi ] ) , we obtain the structure @xmath433 \times \label{eqphistr } \\ \nonumber & \left\ { -\frac{\partial}{\partial\delta}\frac12\big [ \delta(x^- -\pi r\delta)\delta(x^+ -\pi r\bar\delta_\phi ) + \delta(x^- + \pi r\delta)\delta(x^+ + \pi r\bar\delta_\phi)\big]\right\ } \;.\end{aligned}\ ] ] let us recall that , according to eqs . ( [ delta0],[soldelta ] ) , the longitudinal rescattering predicts a shift @xmath434 . we then decide to expand the shift @xmath435 of the transverse field around the constant value @xmath436 , which gives the leading order of the longitudinal shift . by this method , it is easy to evaluate a ( formally ) first order modification @xmath437 of the transverse - field shift , which however is also of leading order in the @xmath188 expansion . by expanding the first term in the l.h.s . of eq . ( [ eqphi ] ) we obtain @xmath438 \times \nonumber\\ & \frac12\big [ \delta(x^- -\pi r\delta ) \delta(x^+ -\pi r\bar\delta ) + \delta(x^- + \pi r\delta)\delta(x^+ + \pi r\bar\delta ) \big]\label{expphi}\end{aligned}\ ] ] so that we can easily match the left and the r.h.s . to get the equation , in the axisymmetric limit , @xmath439 where we work in the regime @xmath440 with @xmath441 . the expression of @xmath323 is found by the ram equation @xmath442 which was provided before ( [ eomrho ] ) . by inserting proper boundary conditions , in our regime @xmath415 we can set @xmath443 ^ 2 $ ] and tentatively look for a solution @xmath444 , where @xmath445 and @xmath446 are constants . a simple calculation shows that @xmath447 + \frac1{2\pi } \;,\ ] ] so that @xmath435 is @xmath335-dependent and of leading order , while the correction @xmath448 is sizeable . furthermore , @xmath435 becomes cutoff independent , and pretty small at @xmath423 . the above conclusion may be unpalatable from a calculational standpoint , but is natural on physical grounds because rescattering occurs for the @xmath323 field at relatively leading level and at all distances , so that the parameter @xmath403 has no particular role and a sizeable difference @xmath449 is expected . one might ask , at this point , whether the ram hierarchy for the action is really satisfied or not . fortunately it remains , because the integral over @xmath207 , @xmath140 of the rescattering vertex with the longitudinal field just vanishes ( @xmath450 ) or , in other words , there is no @xmath451 contribution to the action from diagram [ f : subldiagrams].c . furthermore , the shift does not change the @xmath452 integration in the h - diagram and therefore rescattering contributions . on the other hand , the location of the shock - wave and the evolution of particle trajectories and geodesics does change from the point of view of the metric , so that rescattering is needed in order to have self - consistent calculations even at relative order @xmath0 . all preceding arguments hold for the uv - safe solutions of the ram model in eqs . ( [ eoma],[eomrho ] ) which , by definition satisfy the condition @xmath453 . due to the form of @xmath339 in eq . ( [ rho ] ) , @xmath454 would imply that @xmath455 has a short - distance singularity , @xmath456 has an outgoing flux @xmath457 and therefore @xmath458 has a singular @xmath384-function contribution . the latter behaviour would call for large short - distance effects possibly regularized by the string which are expected , but not considered , in the ram model . restricting the solutions by @xmath453 is possible only if the impact parameter is larger than some critical value @xmath6 which signals a possible classical collapse . in fact , the regular solution ( [ regsol ] ) of eq . ( [ eomrho ] ) has continuous derivative at @xmath459 provided @xmath460 a condition which has real - valued solutions only for @xmath461 . on the other hand , the complex solutions for @xmath8 lead to an exponential damping @xcite of the @xmath1-matrix of type @xmath462 with exponent of the order of a black - hole entropy @xmath463 which leads , eventually , to a violation of unitarity of the ram model @xcite . from the above discussion , some questions arise : what happens to the rescattering solutions for @xmath8 , when they are irregular ? do we see any sign of a possible collapse in the latter , perhaps in relation with the unitarity problem and to a `` fall in the center '' mechanism @xcite ? let us then consider a class of real - valued ram solutions with @xmath454 for @xmath8 , as follows . we start from the general solution for @xmath454 @xmath464 \theta(b^2-\tau)+r^2\cosh^2\chi(\tau)\theta(\tau - b^2)\ ] ] where @xmath465 is now a free parameter which @xmath466 depends on . then we choose @xmath467 so as to minimize @xmath468 , i.e. , @xmath469 . the parameters of this `` minimal '' solution @xmath470 are given by @xmath471 the corresponding rescattering current is @xmath472 the above current is now non - vanishing for @xmath473 , and , in addition , has a non - integrable behaviour for @xmath474 . this means that large rescattering amplitudes are built in the short - distance region @xmath475 and , as a consequence , that exchange and emission of ( massive ) string states can no longer be neglected . nevertheless , we find it instructive to provide here a preliminary analysis of irregular solutions in our effective - theory framework in which only graviton intermediate states are considered . basic string effects like graviton reggeization and ensuing string production at @xmath476 @xcite , as well as diffractive and central string emission induced by tidal forces @xcite - derivatives of phaseshifts @xcite and are thus possibly important in the whole subcritical region @xmath475 where ( [ irrjr ] ) is large , so that unitarity defect @xcite and energy balance @xcite could be sizeably affected . ] should certainly be estimated in the near future in order to see how they affect the picture , but we feel that our effective approach may still provide suggestions and questions to be answered , and is anyway needed as a ground for the estimates just mentioned . in order to get a better insight on the evolution of irregular solutions , we cut - off the current ( [ irrjr ] ) below @xmath477 , and we compute the total charge , related in sec . [ s : rcram ] to the h - diagram action @xmath478 and to the shift @xmath479 as follows : @xmath480\rho(0 ) + \frac{\rho^2(0)}{\lambda_s^2 } + { \mathcal{o}\left(\lambda_s^2\right)}\right ] \ ; , \label{ahirr}\end{aligned}\ ] ] where the singular integration has been performed by a careful integration by parts and in the last equality we have introduced the @xmath384-function singularity in the @xmath346-derivative : @xmath481 this means that we think of @xmath482 for @xmath483 ( a region which is felt by the two - dimensional model through the outgoing flux ) and that the @xmath384-function occurs because of the @xmath484-discontinuity . thus , strictly speaking , we discover that the singular solutions are not _ bona fide _ solutions of the equations of motion up to @xmath485 unless we introduce a singular `` external force '' @xmath486 at @xmath487 . we can do that by starting from the action @xcite @xmath488\ ] ] and by supplementing the lipatov s current @xmath325 of eq . ( [ defhc ] ) with a singular external current @xmath489 by then replacing such expression into eq . ( [ 2dimaction ] ) , i.e. , @xmath490\left ( 2\dot{a}\dot{\bar{a}}+\frac{\rho_m(0)\delta'(\tau)}{(\pi r)^2}\right ) \;,\ ] ] we find that the action enforces eq . ( [ irrdrho ] ) and becomes , on the equation of motion , @xmath491{\mathrm{d}}\tau-\frac12\int_{0^+}^\infty \left(\frac{1-\dot\rho}{\pi r}\right)^2{\mathrm{d}}\tau + \frac12\left(\frac{\rho_m(0)}{\pi r}\right)^2\delta(0 ) \right\ } \nonumber\\ & = a(b ) - \int_{0^+}^\infty \frac{(1-\dot\rho)^2}{2\pi r^2}\;{\mathrm{d}}\tau + \frac{\rho_m^2(0)}{2\pi r^2\lambda_s^2 } \ ; , \label{irract}\end{aligned}\ ] ] thus determining the singular term in the action . as a step - function located around @xmath487 with width @xmath492 and height @xmath493 , hence the identification @xmath494 . ] we are now able to look at the scattering and shift properties of the irregular solutions , by keeping in mind that we need in this case the external current ( [ deltahc ] ) , that we think generated by the short - distance string dynamics and/or by possibly collapsed matter . compared to the complex solutions which have a quantum - tunneling interpretation and cause a probability suppression the irregular ones may provide a probability source or alternatively may carry away the information loss . by thus using the eikonal representation of the @xmath1-matrix with the action ( [ irract ] ) and by carefully computing @xmath4- and @xmath71-derivatives , we get the scattering angle @xmath495 and the time delay @xmath496 \equiv 2\pi r\delta_0 \;,\ ] ] where , by eqs . ( [ irrsol],[deltahc ] ) @xmath497 \;.\ ] ] such results can be further specified by using @xmath498 and the outcomes are plotted in fig . [ f : irraction ] . without the @xmath468 contribution ; _ ( b ) _ scattering angle as derived by eq . ( [ irrtheta ] ) . in both plots @xmath499 ; lengths are measured in units of @xmath500._,title="fig : " ] without the @xmath468 contribution ; _ ( b ) _ scattering angle as derived by eq . ( [ irrtheta ] ) . in both plots @xmath499 ; lengths are measured in units of @xmath500._,title="fig : " ] [ f : irraction ] we note the strong increase of the action in the region @xmath501 due to the positive singular contribution implying in particular a motion with many turns in the region @xmath502 . actually , large scattering angles are reached pretty soon around @xmath503 as solutions of the equation @xmath504 \;,\ ] ] but , due to the strong increase of the action derivative ( fig . [ f : irraction ] ) , the saddle - points are confined to either the critical region @xmath505 or to the small-@xmath4 region @xmath506 , for all real values of @xmath507 . finite @xmath4-values , in the region @xmath508 are strongly suppressed like @xmath509 . the above observation suggests that the singular solutions under study do not actually yield back all the initial information in the physical scattering region , but may carry it away in the small-@xmath4 , string - dominated region . another piece of information comes from the study of the shifts and the corresponding time delay . we first notice , by eq . ( [ irrdelay ] ) , that the short - distance shift of @xmath510\frac{2\rho(b^2)}{2\pi r^2}\ ] ] has a positive singular term meaning a lot of time spent in the interaction region which is essentially twice the singular contribution to @xmath511 in eq . ( [ ahirr ] ) . the factor of 2 comes from @xmath468 scaling like @xmath512 similar to what happens for the h - diagram itself . furthermore , we already know from sec . [ s : rcram ] that the large - distance shift takes contributions from rescattering insertions and produces the additional ( negative ) shift @xmath479 , the one we started with in eq . ( [ deltah ] ) . we thus have @xmath513 we note that the singular term remains positive after subtraction of @xmath479 and of the same order as that of the action apart for the addition of a term proportional to @xmath514 . the very large time delay is thus confirmed at all distances . that means that the singular solutions , wherever they are , spend a long time in the interaction region , before exiting , either around @xmath503 , or at @xmath515 . therefore , they are connected with long - lived states in that region . finally , let us look at the associated metric ( [ metr ] ) , whose geodesics in the `` regular '' case @xmath453 were argued to provide a complementary picture of particle scattering . for the irregular solutions of eqs . ( [ irrsol],[minpar ] ) , the relevant metric coefficients ( inside the light - wedges @xmath516 ) have the form [ gmunu ] @xmath517 where the arrows label the simplified expressions obtained in the @xmath518 limit , which will be used for the purpose of the qualitative discussion below . we note that , for @xmath454 , @xmath519 shows a puzzling zero at @xmath520 , which causes a change of signature of the @xmath106-dimension for @xmath521 a feature to be taken with great caution because our perturbative identification of the metric coefficient is probably invalid in the strong - coupling region we are interested in . nevertheless , let us take the expressions ( [ gmunu ] ) at face - value , and discuss the ensuing geodetic flow , in the transverse plane , which is strongly affected by the @xmath323-field singularity @xmath522 . in order to do that , we shall introduce the test - particle lagrangian ( per unit mass ) @xmath523 with corresponding momenta and hamiltonian ( per unit mass ) [ momenta ] @xmath524 it is convenient to define the momenta per unit energy @xmath525 , so as to allow the treatment of the massless case too . by evaluating the constants of motion in the asymptotic region @xmath526 one finds that @xmath527 , @xmath528 and @xmath529 , where @xmath530 , @xmath531 and @xmath532 are the longitudinal , transverse and total ( asymptotic ) velocity , while @xmath533 is the impact parameter of the test particle w.r.t . the @xmath272 axis . the test - particle motion is eventually described by the constants of motion @xmath534 , @xmath535 , @xmath536 and by the radial `` effective potential '' @xmath537 \longrightarrow -\left(1-\frac{r^*{}^2}{r^2}\right ) v_\perp^2\left(1 -\frac{\beta^2}{r^2+r^*{}^2}\right ) \;,\end{aligned}\ ] ] since @xmath538 . ( solid blue ) , @xmath539 ( dashed red ) , @xmath540 ( dotted green ) , @xmath541 ( dash - dotted brown ) . _ ] [ f : potential ] the @xmath542 geodesics are characterized by the fact that the transverse velocity @xmath531 ( if present ) is purely radial and the potential ( [ potential ] ) is purely repulsive ( fig . [ f : potential ] ) , and would be typical of an angular momentum barrier with @xmath543 that by lack of other explanations we should attribute to the external source @xmath544 in eq . ( [ deltahc ] ) that we have put in at @xmath545 in order to justify the irregular solutions themselves . for @xmath546 the repulsive potential is absent , and the geodesics ( at @xmath547 , say ) will reach the shock - wave ( at @xmath548 , say ) at the initial distance @xmath106 , but will thereby acquire a shift and will be deflected according to the formula @xmath549 which takes into account the `` refraction index '' of the @xmath323-field . once again @xmath550 is unreachable : if initially we try to approach it , the geodesic scattering angle approaches @xmath551 , and the test particle is reflected backwards . we note at this point that the situation does change unexpectedly when @xmath552 increases and overcomes the threshold @xmath553 : in this case , according to eq . ( [ potential ] ) , the centrifugal factor at the origin changes sign , so that the potential becomes partly attractive , close to the origin and above @xmath553 ( fig . [ f : potential ] ) . in particular , for @xmath540 the potential is everywhere attractive ! this somewhat surprising feature opens up the possibility that test particles may reach the small-@xmath106 region , at least at quantum level , because the small barrier has a transmission coefficient of order unity around the value @xmath540 , and is thus able to populate the small-@xmath106 region around @xmath476 in an efficient way . that feature is the counterpart of the selection of possible issues of the particles themselves , which was argued before to have sizeable probability for a small-@xmath4 exit on the basis of the action stationarity condition . furthermore , the test - particle action variable @xmath554 develops a logarithmic behaviour in the small-@xmath106 region with frequency @xmath555 : is that perhaps a possible interpretation of the time nature of the variable @xmath106 in that region ? on the whole , we think , the above discussion shows at a sort of `` phenomenological '' level that the irregular solutions may indeed vehicle the particles information from distances of order @xmath11 down to the string size , thus suggesting that the unitarity loss of the @xmath453 model is due to the opening up of the gate to other worlds . however , due to the lack of the theoretical ingredients from string theory mentioned in the beginning , in the present formulation we have been unable to really assess the reliability of the above suggestion and to discuss the physical interpretation of the state(s ) , living at @xmath476 , which could be responsible for the singular contributions to the action investigated here and for their intriguing consequences . to sum up , in our study of the improved eikonal model with subleading correction we have found that : * the source particles , after the interaction , besides being deflected , suffer a time delay which can be interpreted as a shift in the light - cone variables ( sec . [ s : ilepp ] ) ; * the shock - wave metric fields @xmath556 generated by the source particles are also shifted in an analogous way , after the interaction described by the eikonal amplitude ( sec . [ s : ssw ] ) ; * the subleading contributions to the eikonal , represented by the ( multi ) h - diagram and by the rescattering terms , provide corrections to the scattering angle of the particles and to the shifts of the metric fields @xmath556 and @xmath296 ( sec.[s : rcram ] ) . due to rescattering , the large - distance shift of the particle fields agrees with the scattering - angle description based on the action . in order to find an adequate form for the self - consistent metric which embodies all these features , we have to modify the acv expression ( [ metr ] ) . as far as the shock waves are concerned , the improved expressions are obtained from eq . ( [ hppr ] ) by replacing the leading profile function @xmath373 with the full profile function @xmath134 derived in sec . [ s : hd ] and by accordingly shifting the location of the shock waves of the amount @xmath557 as explained in sec . [ s : rssf ] . in turn , the support @xmath558 of the transverse field @xmath296 is also shifted from the region @xmath516 to the two ( past and future ) disconnected regions @xmath559 where in the region @xmath560 after the interaction we have used the rotated coordinates of eq . ( [ rotation ] ) . equivalently , the characteristic function of @xmath561 reads @xmath562 the ensuing improved metric we propose is thus @xmath563 \nonumber \\ & + ( { \mathrm{d}}x^+)^2 \delta\big(x^{+}-\pi r { \epsilon}(x^-)\bar\delta({\boldsymbol{x}})\big ) \left [ 2\pi r \bar{a}({\boldsymbol{x } } ) -{{\textstyle\frac{1}{4}}}(1-\dot\rho ) |x^{-}-\pi r { \epsilon}(x^+)\delta({\boldsymbol{x}})| \right ] \nonumber \\ & + ( { \mathrm{d}}x^-)^2 \delta(x^{-}-\pi r { \epsilon}(x^+)\delta({\boldsymbol{x}})\big ) \left [ 2\pi r a({\boldsymbol{x } } ) -{{\textstyle\frac{1}{4}}}(1-\dot\rho ) |x^{+}-\pi r { \epsilon}(x^-)\bar\delta({\boldsymbol{x}})| \right ] \nonumber \\ & + { \mathrm{d}}r^2\left[1 + 2(\pi r)^2\chi({\sigma})\dot{\phi } \right ] + { \mathrm{d}}\theta^2\,r^2 \left[1 + 2(\pi r)^2\chi({\sigma})(\dot{\phi } + 2r^2\ddot{\phi})\right ] \;. \label{metrica}\end{aligned}\ ] ] it takes into account all the physics results following the analysis of the improved eikonal model and the acv effective action , and is self - consistent , in the sense that the trajectory shifts caused by the shock - wave metric are felt by the metric itself . of the metric field @xmath296 ( yellow ) . the ( pink ) rectangle @xmath564 actually corresponds to two different regions of space - time , which are accessed either from the left or from the right . _ ] [ f : wedgeslice ] a word of caution has to be spent concerning the form of the metric in eq . ( [ metrica ] ) in the region between the forward and past light - wedges . in fact , the set @xmath565 , @xmath566 enclosed by the continuation of the shock waves represented by a rectangle @xmath564 for each longitudinal section at fixed @xmath142 as depicted in fig . [ f : wedgeslice ] actually represents two distinct physical regions in our coordinate system . each region is unambiguously described by continuing the metric either from the left or from the right . for instance , by continuing the metric from the left across the boundary @xmath567 , we enter the region at the left of both shocks , where shock 1 travels from @xmath44 to @xmath568 , collides with shock 2 in @xmath568 , and after that shock 2 moves from @xmath568 to @xmath536 and goes on in the @xmath140 direction , as shown . on the other hand , either continuation leads to the same form of the metric inside the future light - wedge @xmath560 . we have also investigated here the scattering features of uv - sensitive solutions of the model in the subcritical range @xmath569 , without making a real attempt to restore the string degrees of freedom which are nevertheless called for by the singular short - distance behaviour of the rescattering current . this preliminary analysis leads to the suggestion that such solutions may indeed carry the information from distances of order @xmath11 to distances of order @xmath476 , but a dynamical analysis in the string - dominated region is needed to find out about their fate . it is a pleasure to thank gabriele veneziano for various exchanges on the topics presented here , and domenico seminara for useful discussions . we start noticing that the field - insertions we shall consider are basically either _ ( a ) _ external ( fig . [ f : hdiagraminsertions].a ) or _ ( b ) _ internal ( fig . [ f : hdiagraminsertions].b , c ) . here we shall concentrate on the final external insertion kinematics which in the frame characterized by initial particle s momenta @xmath570 and final momenta @xmath571 is able to provide the full shift @xmath572 from past to future . the field @xmath573 is coupled to the charge @xmath574 and thus the insertion factor in front of the amplitude @xmath575 is @xmath576 \left(q^- - \frac{{\boldsymbol{q}}^2-{\mathrm{i}}{\varepsilon}}{q^+}\right ) q^+ } \frac{{\mathrm{d}}^2{\boldsymbol{q}}}{(2\pi)^2 } { \mathrm{e}^{{\mathrm{i}}{\boldsymbol{q}}\cdot{\boldsymbol{x}}}}\ ] ] where @xmath577 and @xmath578 are both taken to the positive and larger than @xmath579 . the value of @xmath580 is fixed by the mass - shell and the @xmath581-integration is done in the lower half - plane because @xmath582 picking up the pole at @xmath583 . the result is @xmath584 { \mathrm{e}^{{\mathrm{i}}{\boldsymbol{q}}\cdot{\boldsymbol{x } } } } \label{htpp2}\end{aligned}\ ] ] here we have kept for completeness the soft - emission contribution at @xmath585 also which is however suppressed by the phase factor exponent @xmath586 in the fast - emitting particle kinematics we are using . the dominant contribution , due to the @xmath587 pole , has a small phase factor @xmath588 , and thus it provides the @xmath276 result that we have anticipated in the text . limit the whole contribution ( [ htpp ] ) vanishes for @xmath589 , because the @xmath581-integrand has both poles in the same half - plane . ] note also that the @xmath204-translation in the @xmath590 denominator is due to the scattering angle implied by the @xmath591 momentum transfer , which in a single - hit process is small , @xmath592 and is thus negligible . on the other hand , in the full scattering process we have that @xmath593 is sizeable , and the shock - wave is correspondingly rotated , as in eqs . ( [ hr],[hppr ] ) of the text . therefore , for insertion on the eikonal lines in which @xmath594 and the amplitude factor is @xmath595 , we recover the insertion factor @xmath596 that we have repeatedly used in the text . here @xmath597 is the leading profile function of eq . ( [ a0 ] ) and the possible rotation along the scattered beams is understood . in the case of the h - diagram , besides the insertion _ ( a ) _ on external lines ( already computed in the text ) we have in principle two insertions , one of which is on the internal ( eikonal ) propagators . we think that this one ( fig . [ f : hdiagraminsertions].b ) is already included by the lipatov s vertices , which by definition count the external particles emissions also . therefore , in order to avoid double - counting we only estimate here the insertion on the rescattering graviton propagator of momentum @xmath598 in fig . [ f : hdiagraminsertions].c , as emitted by lipatov s vertices . by going through the same steps as before , we perform the @xmath581-integration first ( by keeping the leading @xmath587 contribution only ) and we have , in addition , a weighted @xmath600-integration , as follows : @xmath601 in the strong ordering region @xmath602 we estimate the effect of the ( small ) @xmath204-translation ( or @xmath591-dependent rotation ) by redefining @xmath603 and neglecting the translation in the matrix element @xmath604 . the integral in @xmath605 becomes @xmath606 by thus factorizing the exponential @xmath607 which modifies the subsequent @xmath578 and @xmath216 integrations . by performing the latter in the phase - space @xmath608 , we obtain in turn the factor @xmath609 \nonumber \\ & \simeq \log(\sqrt{s}b ) \;{\mathrm{i}}\partial_- \delta(x^- ) \times \left[1+{\mathcal{o}\left(\frac1{\sqrt{s}b}\right)}\right ] \;.\end{aligned}\ ] ] we can see here the small shock - wave rotation @xmath610 which has been neglected in front of the logarithmic phase space . by then collecting the various factors we obtain the result @xmath611 \;,\ ] ] where we have defined @xmath612({\boldsymbol{x } } ) & \equiv \pi r^2 y\int{\mathrm{d}}^2 { \boldsymbol{x}}'\;|{\tilde{h}}({\boldsymbol{x}}')|^2 g_0({\boldsymbol{x}}-{\boldsymbol{x } } ' ) \stackrel{r\gg b}{\simeq } a_0(r)\cdot\im a_h(b ) \ ; , \\ \im a_h & = \frac{2y}{\pi}\re a_h \;. \nonumber\end{aligned}\ ] ] this shift of the imaginary part and the corresponding one for the real part add up to the total shift @xmath613 of both real and imaginary parts due to the external line insertion @xmath614 and thus occurs at large distances only with a relative minus sign , as emphasized in the text . 99 d. amati , m. ciafaloni and g. veneziano , `` superstring collisions at planckian energies , '' phys . b * 197 * ( 1987 ) 81 ; `` classical and quantum gravity effects from planckian energy superstring collisions , '' int . j. mod . phys . a * 3 * ( 1988 ) 1615 ; `` can space - time be probed below the string size ? , '' phys . b * 216 * ( 1989 ) 41 . d. j. gross and p. f. mende , `` the high - energy behavior of string scattering amplitudes , '' phys . b * 197 * ( 1987 ) 129 ; `` string theory beyond the planck scale , '' nucl . b * 303 * ( 1988 ) 407 . d. amati , m. ciafaloni and g. veneziano , `` higher order gravitational deflection and soft bremsstrahlung in planckian energy superstring collisions , '' nucl . b * 347 * ( 1990 ) 550 . d. amati , m. ciafaloni and g. veneziano , `` effective action and all order gravitational eikonal at planckian energies , '' nucl . b * 403 * , 707 ( 1993 ) . d. amati , m. ciafaloni and g. veneziano , `` towards an @xmath1-matrix description of gravitational collapse , '' jhep * 0802 * ( 2008 ) 049 [ arxiv:0712.1209 [ hep - th ] ] . m. ciafaloni and d. colferai , `` s - matrix and quantum tunneling in gravitational collapse , '' jhep * 0811 * ( 2008 ) 047 [ arxiv:0807.2117 [ hep - th ] ] . m. ciafaloni and d. colferai , `` quantum tunneling and unitarity features of an @xmath1-matrix for gravitational collapse , '' jhep * 0912 * ( 2009 ) 062 [ arxiv:0909.4523 [ hep - th ] ] . p. c. aichelburg and r. u. sexl , `` on the gravitational field of a massless particle , '' gen . rel . grav . * 2 * ( 1971 ) 303 . a. einstein , l. infeld and b. hoffmann , `` the gravitational equations and the problem of motion , '' annals math . * 39 * ( 1938 ) 65 . + a. einstein and l. infeld , `` the gravitational equations and the problem of motion . 2 . , '' annals math . * 41 * ( 1940 ) 455 . m. ciafaloni , d. colferai and g. falcioni , `` unitarity alternatives in the reduced - action model for gravitational collapse , '' jhep * 1109 * ( 2011 ) 044 [ arxiv:1106.5628 [ hep - th ] ] . g. t hooft , `` graviton dominance in ultrahigh - energy scattering , '' phys . b * 198 * ( 1987 ) 61 . d. amati , m. ciafaloni and g. veneziano , private communication . l. n. lipatov , `` high - energy scattering in qcd and in quantum gravity and two - dimensional field theories , '' nucl . phys . b * 365 * ( 1991 ) 614 . r. kirschner and l. szymanowski , `` effective action for high - energy scattering in gravity , '' phys . rev . d * 52 * ( 1995 ) 2333 [ arxiv : hep - th/9412087 ] . l. n. lipatov , `` multi - regge processes in gravitation , '' sov . jetp * 55 * ( 1982 ) 582 [ zh . * 82 * ( 1982 ) 991 ] . g. dappollonio , p. di vecchia , r. russo and g. veneziano , `` high - energy string - brane scattering : leading eikonal and beyond , '' jhep * 1011 * ( 2010 ) 100 [ arxiv:1008.4773 [ hep - th ] ] . s. b. giddings , d. j. gross and a. maharana , `` gravitational effects in ultrahigh - energy string scattering , '' phys . rev . d * 77 * ( 2008 ) 046001 [ arxiv:0705.1816 [ hep - th ] ] . v. s. rychkov , private communication .

galaxy clusters are extensively observed in optical , x - ray and radio bands . in the radio band , a cluster can be observed in the the rayleigh - jeans side of the cosmic microwave background spectrum , as a dip in the brightness temperature , due to sunyaev - zeldovich effect ( sz ) ( sunyaev & zeldovich 1972 ; for a comprehensive review , see birkinshaw 1999 ) . the sz distortion appears as a decrement for wavelengths @xmath4 mm ( frequencies @xmath5 ghz ) and as an increment for wavelengths @xmath6 mm . sz effect has the advantage that the sz intensity , unlike that of the x - ray , does not suffer from the @xmath7 cosmological dimming . as discussed by numerous authors ( birkinshaw & hughes 1994 ; silverberg et.al . 1997 ) , one can combine the x - ray and radio observations for clusters to determine cosmological parameters . this has been done in the recent years to determine the hubble constant @xmath0 ( birkinshaw 1999 ) . the sz signal is , however , weak and difficult to detect . recent high signal to noise detection have been made over a wide range in wavelengths using single dish observations : at radio wavelengths ( herbig et al , 1995 , hughes & birkinshaw 1998 ) , millimeter wavelengths ( holzapfel et al . 1997 , pointecouteau et al . 1999 ) and submillimeter wavelengths ( komatsu et al . 1999 ) . interferometric observations have also been carried out to image the sz effect ( jones et al . 1993 , saunders et al . 1999 , reese et al . 1999 , grego et al . other than estimating @xmath8 combining sz and x - ray data , sz effect alone can also be used to determine the cosmological mass density @xmath9 of the universe ( bartlett & silk 1994 , oukbir & blanchard 1992 , 1997 , blanchard & bartlett 1998 ) . however , these procedures generally assume the cluster gas is spherical , unclumped and isothermal . almost all clusters , however , show departures from these simplistic assumptions with some to a large extent . departures from these simple assumptions can lead to systematic errors in the determination of the different cosmological parameters ( inagaki et al 1995 ) , especially , it was seen that non - isothermality of the cluster can lead to a substantial error in values of the cosmological parameters . temperature structure in a cluster can be the result of the shape of the gravitational potential ( navarro et al 1997 , makino et al 1998 ) , or it can arise due to the fact that the initial falling gas in the cluster potential is less shock heated than the later falling gas ( evrard 1990 ) . in fact hydrodynamical simulations of isolated clusters also show a definite temperature structure and can introduce error in the value of @xmath0 when compared to the traditional isothermal @xmath10-models ( yoshikawa 1998 ) . roettiger et al ( 1997 ) have shown that cluster mergers can result in deviations from both sphericity and isothermality . observationally , the main handicap arises from the fact that the thermal structure of clusters are hard to measure , and the temperatures generally taken in analyses are the x - ray emission weighted temperature , usually measured over a few core radii . so , in general an isothermal description of the cluster is taken ( or sometimes a phenomenological temperature model based on the coma cluster : for example see eq 73 of birkinshaw 1999 ) . in this paper , we study another important phenomenon that can substantially change the temperature structure , _ viz . _ a cooling flow . cooling flows in clusters of galaxies ( for an introduction , see fabian et al 1984 ) is a well established fact by now , and it is seen that around @xmath11 of clusters exhibit cooling flows in their core with @xmath12 of them having cooling flows of more than @xmath13 ( markevitch et al 1998 , peres et al 1998 , allen et al 1999 ) . in the largest systems , the mass deposition rate can be as high as @xmath14 ( allen 2000 ) . the idealised picture of a cooling flow is as following : initially when the cluster forms , the infalling gas is heated from gravitational collapse . with time this gas cools slowly and a quasi hydrostatic state emerges . however in the central region , where energy is lost due to radiation faster than elsewhere , an inward ` cooling flow ' initially arises due to the pressure gradient ( fabian 1994 ) . this can modify the sz decrement and act as a systematic source of error in the determination of the cosmological parameters . schlickeiser ( 1991 ) has shown that free - free emission from cold gas in the cooling flow can actually lead to an apparent decrease of the sz effect at the centre . since the central cooling flow region is generally very small , the isothermal @xmath10-model of cluster gas can still be used for the majority of cluster region even for cooling flow cluster , with the extra precaution of excluding the central x - ray spike from the x - ray fit , and a corresponding change made in the fitting of the sz decrement . this is only possible for nearby clusters , however with well resolved cluster cores . naively , the change in the central sz decrement @xmath15 can be seen as follows : for a non cooling flow cluster , the central decrement is given by the line of sight integral of the electron pressure through the cluster centre along the full extent of the cluster . if the cluster has a maximum radius @xmath16 , then the central sz decrement at rj wavelengths can be written as @xmath17 . for a cluster with a cooling flow , let us suppose that the electron pressure @xmath18 , drastically falls below a certain radius @xmath19 , which is typically well inside the core of the cluster . the resulting central decrement is then @xmath20 . depending on the distance of @xmath19 from the cluster centre [ @xmath21 to @xmath22,there wiil be a change in the value of @xmath15 by @xmath23 . however , this simplistic view may not be true . this estimate assumes that the pressure profile remains a @xmath10 profile outside the radius @xmath19 . the pressure profile , however , need not follow the @xmath10 profile once cooling flow starts and it can deviate from it substantially even for radii much larger than @xmath19 . as a matter of fact , there can actually be an increase in the pressure for a large region inside the cooling flow , before a sudden drop inside @xmath19 . since the usual proecedure for estimating the hubble constant depends on fitting @xmath10 profiles to the sz and x - ray profiles , to estimate @xmath24 , this change in the pressure profile due to cooling flow can distort the estimation of @xmath24 and hence , the value of @xmath8 in a non - trivial way . we study this effect in detail in later sections . in this paper , we have investigated the problem of cooling flow induced change in the temperature and density profile , its effect on the sz effect , and its subsequent effect on the determination of cosmological parameters . in 2 we briefly review the physics of sunyaev zeldovich effect ; 3 is devoted to the physics of cooling flows and discussing the cooling flow solutions ; in 4 we look at the effect of cooling flow solutions on sz effect , both on the determination of @xmath0 and @xmath9 ; we conclude in 5 with a brief comment on how this work differs from other work and the relevance of this paper . for the sz effect , our notation and approach mainly follows that described in barbosa et al(1996 ) . the integral of the electron pressure along any line - of - sight through the cluster determines the magnitude of the distortion of the apparent brightness temperature of the cosmic microwave background ( cmb ) due to sz effect . this is quantified in terms of the compton @xmath25-parameter : @xmath26 where @xmath27 is the boltzman constant , @xmath28 is the gas temperature , @xmath29 is the electron rest mass , @xmath30 is the electron number density , @xmath31 is the velocity of light and @xmath32 is the thomson scattering cross - section . this occurs through the inverse - compton scattering , by the hot intracluster gas , of the cmb photons propagating through the cluster medium , and the energy transfer in this interaction between hot electrons and cmb photons resulting in a distortion to the cmb spectrum . the sz surface brightness at a position @xmath33 of the cluster with respect to the mean cmb intensity is given by @xmath34 _ x _ is a dimensionless frequency parameter @xmath35 where @xmath36 is the planck constant , @xmath37 is the observing frquency and @xmath38 is the cmb temperature at the present epoch : @xmath39k . the function @xmath40 describes the spectral shape of the effect @xmath41.\ ] ] since the total photon number is conserved in the inverse compton scattering process , upscattering of the photons , the spectral dependance gets an unique shape , through a decrement in the brightness temperature at lower frequencies while an increase is observed at higher frequencies . the sunyaev - zeldovich effect provides a unique observing approach to traditional methods which use x - ray temperature and x - ray luminosity . the x - ray studies have a major disadvantage because of ` cosmological dimming ' , the surface brightness of distant x - ray sources falls off as @xmath42 , and for this reason , obtaining samples of clusters at cosmological distances is challenging . the sz effect has the distinct advantage of being independent of the distance to the cluster . the sz flux density from a cluster will diminish with distance to the cluster as the square of the angular - size distance ; in contrast to x - ray flux densities from clusters which diminish as the square of the luminosity distance to the cluster . if the observations resolve clusters , particularly at lower redshifts , the observed sky sz temperature distribution will be sensitive to the thermal electron temperature structure within the clusters ; once again this may be contrasted with x - ray emission images of cluster gas distributions which are mainly sensitive to the gas density distribution . the method for the determination of hubble constant using sunyaev - zeldovich effect uses two observable quantities : 1 ) @xmath43 of the cmb due to sz effect ; 2 ) the x - ray surface brightness @xmath44 of the cluster . these can be written as ( r ) = -2_l_min^l_maxk_b t_e _ t n_e dl , s_x(r ) = 1 _ l_min^l_max dl_x dl , where @xmath45 is the distance to the line of sight from the cluster centre , @xmath46 and @xmath47 give the extension of the cluster along the line of sight , @xmath48 is the x - ray emissivity and @xmath49 the line element along the line of sight . the x - ray emissivity in the frequency band @xmath50 to @xmath51 can be written as = n_e^2 ( t_e;_1 , _ 2 , z ) , where ( t_e;_1 , _ 2 , z ) = 2 ^ 1/2 16e^6 a ( t_e;_1 , _ 2 , z ) , where @xmath52 du .\end{aligned}\ ] ] in the above equations we have assumed primordial abundance of hydrogen and helium and have set @xmath53 , @xmath54 is the electron charge , @xmath55 , @xmath56 , and @xmath57 is the velocity averaged gaunt factor for the ion of charge ze ( kellog 1975 ) . traditionally , to model the cluster gas distribution one takes the following density and temperature profiles ( cavaliere & fusco - femiano 1978 ) n_e(r ) = n_e0 ^ -3/2 , t_e(r ) = t_iso = constant , where @xmath58 is the central electron density and @xmath24 the core radius of the cluster . the above expressions are used as an empirical fitting model , and the parameter @xmath59 is regarded as the fitting parameter . the equation holds for @xmath60 , where @xmath61 is the maximum ` effective ' extension of the cluster . conventionally , @xmath62 , and then from equations ( 5),(6),(10),(11 ) , we get @xmath63}^ { 1/2 - 3\beta /2 } , \end{aligned}\ ] ] @xmath64}^{1/2 - 3\beta } , \end{aligned}\ ] ] where @xmath65 is the gamma function . since both the central cmb decrement and the x - ray surface brightness are observed , one can then combine equations ( 12 ) and ( 13 ) to estimate the core radius as r_c , est & = & ^2_obs ( 3 - 1/2)(3/2)^2 + & & m_e^2 c^4 ^-1/2 . in the above equation @xmath66 is the angular core radius observed in x - ray , and @xmath67 is the x - ray flux averaged temperature ( obtained from fitting the observed x - ray spectrum to the theoretical spectrum expected from isothermal case ) . this x - ray emission weighted temperature is given by t_iso . the point to be noted is that @xmath68 is the virial radius of the cluster and its choice depends on the observer . if the temperature has a spatial structure then the @xmath69 inferred from such a procedure may give different values depending on how much of the cluster is taken in making the above average . it has been seen ( yoshikawa et al 1998 ) , that this can lead to a substantial change in sz effect inferred , and thus to the value @xmath0 . the angular diameter distance @xmath70 can be approximated for nearby ( @xmath71 ) clusters as d_a = cz . thus finally we have got an estimate of @xmath72 as @xmath73 . now , if from other observations we know the cosmological parameters @xmath9 and @xmath74 , then one can estimate the hubble constant as can be seen from the above equations , the value of @xmath0 depends crucially on the many assumptions of isothermality and @xmath10-model density distribution of the cluster . cooling flow changes both of these and so it can significantly affect the value of the hubble constant . from x - ray spectra of clusters , it is known that the continuum emission is thermal bremsstrahlung in nature and originates from diffuse intracluster gas with densities @xmath75 and temperatures around @xmath76 k. the gas is usually believed to be in hydrostatic equilibrium . if , however , the density in the inner region is large enough , so that the cooling time is less than the age of the cluster , then there is a ` cooling flow ' ( fabian et al 1984 and references therein ) . of course there would be a flow only when the dynamical time is also shorter than the cooling time scale ( @xmath77 ) . the basic equations of cooling flow are : & = & -u _ b d _ b dr - 2 u r - _ b , + udu+ 1 d(_b ) & = & - gm_t(r ) , + ud(3 ) - u d_b & = & ( ) , [ eq : flow ] where @xmath78 , @xmath79 is the mean molecular weight , and @xmath80 is the proton mass . for steady flows with constant mass flux , this implies @xmath82 for steady flows . ( note that in cooling flows both @xmath83 and @xmath84 are negative . the subscripts @xmath85 refers to baryons and @xmath86 refers to total i.e baryons + dark matter . however , we assume the baryonic contribution to the total mass negligible w.r.t to the dark matter contribution . ) @xmath87 describes the distribution of the total mass and depends on the details of dark matter density profiles ( see below ) . @xmath88 is the cooling function defined so that @xmath89 is the rate of cooling per unit volume . we use an analytical fit to the optically thin cooling function as given by sarazin & white ( 1987 ) , & & ( ( ) 10 ^ -22 erg cm^3 s^-1 ) = 4.7 + & + & 0.313 t^0.08 + & + & 6.42 t^-0.2 + & + & 0.000439 t^0.35 . this fit is accurate to within @xmath90% accuracy , for a plasma with solar metallicity , within @xmath91 k. for @xmath92 k , it underestimates cooling by a factor of order unity ( compared to the exact cooling function , as in , e.g. , schmutzler & tscharnuter , 1993 ) , and therefore is a conservative fit to use , as far as the effect of cooling is concerned . for non - steady flows , we adopt the formalism of white & sarazin ( 1987 ) , where the mass deposition rate , @xmath93 , is characterised by a ` gas - loss efficiency ' parameter @xmath94 . one writes @xmath95 where @xmath96 is the local isobaric cooling rate ( @xmath97 ) . it has been found that @xmath98 models can reproduce the observed variation of mass flux ( @xmath99 ) ( sarazin & graney 1991 ) . fabian ( 1994 ) has noted that these models of white & sarazin ( 1987 ) yield good approximations to the emission weighted mean temperature and density profiles for cooling flow clusters . we also note that rizza et al ( 2000 ) have used the steady flow models of white & sarazin ( 1987 ) to simulate cooling flows . we first discuss cooling flows with @xmath100constant . with @xmath101 , one can eliminate the density from eq . [ eq : flow ] to get two differential equations : & = & u + ddr & = & 2 [ eq : steady ] these equations have singularities at the sonic radius @xmath19 where @xmath102 . a necessary condition of singularity is that the numerators of eq . [ eq : steady ] vanish at the sonic radius . therefore ( mathews & bregman 1978 ) r_s = ( 3/10_s ) we have used two different dark matter profiles for the cluster . the first model ( model a ) has been discussed earlier in the literature in the context of cooling flows in cluster ( white & sarazin 1987 ; wise & sarazin 1993 ) with a density profile , _ d=\ { ll _ o 1+(r / r_core)^2 + _ o , g 1+(r / r_c , g)^2 & + _ o 1+(r / r_core)^2 & . here @xmath103 gm @xmath104 and @xmath105 kpc describe the profile of the cluster mass , and @xmath106 gm @xmath104 and @xmath107 kpc describe the profile of the galaxy in the centre of the cluster . model b does not have the galaxy in the center , and so it is described simply by @xmath108 $ ] . with these assumptions , the solutions for steady cooling flows , @xmath100 constant ) are fully characterized by ( 1 ) the inflow rate , @xmath84 , and ( 2 ) the temperature of the gas @xmath109 at the sonic radius @xmath19 . obviously , the cooling flow solutions are only valid within the cooling radius @xmath110 where @xmath111 . we assume a value of @xmath112 gyr for all models . we assume that outside the cooling radius , gas obeys quasi - hydrostatic equilibrium ( sarazin 1986 ) . although this means matching the cooling flow solutions with nonzero @xmath83 to @xmath113 solutions outside , in reality the velocity of gas at the cooling radius is very small ( for a @xmath114 m@xmath115 yr@xmath116 with @xmath117 gm @xmath104 , at @xmath118 kpc implies a velocity of @xmath119 km s@xmath116 ) , which is close to the limit of turbulence in the cluster gas ( jaffe 1980 ) , and smaller than the sound velocity ( @xmath120 km s@xmath116 ) . the velocity of the flow at the cooling radius is , therefore , for all practical purposes , sufficiently small to be matched to the solution of hydrostatic equilibrium outside . ( in this approach , we avoid the time consuming search for the critical value of @xmath84 for which the flow solutions behave isothermally at @xmath121 ( see sulkanen 1993 ) . ) as in the usual assumptions for the interpretation of sz effect , we assume that the gas outside the cooling radius is isothermal , with a constant temperature profile . the density , therefore , obeys @xmath122^{-3 \beta /2}$ ] , where @xmath123 , and @xmath69 is the temperature of the gas at and outside the cooling radius . for models with non - zero @xmath94 ( model c : has the same mass profiles as model a ) , the solutions are characterized by @xmath109 and the value of @xmath84 at the cooling radius , @xmath124 . since a fraction of mass drops out of the flow in this case , the inflow velocity need not rise fast and so it is possible to find completely subsonic solutions . we numerically solved the flow equations for the parameters listed in table 1 . the density , temperature and pressure profiles for three cases are presented in figures 1 , 2 & 3 . we mark the position of @xmath110 in each case , and we mark @xmath19 for the cases of transonic flows ( when @xmath125 constant ) . beyond @xmath110 we match a hydrostatic solution , as explained above , for the respective potentials . we also present , for comparison , the behaviour if the solutions outside @xmath110 are assumed to extend inwards ( that is , if no cooling flow is assumed ) . we will postpone the discussion on the effect of these profiles on the sz decrement to a later section , and only discuss the qualitative aspects of the solutions here . the solution a1 is similar to that presented by wise & sarazin ( 1993 ) ( their figure 1 ; although they chose to characterise the solutions by the temperature at @xmath110 and not @xmath109 as we have done here ) . it is also similar ( qualitatively ) to the solution for a262 presented by sulkanen ( 1989 ) . as the latter authors have noted , the effect of having a galactic potential in the center is to have a flatter temperature profile for @xmath126 , than in the case of no galactic potential . this aspect is clearly seen while comparing our solutions with and without galactic potentials in the center . our calculations for the case without the central galaxy are admittedly flawed in the very inner regions where the gas density is larger than the dark matter density , which results in an incorrect determination of the graviational potential in the inner region . however , this happens only inside a region @xmath127 kpc from the centre , and should not influence our final results by a large extent . a word of explanation for the pressure profiles is in order here . naively speaking , it would appear that the pressure profile inside the cooling radius should have lower values than the corresponding case of hydrostatic equilibrium . the fact that it is not always so has been noted in the literature ( e.g. , soker & sarazin 1988 , fig 1 of sulkanen 1989 ) . the reason for the pressure bump just outside of the sonic radius is that the flow in this inner region is not pressure driven , but rather by gravity ( see also soker & sarazin 1988 ) . this is why the bump in the profile depends on the presence and absence of the galaxy in the centre . and this profile leads to the curious result that the presence of cooling flow can lead to the overestimation of the hubble constant as discussed in the next section the model with mass deposition ( c1 ) is shown in figure 3 . the local mass flux is found to be almost proportional to the radius , consistent with various observations ( fabian , nulsen & canizares 1984 ; thomas 1987 ) , and , therefore , is probably a realistic model for cooling flow clusters . in this case the temperature drops gradually all the way through , since the velocity does not rise too fast . the deposited mass is assumed to impart negligible pressure and the pressure refers only to that of gas taking part in the flow . in this section we discuss the sz and x - ray profiles of clusters with cooling flows . we compare these with profiles from cluster having gas in hydrostatic equilibrium , and comment on the reliability of measuring hubble constant . the effect of cooling flow and the subsequent increased bremsstrahlung emmison is seen in the sudden increase in the x - ray flux in the innermost region of the cluster ( figure 5 ) . the signature of the cooling flow is seen in the form of the central spike in the x - ray profile . the x - ray profile is only affected slightly by the drop in temperature and it is the dependence on the gas density that holds . the temperature dependance becomes important only near the sonic point . outside @xmath110 , the x - ray profile is the same as that in the hydrostatic cases . the sz distortion is proportional to the line of sight integral of the pressure , and the sudden increase of the gas density inside the cooling radius is moderated by the decrease of the gas temperature . as a result there is a gradual increase in the gas pressure . near the sonic point the temperature drops drastically by orders of magnitude , and results in sudden decrease in pressure . however , since this change in pressure becomes acute only within @xmath128 of the core radius , it contributes negligibly to the line - of - sight integral of the gas pressure , and leads to an increase in the sz distortion inside the cooling radius for all models considered ( see figure 4 ) . like the x - ray profiles , the sz profiles outside @xmath110 is the same as that for the corresponding hydrosstatic cases . the sz profiles have been calculated in the rayleigh - jeans limit ( @xmath129 ) where @xmath130 of eqn . 4 goes to @xmath131 . in general , however , the profiles should be calculated using eqn 2 . our results below are independent of the observational frequency , since the profiles at different frequencies have similar shapes , with the amplitude of the sz distortion scaled either up or down . once both profiles are known , one can determine the deviation in the value of the hubble constant using eqns(14 ) & ( 16 ) . the deviation from the idealistic case can be parametrised as f_h = h_0,esth_0,true the above formula has been used to determine the deviation of the estimated value of @xmath0 from the actual value , for models listed in table-1 . the effect of cooling flow on the determination of the cosmological parameters are summarised in table-2 . to begin with , one has to get best fitted values for @xmath24 ( or @xmath132 ) and @xmath10 from different profiles . since , the estimation of the hubble constant depends on the determination of these parameters from the profiles , we look at this issue in more detail . we must keep in mind that the best fitted value of @xmath24 ( or @xmath132 ) and @xmath10 depends on whether one decides to fit the x - ray or the sz profiles , and the choice can lead to significant differences in the estimated value of @xmath0 . one of the reasons for the strong dependance on the nature of the profile can be the non - isothermality of the cluster gas . recent observations indicate that intracluster gas has a temperature structure , see markevitch et al 1998 . this is because the y - parameter depends on the integral over @xmath28 , while emissivity of thermal bremsstrahlung depends on @xmath133 . the dependance of the gaunt factor on @xmath28 isindirect and weak . yoshikawa et al 1998 have shown that gas temperature drop in the central regions ( their fig . 3 ) , should increase both @xmath24 and @xmath10 fitted to @xmath134 , and to alesser extent to @xmath135 , as compared to those compared to @xmath136 . this discrepancy increases at higher redshifts . however , in their case , there is little change in the gas density profile . clumpiness can also give rise to different fits , resulting in an overestimation of the hubble constant ( inagaki etal 1995 ) . there are two other important points that have to be kept in mind while fitting the profiles . first , we must remember that we are trying to fit a cluster having a finite profile with the formulae ( eqns 12 & 13 ) for isothermal @xmath10 profiles which is derived assuming the cluster to be of infinite extent . this can , by itself , lead to an overestimation of @xmath0 ( inagaki et al 1995 ) . thus to have a good fit one must choose a segment of the profile such that , within that segment , the profiles ( sz or x - ray ) for a finite cluster do not differ much from those of a _ hypothetical _ cluster of infinite extent . we found that sz and x - ray profiles of clusters start differing from those of infinite size at radii greater than @xmath137 times the core radius . hence , we have restricted our fitting to radii within @xmath138 . next , one must also be careful to exclude the region close to the sonic point , so that the x - ray spike is excluded from the fit . also , the central portion in the sz profile should be avoided as its inclusion can give an _ apparent _ central decrement less than its neighbouring points ( see schlickeiser 1991 ) . we have fitted the sz and x - ray profiles varying the inner cutoff radius and the results for a representative solution for each class of model are tabulated in tables 3 & 4 . thus , all fittings were done for profiles extending from @xmath139 to @xmath140 . as can be seen from table-2 , cooling flows can lead to an overestimation of the hubble constant . however , we must emphasise , that it may not be possible to _ a priori _ estimate the amount of bias introduced in the measurement of the hubble constant due to cooling flows . there is no simple correlation between the amount of cooling ( i.e @xmath84 ) and the change in the estimated @xmath0 from the actual value . the total change depends not only on @xmath84 , but also on the position of cooling radius , sonic radius , temperature at the sonic point and the isothermal temperature characterising the hydrostatic cases , with which comparisons are made . specifically , the fitted values of @xmath141 and @xmath10 for cooling flow models differ from hydrostatic models according to shape of underlying profiles , which is marked by two important features , firstly the central excess of x - ray flux ( or excess decrement of sz flux ) , and secondly the the deviation from the smooth hydrostatic profile inside @xmath110 , the amount of overestimation mainly depends on these factors . for models with a central galaxy potential , there is always an over - estimation of @xmath0 , which is greater than the models without the central galaxy . for the realistic cases of models c1 and c2 , where we have a variable @xmath84 with @xmath45 inside the cooling radius , the deviation of estimated hubble constant from its actual value is almost the same . they are also greater than that of models a and b , having similar mass flow rates . this may be due to the fact that the maximum deviation in pressure from the hydrostatic cases is more in non steady cases , than in steady flows . also non - steady cases are marked by the absence of the sonic radii and the subsequent drop in temperature . we note that although the different choice of fitting may change the absolute determination of cosmological parameters , the trend i.e deviation from the correct values remains more or less unaffected . it is interesting to note that for b - type model ( c1 ) , which include mass deposition in cooling flows , the deviations decrease as one excludes a greater part of the cooling flow region ( table 4 & 5 ) . the other models show an increase instead . here , we remind ourselves that models with mass deposition i.e c - type models are more realistic ( fabian 1994 ) . it is possible that the unsually high value of deviation ( table 4 ) and the counter - intuitive trend of increasing deviation with decreasing portion of cooling flow region used for fitting ( table 4 & 5 ) , arise because of the unrealistic modelling of cooling flows . if we take the model b1 as a realistic one , then table 4 & 5 show that to obtain a value of the hubble constant within an accuracy of @xmath2 , one should have @xmath142 . in most cases , @xmath143 ( fabian 1984 ) . however , as @xmath110 can not be determined without actually detecting a cooling flow in a cluster , we suggest that _ a significant portion of the profile within @xmath24 should be excluded as a precaution_. the sz and the x - ray profiles for the different models are shown in figures 4 & 5 . our work on the effect of temperature structure of clusters and its effect on sz decrement differs from other previous work of this nature in following way : this work takes into account the change in density profile as well as the temperature profile since both becomes important in the central region of the cluster . also , previous authors have looked at the issue of non - isothermality of a cluster at radii greater than the core radius of the cluster , whereas we look at temperature change at regions inside the core radius . for them the density profile can still be well approximated by a @xmath10 profile , whereas for cooling flow solutions , density profile is vastly different . further , they have neglected radiative cooling in their work . we for the first time look at sz effect in presence of radiative cooling , by first solving the cooling flow equations for reasonable and physical solutions . in summary , we find that the presence of a cooling flow in a cluster can lead to an overestimation of the hubble constant determined from the sunyaev - zeldovich decrement . we have used different models of cooling flows , with and without mass deposition , and found the deviation in the estimated value of the hubble constant in the case of a cooling flow from that of hydrostatic equilibrium . we have used the usual procedure of fitting the sz and x - ray profile with a @xmath10 profile to get an estimated value of @xmath24 , and then compared with that for the case of gas in hydrostatic equilibrium in order to estimate the deviation in the hubble constant . for the more realistic models with mass deposition ( varying @xmath1 with radius ) , we found that the deviation decreases with the exclusion of greater portions of the cooling flow region . quantitatively , we found that for the deviation to be less than @xmath2 , one should exclude a portion of the profile upto @xmath144 . since @xmath110 is difficult to estimate without actually detecting a cooling flow , we have suggested that a significant portion of the profile inside @xmath24 should be excluded , to be safe . there can be another important implication of the effect of cooling flows . with the upcoming satellite missions ( map & planck ) , we have come to the point where there are efforts to constrain @xmath145 with surveys of blank sz fields ( bartlett et al 1998 ; bartlett 2000 ) , ultimately giving rise to sz - selected catalog of clusters ( aghanim et al 1997 ) . this method relies on estimating the number of sz sources brighter than a given threshold flux ( barbosa et al 1996 ) . the point to be noted is that since these surveys are essentially flux limited in nature , the validity of the analysis to determine @xmath145 depends crucially on the _ one - to - one _ association of flux - limits to corresponding mass limits of clusters . from our analysis above , it seems that it may not be possible to associate a unique cluster mass to a given sz distortion , given the uncertainty due to the presence of cooling flows . this might lead to contamination in sz cluster catalogs and the inference of @xmath145 . recently , attempts have been made to constrain @xmath9 from variance measurement of brightness temperature in blank fields ( subrahmanyan et al 1998 ) and comparing them to simulated fields ( majumdar & subrahmanyan , 2000 ) , of cumulative sz distortions from a cosmolgical distribution of clusters . these results may also be systematically affected due to the presence of clusters having cooling flows . the estimations made in this paper strictly applies to cases where the image of the sz effect is directly obtained by single dish observations . for interferometric observations , the interferometer samples the fourier transform of the sky brightness rather than the direct image of the sky . the fourier conjugate variables to the right ascension and declination form the @xmath146 plane in the fourier domain . due to spatial filtering by an interferometer , it is necessary that models be fitted directly to the data in the @xmath146 plane , rather that to the image after deconvolution . we do not forsee drastic change from our inferences in such cases since the result mainly depends on the deviation of the sz and x - ray profile in case of a cooling flow from those in hydrostatic equilibrium . this , however , should be looked in greater detail in future . we also note that with the growing number of high quality images of the sz effect with interferometers , which have greater resolution than single dish antennas , the shape parameters of the clusters can be directly determined from the sz dataset rather than from an x - ray image ( grego et al . 2000 ) . finally , we would like to add , though the calculations presented in this paper were done using the dark matter profile ( eqn . 21 ) , which is `` commonly used '' for calculating cooling flow solutions , it is inconsistent with the dark matter profile ( navarro et al 1997 ) found in numerical simulations . ( for a comparison of mass and gas distribution in clusters having cooling flows with different dark matter profiles , see waxman & miralda - escude , 1995 ) . moreover , we have neglected the self gravity of the gas . suto et al ( 1998 ) have calculated the effect of including the self gravity of the gas in determining the gas density profile . to make strong conclusions about the effect of cooling flows in the determination of the hubble constant , one should take both the points mentioned above into account . we are grateful to joseph silk for his comments on the manuscript . we also thank the referee , dr asantha cooray , for numerous suggestions which have helped us in improving the paper . one of the authors ( sm ) would like to thank the raman research institute , bangalore , for hospitality and for providing computational facilities . sm acknowledges help recieved from r.sridharan on using idl . finally , he would also like to express his gratitude to pijushpani bhattacharjee for his constant encouragement . aghanim , n. , de luca , a. , bouchet , f. r. , gispert , r. & puget , j. l. , 1997 , a & a , 325 , 9 allen , s. w. , fabian , a. c. , johnstone , r. m. , arnaud , k. a. , & nulsen , p. e. j. , 1999 : astro - ph/9910188 allen , s. w. , 2000 : astro - ph/0002506 barbosa , d. , bartlett , j. g. , blanchard , a. & oukbir , j. , 1996 , astro - ph/9607036 barbosa , d. , bartlett , j. g. , blanchard , a. & oukbir , j. , 1996 , a & a , 314 , 13 bartlett , j. g. & silk , j. , 1994 , apj , 423 , 12 bartlett , j. g. , blanchard , a. & barbosa , d. , astro - ph/9808308 bartlett , j. g , . 2000 : astro - ph/0001267 birkinshaw , m. & hughes , j. p. , 1994 , apj , 420 , 33 birkinshaw , m. 1999 , phys rep . , 310 , 97 blanchard , a & bartlett , j. g. , 1998 , a & a , 332 , 49l cavaliere , a. & fusco - femiano , r. , 1978 , a & a , 70 , 677 evrard , a. e. , 1990 , apj , 363 , 349 fabian , a. c. , nulsen , p. e. j. & canizares , c. r. 1984 , nature , 310 , 733 fabian , a. c. 1994 , araa , 32 , 277 grego , l. , carlstrom , j. e. , joy , m. k. , reese , e. d. , holder , j. p. , patel , s. , cooray , a. , & holzapfel , w. l. , 2000 , : astro - ph/0003085 herbig , t. , lawrence , c. r. , redhead , a. c. s. , & gulkis , s. , 1995 , apj , 449 , l5 holzapfel , w. l. , et al . 1997 , apj , 480 , 449 reese , e. d. , 1999 , preprint ( astro - ph/9912010 ) hughes , j. p. , & birkinshaw , m. , 1998 , apj , 501 , 1 inagaki , y. , suginohara , t. & suto , y. , 1995 , pasj , 47 , 411 jaffe , w. 1980 , apj , 241 , 925 jones , m. et al . 1993 , nature , 365 , 320 kellog , e. , baldwin , j. r. & koch , d. , 1975 , apj , 199 , 299 komatsu , e. , kitayama , t. , suto , y. , hattori , m. , kawabe , r. , matsuo , h. , schindler , s. , & yoshikawa , k. , 1999 , apj , 516 , l1 majumdar , s. & subrahmanyan , r. , 2000 , mnras , 312 , 724 makino , n. , sasaki , s. & suto , y. , 1998 , apj , 497 , 555 markevitch , m. , forman , w. r. , sarazin , c. l. & vikhlinin , a. , 1998 , apj , 504 , 27 mathews , w. g. , & bregman , j. n. , 1978 , apj , 224 , 308 navarro , j. f. , frenk , c. s. , & white , s. d. m. , 1997 , apj , 40 , 493 oukbir , j. & blanchard , a. , 1992 , a & a , 262 , 21l oukbir , j. & blanchard , a. , 1997 , a & a , 317 , 1 padmanabhan , t. , 1993 , structure formation in the universe , cambridge university press , cambridge peres , c. b. , fabian , a. c. , edge , a. c. , allen , a. w. , johnstone , r. m. , & white , d. a. , 1998 , mnras , 298 , 416 pointecouteau , e. , giard , m. , benoit , a. , dsert , f. x. , aghanim , n. , coron , n. , lamarre . , j. a. , & delabrouille , j. , 1999 , apj , 519 , l115 press , w. h. & schecter , p. , 1974 , apj , 187 , 225 reese , e. d. , et al . 1999 : astro - ph/9912071 rizza , e. , loken , c. , bliton , m. , roettiger , k. , & burns , j. , 2000 , apj , 119 , 21 roettiger , k. , stone , j. m. & mushotzky , r. f. , 1997 , apj , 482 , 588 sarazin , c. l. 1986 , rev . phys . , 58 , 1 sarazin , c. l. & white , r. e. 1987 , apj , 320 , 32 sarazin , c. l. & graney , c. m. 1991 , apj , 375 , 532 saunders , r. et al . 1999 : astro - ph/9904168 schlickeiser , r. 1991 , a & a , 128 , 23l schmutzler , t. & tscharnuter , w. m. 1993 , a & a , 273 , 318 silverberg , r. f. et al , 1997 , apj , 485 , 22 soker , n. & sarazin , c. l. 1988 , apj , 327 , 66 subrahmanyan , r. , kesteven , m. j. , ekers , r. d. , sinclair , m. & silk , j. 1998 , mnras , 298 , 1189 sulkanen , m. e. , burns , j. o. & norman , m. l. 1989 , apj , 344 , 604 sunyaev , r. a. & zeldovich , ya . b. 1972 , comm . space phys.,4 , 173 suto , y. , sasaki , s. & makino , y. , 1998 , apj , 509 , 544 thomas , p. , fabian , a. c. & nulsen , p.e.j . 1987 , mnras , 228 , 973 . waxman , e. & miralda - escude , j. , 1995 , apj , 451 , 451 white , r. e. , & sarazin , c. l. , 1987 , apj , 318 , 629 wise , m. w. & sarazin c. l. , 1993 , apj , 415 , 58

building scalable multi - qubit systems is presently the main challenge towards the implementation of a solid - state quantum information processor @xcite . high - fidelity single qubit gates based both on semi- @xcite and super - conducting technologies are nowadays available . in particular , for the three basic types of superconducting qubits ( charge , flux and phase ) single - qubit operations with high quality factor have been demonstrated in different laboratories @xcite . further improvement has been recently achieved with circuit - qed architectures @xcite . multi - qubit systems instead have been proved harder to operate , the main limitation arising from the broadband solid state noise . the requirements for building an elementary quantum processor are in fact quite demanding on the efficiency of the protocols . this includes both severe constraints on readout and a sufficient isolation from fluctuations to reduce decoherence effects . based on experiments on the different josephson junction ( jj ) setups , there is presently a general consensus on the most common noise spectral forms and on the main consequences on systems evolutions . typically noise is broadband and structured , i. e. the noise spectrum extends to several decades , it is non - monotonic , sometimes a few resonances are present . noise with @xmath0 spectrum is common to virtually all nanodevices . its physical origin varies from device to device and depends on the specific material . different implementations are in fact more sensitive either to charge , flux or to critical current fluctuations with spectral density scaling as the inverse of the frequency . the presence of slow components in the environment makes the decay of the coherent signal strongly dependent on the experimental protocol being used @xcite . measurements protocols requiring numerous repetitions are particularly sensitive to the unstable device calibration due to low - frequency fluctuations . the leading effect is defocusing of the measured signal , analogous to inhomogeneous broadening in nmr @xcite . the intrinsic high - frequency cut - off of @xmath0 noise is hardly detectable , measurements typically extending to @xmath1 hz ( recently charge noise up to @xmath2 mhz has been detected in a set @xcite ) . incoherent energy exchanges between system and environment , leading to relaxation and decoherence , occur at typical operating frequencies ( about @xmath2 ghz ) . indirect measurements of noise spectrum in this frequency range often suggest a white or ohmic behavior @xcite . in addition , narrow resonances at selected frequencies ( sometimes resonant with the nanodevice relevant energy scales ) have being observed @xcite . in certain devices they originate from the circuitry @xcite and may eventually be reduced by circuit design . more often , resonances are signatures of the presence of spurious fluctuators which also show up in the time resolved evolution , unambiguously proving the discrete nature of the noise sources @xcite . such fluctuators may severely limit the reliability of nanodevices @xcite . explanation of this rich physics is beyond phenomenological theories describing the environment as a set of harmonic oscillators . on the other side , an accurate characterization of the noise sources might be a priory inefficient , since a huge number of parameters would be required for a microscopic description . therefore one may follow a different route , consisting in classifying noise sources on the basis of their effects instead of on their nature and to understand , case by case , which is the efficient description of the environment . the required information may in fact depend on the specific protocol , especially when the environment is long - time correlated . this program is meaningful for quantum information where relevant time scales are much smaller than the decoherence time . this means that in the generic favorable situation coupling with the environment has simple effects on the system dynamics . as a difference with condensed matter physics where long time behavior is emphasized and interesting problems involve entanglement of a many - body system . here we illustrate a road - map to treat broadband noise which allows to obtain reasonable approximations by systematically including only the relevant information on the environment , out of the huge parametrization needed to specify it . the multi - stage approach for the different classes of broadband noise has been originally introduced for single qubit gates @xcite . the obtained predictions for the decay of the coherent signal are in agreement with observations in various jj implementations and in different protocols @xcite . we mention the observed decay of ramsey fringes in charge - phase qubits @xcite and recent results on flux / phase qubits @xcite . here we extend the procedure to complex solid - state architectures . as an illustrative case , we perform a systematic analysis of the effects and interplay of low and high frequency noise components in a two - qubit gate in a fixed coupling scheme . such a systematic analysis points out that efficient operations in the solid state require an accurate preliminary characterization of the noise spectral characteristics and tuning appropriately the device working point . the paper is organized as follows : in section 2 we introduce the general framework of solid - state multi qubit gates and illustrate the characteristics of the most common spectral forms . in section 3 we propose a classification of the noise sources and in section 4 we present a multi - stage approach to deal with broadband noise . section 5 is dedicated to adiabatic noise . in section 6 the multistage approach is applied to a universal two - qubit gate . section 7 summarizes our main findings . a multi - qubit gate can be modeled by the following hamiltonian ( @xmath3 ) @xmath4 where for each qubit , labeled by @xmath5 , @xmath6 includes set of parameters intrinsic to the device and time dependent classical control fields . manipulation of tunable fields allows control of the dynamics and design of arbitrary single qubit operations via unitary transformations . interactions among qubits and the needed additional parametrization to describe the multi - qubit gate are included in @xmath7 . depending on the design , the control hamiltonian may span a reduced subspace of the qubit liouville space . in other words , there is a limited number of ports available for control , for tuning , for state preparation and measurement . noise is also coupled via these ports . it is rather usual that only one of the control fields allows the required fast addressing of each qubit . therefore we focus on a model where both the control fields , @xmath8 , and the environment are coupled to a single qubit operator , say @xmath9 , @xmath10 where the polar angles @xmath11 define qubit-@xmath12 working point . the environment hamiltonian is @xmath13 and @xmath14 is a collective environment variable acting on qubit @xmath12 . model ( [ h ] ) implies a projection of the device hamiltonian onto the subspace spanned by the two lowest energy eigenstates for each qubit . this description is valid provided that manipulations with the control fields do not induce leakage to higher energy states of the device @xcite . we leave unspecified the nature of the noise sources described by @xmath15 , relevant cases being either discrete or gaussian fluctuations . in addition , the environment hamiltonian may include correlations among noise sources affecting each qubit . in the spirit of the present analysis , we assume that the only information at our disposal is the power spectrum of @xmath16 fluctuations @xmath17 where @xmath18 denotes the equilibrium average with respect to @xmath13 . we remark that the effect of the environment on the system dynamics is completely characterized by @xmath19 only if the environment is composed of harmonic oscillators or if it is weakly coupled to the system and short time correlated . of course this is not the case if low - energy excitations determine memory effects . this is the typical situation in the solid state where in general additional statistical information on the environment is required . we assume that each noise component @xmath16 has broadband spectrum @xmath20 , @xmath21 followed by a white or ohmic flank at frequencies @xmath22 . low- and high - frequency cutoffs depend on the specific setup . impurities of various origin responsible for random telegraph fluctuations contribute lorentzian peaks to the power spectrum @xmath23 . such a spectrum originates from classical fluctuations @xmath24 , where @xmath25 randomly switches between two values @xmath26 with rate @xmath27 . an ensemble of @xmath28 bistable impurities with a distribution of switching rates @xmath29 , @xmath30 $ ] and average coupling strength @xmath31 , gives rise to @xmath0-spectrum , @xmath32\,\,\omega^{-1}$ ] @xcite . selected lorentzian peaks may be visible in the spectrum if individual impurities are strongly coupled , i. e. when @xmath33 . instead , damped coherent fluctuators at frequencies @xmath34 , in the simplest cases contribute to the total spectrum with additional peaks , @xmath35 . spectroscopic evidence of coherent impurities of frequency @xmath36 close to the qubit larmor frequency has been reported in @xcite . remarkably , the possibility to exploit spurious quantum two - level systems as qubits @xcite or for quantum memory operations has been recently demonstrated @xcite . in these cases a quantum description of the impurity is required @xcite . the above description illuminates that in the solid state we have to deal with broadband and structured noise . in other words , the noise spectrum extends to several decades , it is non - monotonic , sometimes a few resonances are present . the various noise sources responsible for the above phenomenology have a qualitative different influence on the system evolution . this naturally leads to a classification of the noise sources according to the effects produced rather than to their specific nature . the effects of high - amplitude noise at low frequencies , like @xmath0 noise , vary from protocol to protocol . this feature is typical of non - markovian baths . quantum operations necessarily require repetitions of single detections , each leading to boolean answer . therefore , even `` single shot '' measurements result from numerous repetitions of single runs in an overall process which may last minutes . in the presence of low frequency fluctuations this leads to unstable device calibration and random clock frequencies in the various repetitions . as a result , a de - focused signal is observed , a phenomenon analogous to inhomogeneous broadening in nmr . re - focusing protocols , like echo or some dynamical decoupling schemes , allow partial recovery of the signal @xcite . since the environment is long - time correlated , statistical information beyond the power spectrum may be required to describe its effects . this is the case for instance in echo protocols @xcite . environments with long - time memory belong to the class of adiabatic noise , for which the born - oppenheimer approximation is applicable . we classify this part of the noise spectrum as `` adiabatic noise '' . noise at higher frequency , around the system typical scales ( e.g. single qubit larmor frequencies ) , results in incoherent energy exchanges between the quantum device and the environment . in particular , high - frequency noise is responsible for spontaneous decay . for systems relevant for quantum information , the system - bath coupling is rather small , in addition high - frequency noise is short - time correlated . therefore , in simplest cases , effects can be described by a born - markov master equation @xcite . for single qubit gates it leads to the relaxation and secular dephasing times , @xmath37 and @xmath38 @xcite . we classify this part of the noise spectrum as `` quantum noise '' . finally , resonances in the spectrum unveil the presence of discrete noise sources which severely effect the system performances , in particular reliability of devices . this is the case when classical impurities are slow enough to induce a visible bistable instability in the system intrinsic frequency . for instance , single qubit gates in the presence of random telegraph noise ( switching rate @xmath27 ) may display two effective frequencies , @xmath39 and @xmath40 , depending on the impurity state . their visibility is measured by the ratio @xmath41 . beatings can be observed in the `` strong coupling '' regime @xmath42 @xcite . quantum impurities may also entangle with the device . this additionally leads to a variety of features , like peculiar temperature dependencies of decay rates @xcite . under these conditions , knowledge of the power spectrum is absolutely insufficient and in order to describe these effects the relevant system hilbert space has to be enlarged to include the responsible environmental degrees of freedom . effects in general depend on the specific protocol and require a microscopic model of the fluctuators . we classify this part of the noise spectrum as `` strongly - coupled noise '' . each noise class requires a specific approximation scheme , which is not appropriate for the other classes . the overall effect results from the interplay of the three classes of noise . in the following section we will illustrate a multi - scale theory to deal with solid state broadband noise . we are interested to a reduced description of the @xmath43-qubit system , expressed by the reduced density matrix @xmath44 . it is formally obtained by tracing out environmental degrees of freedom from the total density matrix @xmath45 , which depends on quantum ( q ) , adiabatic ( a ) and strongly coupled ( sc ) variables . the elimination procedure can be conveniently performed by separating in the interaction hamiltonian , @xmath46 , various noise classes , e.g. by formally writing @xmath47 adiabatic noise is typically correlated on a time scale much longer than the inverse of the natural frequencies @xmath48 , then application of the born - oppenheimer approximation is equivalent to replace @xmath49 with a classical stochastic field @xmath50 . this approach is valid when the contribution of adiabatic noise to spontaneous decay is negligible , a necessary condition being @xmath51 . this condition is usually satisfied at short enough times , since @xmath52 is substantially different from zero only at frequencies @xmath53 . this fact already suggests a route to trace - out different noise classes in the appropriate order . the total density matrix parametrically depends on the specific realization of the slow random drives @xmath54 and may be written as @xmath55 . the first step is to trace out quantum noise . in the simplest cases this requires solving a master equation . in a second stage , the average over all the realizations of the stochastic processes , @xmath56 , is performed . this leads to a reduced density matrix for the @xmath43-qubit system plus the strongly coupled degrees of freedom . these have to be traced out in a final stage by solving the heisenberg equations of motion , or by approaches suitable to the specific microscopic hamiltonian or interaction . for instance , the dynamics may be solved exactly for some special quantum impurity models at pure dephasing , @xmath57 , when impurities are longitudinally coupled to each qubit @xcite . the ordered multi - stage elimination procedure can be formally written as @xmath58 \ , p[\vec x(t ) ] \ ; tr_{q } \big [ \ , w^{q , sc}\big(t | \vec x(t)\big ) \ , \big]\right\ } \ , .\ ] ] in the following section we concentrate on the elimination of adiabatic noise . the ordered procedure will be illustrated in section 6 for a two - qubit gate . in general , the adiabatic approximation holds true for times short enough to fulfill the necessary condition , @xmath59 . in the peculiar , pure dephasing regime , @xmath57 , relaxation processes are forbidden and the adiabatic approximation is exact for any @xmath60 . in addition , the adiabatic scheme can be applied also in the presence of correlations between processes @xmath61 and @xmath62 @xcite . for the sake of clarity here we consider adiabatic noise affecting independently each qubit . moreover , we exclude time - dependent drives in @xmath8 and in @xmath63 . the procedure can be straightforwardly extended for instance to rabi oscillations and other ac - drives @xcite . suppose we are able to diagonalize , for each realization @xmath56 of the stochastic processes , the system hamiltonian , @xmath64 and denote @xmath65 an instantaneous eigenstate of ( [ eq : hamiltonian - stochastic ] ) with eigenvalue @xmath66 . if the system is prepared in a pure state , @xmath67 , in the adiabatic approximation , each component @xmath68 evolves in time according to @xmath69 where @xmath70 is the dynamic phase . the stray fields @xmath56 affect the time evolution of the gate fidelity via modifications of the phases @xmath71 and of the eigenstates @xmath72 . in addition , @xmath73 reflects imperfect preparation due to the random field . this occurs , for instance , when the system is prepared by a reset to a pure state vector @xmath67 followed by an initialization pulse whose effect depends also on the stray field at @xmath74 . the evolution of the system conditional density matrix can be presented in a compact form @xmath75 \ ; { \mathrm e}^{- i \int_0^t d s \ , \omega_{m p}(s ) } \ , , \end{aligned}\ ] ] where the instantaneous splittings appearing in the phase factor are @xmath76 and we have introduced the operator @xmath77 = \ , { \,| \ , { m(\vec x_t ) } \,\rangle \,}{\ , \langle\,{m(\vec x_0 ) } \,|\ , } \rho(0 ) { \,| \ , { p(\vec x_0 ) } \,\rangle \ , } { \ , \langle\,{p(\vec x_t ) } \,|\,}\ ] ] which contains information about the preparation and the eigenstates errors . finally , the average of @xmath78 over the stochastic process yields the system density matrix , which can be presented as a path integral @xmath79 \ , p[\vec x(s ) ] \ , \rho^n\big(t|\vec x(s)\big ) \ , .\end{aligned}\ ] ] here @xmath80 $ ] contains information both on the stochastic processes and on details of the specific protocol . it is convenient to split it as follows @xmath81 \,=\ , f[\vec x(s ) ] \ ; p[\vec x(s ) ] \ , , \ ] ] where @xmath82 $ ] is the probability of the realization @xmath83 . the filter function @xmath84 $ ] describes the specific operation . for most of present day experiments on solid - state qubits @xmath84=1 $ ] . for an open - loop feedback protocol , which allows initial control of some collective variable of the environment , say @xmath85 , we should put @xmath84 \propto \ , \delta(\vec x_0 ) = \ , \pi_i \delta(x_{i0})$ ] . a critical issue is the identification of @xmath82 $ ] for the specific noise sources , as those displaying @xmath0 power spectrum . if we sample the stochastic process at times @xmath86 , with @xmath87 and @xmath88 , we can identify @xmath89 = \lim_{m \to \infty } p_{m+1}(\vec x_m , t;\dots ; \vec x_1 , t_1 ; \vec x_0 , 0 ) \ , , \label{sampling}\ ] ] where @xmath90 is a @xmath91 joint probability and we have used the shorthand notation @xmath92 . in the following we will propose a systematic method to select only the relevant statistical information on the stochastic process out of the full characterization included in @xmath82 $ ] . we would like to remark , at this point , that in the adiabatic treatment the word decoherence is perhaps abused . decoherence is ultimately due to entanglement of the qubit to a quantum environment @xcite , whereas classical adiabatic noise produces only de - focusing . however , unless the signal can be _ totally _ re - focused , and this is impossible in practice , the behavior of @xmath93 is the same as for true decoherence @xcite . in other words , although proper and improper mixed states at the fundamental level are well distinct concepts @xcite , they are not distinct in the density matrix description . we also observe that present day `` applied '' research on solid - state coherent nanodevices for quantum information focuses on the `` short - time '' dynamics , since a signal which is almost decayed is useless . in this context , methods as the adiabatic approximation are valuable even if they are not accurate at long time scales and are not valid down to zero temperature . the longitudinal approximation consists in neglecting modifications of the eigenstates @xmath72 and preparation effects . without loss of generality , we may assume vanishing average of the stochastic processes after the preparation pulse , @xmath94 . the longitudinal approximation amounts to put @xmath95 where @xmath96 is an eigenstate of the system hamiltonian @xmath97 . then @xmath98 \approx { \,| \ , { m } \,\rangle \ , } \rho^n_{mp}(0 ) { \ , \langle\,{p } \,|\,}$ ] is a projected element of the initial density matrix and eq.([eq : pathint - adiabatic ] ) simplifies to @xmath99 \ , p[\vec x(s ) ] \ ; { \mathrm e}^{- i \int_0^t \!\!d s \ , \omega_{m p}(s ) } \ , .\end{aligned}\ ] ] the significance of the longitudinal approximation is easily illustrated for a single qubit gate , when the system hamiltonian reduces to @xmath100 in this case , retaining variations of the splitting @xmath39 amounts to consider only fluctuations of the length of the vector @xmath101 . this would be the only effect if the noise would act longitudinally , @xmath102 . variations of the eigenstates @xmath103 originate instead from `` transverse '' variations of @xmath101 . the longitudinal assumption has been performed in ref.@xcite to discuss the effect of gaussian adiabatic environments . the present approach automatically provides constraints on its validity and shows that whereas errors due to transverse fluctuations are weakly dependent on time , phase errors accumulate . therefore , transverse noise in the adiabatic approximation has possibly some effect only at very short times , but the phase damping channel eventually prevails . these considerations strongly depend on the amplitude of the noise . we checked analytically and with simulations that they hold true for realistic figures of noise as those measured in experiments by zorin et al . @xcite . in the longitudinal approximation , diagonal elements of the reduced density matrix in the eigenstate basis , @xmath104 , do not decay . instead , the decay of the off - diagonal elements results from @xmath105 $ ] , where the complex average phase is given by @xmath106 \ , p[x(s ) ] \ , { \mathrm e}^{i \int_0^t \!\!ds \ , \omega[x(s ) ] } \ , .\ ] ] we notice that the longitudinal approximation may be exact for certain protocols , which therefore are not affected by adiabatic transverse noise . for instance , this is the case if the system is described by ( [ eq : hamiltonian - stochastic2 ] ) and it is prepared in a state @xmath107 . if we measure @xmath108 , since @xmath109 do not depend on @xmath110 , the result is not affected by transverse fluctuations and @xmath111 \ , \exp\{{{\cal i } m \,}\ ! \phi(t)\}$ ] . this the case of the decaying oscillations pattern measured with ramsey interference @xcite , if the effect of imperfect @xmath112 pulses is negligible . a standard approximation of the path integrals ( [ eq : pathint - adiabatic],[eq : pathint - adiabatic - longitudinal ] ) consists in neglecting the time dependence in the path , @xmath113 and taking @xmath114=1 $ ] . in this static path approximation ( spa ) the problem reduces to ordinary integrations with @xmath115 . in the single qubit case , for instance , eq.([eq : blur ] ) gives average phase shift @xmath116 where @xmath117 ^ 2 + ( \omega \cos \theta)^2}$ ] . eq.([eq : blur - static ] ) describes the effect of a distribution of stray energy shifts @xmath118 and corresponds to the rigid lattice breadth contribution to inhomogeneous broadening @xcite . in experiments with solid state devices this approximation describes the measurement procedure consisting in signal acquisition and averaging over a large number @xmath119 of repetitions of the protocol , for an overall time @xmath120 ( which may also be minutes in actual experiments ) . due to slow fluctuations of the solid state environment calibration , the initial value , @xmath121 , fluctuates during the repetitions blurring the average signal , independently on the measurement being single - shot or not . the probability @xmath122 describes the distribution of the random variable obtained by sampling the stochastic process @xmath25 at the initial time of each repetition , i. e. at times @xmath123 , @xmath124 . if @xmath125 results from many independent random variables of a multimode environment , the central limit theorem applies and @xmath122 is a gaussian distribution with standard deviation @xmath126 @xmath127 with integration limits @xmath128 , and the high - frequency cut - off of the @xmath0 spectrum , @xmath129 . in the spa the distribution standard deviation , @xmath126 , is the only adiabatic noise characteristic parameter . if the equilibrium average of the stochastic process vanishes , eq.([eq : blur - static ] ) reduces to @xmath130 ^ 2 + ( \omega \cos \theta)^2 } } \ , . \label{spa - oneqb}\ ] ] a convenient approximation is obtained by expanding @xmath131 to second order in @xmath125 , which leads to @xcite @xmath132 the short - times decay of coherent oscillations qualitatively depends on the working point . in fact , the suppression of the signal , @xmath133 $ ] , turns from a @xmath134 behavior at @xmath135 to a power law , @xmath136^{-1/4}$ ] , at @xmath137 . in these limits eq.([ref : quadratic ] ) reproduces known results for gaussian @xmath0 environments . in particular , at @xmath138 we obtain the short - times limit , @xmath139 , of the exact result of ref.@xcite . at @xmath140 the short and intermediate times result of ref . @xcite is reproduced . the fact that results of a diagrammatic approach with a quantum environment , as those of ref . @xcite , can be reproduced and generalized already at the simple spa level makes the semi - classical approach quite promising . it shows that , at least for not too long times ( but surely longer than times of interest for quantum state processing ) , the quantum nature of the environment may not be relevant for the class of problems which can be treated in the born - oppenheimer approximation . notice also that the spa itself has surely a wide validity since it does not require information about the _ dynamics _ of the noise sources , provided they are slow @xcite . going beyond the spa amounts to sample more accurately the stochastic process @xmath83 in ( [ sampling ] ) . the first correction to the spa is obtained by parametrizing the random process as follows @xmath141 . inserting this expression in ( [ eq : pathint - adiabatic ] ) and approximating @xmath82 $ ] in ( [ sampling ] ) with the conditional probability @xmath142 $ ] , we obtain @xmath143 the joint probability depends on the statistics of the noise sources . therefore the first correction to the spa distinguishes discrete and gaussian processes . considering again a single qubit , the average phase ( [ eq : blur ] ) at the working point @xmath144 for generic gaussian noise becomes @xmath145 t \over \omega } \big ] + { 1 \over 2 } \ln \big[1 + i { \sigma^2 \pi(t)\,t \over 3 \ , \omega } \big ] \ , , \ ] ] where @xmath146 is a transition probability , depending on the stochastic process . for ornstein - uhlenbeck processes it reduces to the result of ref . the first correction suggests that the spa , in principle valid for @xmath147 , may have a broader validity . this is illustrated in fig.([figure1 ] ) where the adiabatic approximation ( numerical evaluation of eq.([eq : blur ] ) ) is compared with the exact ( numerical ) dynamics in the presence of @xmath0 noise and with the analytic forms resulting from the spa and its first correction . for @xmath0 noise due to a set of bistable impurities the spa valid also for @xmath148 , if @xmath149 . of course the adiabatic approximation is tenable if @xmath150 . by sampling more accurately the adiabatic process @xmath56 it is possible to selectively include the statistical information needed for the specific measurement process . for instance , echo protocols are able to partly re - focus the signal , in other words de - focusing described by the spa is almost canceled by the echo pulse sequence . the decay of the echo signal is due to the un - canceled dynamics of the low - frequency fluctuations and to quantum noise . the leading effect of adiabatic noise can be estimated by a proper parametrization of @xmath56 , similar to the one considered in the present paragraph @xcite . in the present section we apply the multi - stage approach to a universal two - qubit gate based on a fixed coupling scheme . capacitive and inductive fixed couplings @xcite have been used to demonstrate two - qubit logic gates in different jj implementations @xcite . entanglement is generated by tuning single - qubit energy spacing to achieve mutual resonance . to this end , during the gate operation at least one qubit has to be moved away from the working point of minimal sensitivity to parameters variations , the `` optimal point '' . this has so far represented the main drawback of the fixed - coupling scheme for josephson implementations , with the exception of phase qubits @xcite . more recent proposals have attempted to solve this problem by introducing tunable coupling schemes @xcite . most of them rely on additional circuit elements and gain their tunability from ac - driving @xcite or from `` adiabatic '' couplers @xcite . some of these coupling schemes have been tested in experiments and are potentially scalable @xcite . none of these implementations is however totally immune from imperfections . in general , introducing additional on - chip circuits elements opens new ports to noise . the possibility to employ `` minimal '' fixed coupling schemes has been recently reconsidered in ref . @xcite , pointing out the possibility to single out `` optimal coupling '' conditions which ensure reasonable protection from @xmath0 fluctuations . in order to implement a @xmath151 gate in a fixed coupling scheme we need two resonant qubits with a transverse coupling , as described by the hamiltonian @xmath152 where @xmath153 is the coupling strength , and @xmath154 the pseudo - spin operators whose eigenstates @xmath155 ( eigenvalues @xmath156 ) are the computational states of qubit @xmath12 . eigenvalues and eigenvectors of @xmath157 are given by -6pt @xmath158 where @xmath159 , @xmath160 with @xmath161 and we have used the shorthand notation @xmath162 . the level structure of the coupled system is schematically illustrated in fig.([splittings ] ) . as a result of the diagonal block structure of the hamiltonian ( [ eq : hamiltonian ] ) in the computational space , the two - qubit hilbert space is factorized in two subspaces spanned by pairs of eigenvectors . a system prepared in @xmath163 , freely evolving for a time @xmath164 , yields the entangled state @xmath165 , corresponding to a @xmath151 operation . the pair of states @xmath166 and @xmath167 span the subspace where the @xmath151 gate is realized , which we name swap - subspace . the subspace spanned by the pair of states @xmath168 and @xmath169 is termed z - subspace . eq.([eq : hamiltonian ] ) models , for instance , capacitive coupled charge qubits or inductively coupled flux qubits @xcite . typically , these system can be driven by pulses acting along the transverse direction , @xmath9 . the same port introduces noise into the system @xcite . this is the case of the charge - phase two - port design of ref . @xcite . ] . here we consider noise sources acting transversely with respect to each qubit , i. e. the interaction hamiltonian takes the form @xmath170 each noise component @xmath16 has broadband spectrum @xmath20 , @xmath21 followed by a white flank at frequencies @xmath171 . correlated noise sources acting on both qubits have been addressed in @xcite . if the system is initialized in the swap subspace , for instance in the state @xmath172 , bitwise readout gives the qubit 1 switching probability @xmath173 , i.e. the probability that it will pass to the state @xmath174 , and the probability @xmath175 of finding the qubit 2 in the initial state @xmath174 . cyclic anti - correlation of the probabilities signals the formation of the entangled state during the @xmath151 operation . in terms of the two qubit reduced density matrix in the eigenstate basis the switching probabilities read @xmath176 + \rho_{00}(t ) + \left [ \ , \rho_{33}(t ) - \rho_{00}(t ) \right ] \sin^2 \frac{\varphi}{2 } \nonumber \\ & & + { \rm re}[\rho_{12}(t ) ] + { \rm re } [ \rho_{03}(t ) ] \sin \varphi \label{eq : psw1 } \\ \!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!\!p^{(2)}(t ) = \langle - | { \rm tr}_1 \rho(t ) | - \rangle & = & \frac{1}{2 } \left [ \ , \rho_{11}(t ) + \rho_{22}(t ) \right ] + \rho_{00}(t ) + \left [ \ , \rho_{33}(t ) - \rho_{00}(t ) \right ] \sin^2 \frac{\varphi}{2 } \nonumber \\ & & - { \rm re}[\rho_{12}(t ) ] + { \rm re}[\rho_{03}(t ) ] \sin \varphi . \label{eq : psw2}\end{aligned}\ ] ] the matrix elements entering the above probabilities can be evaluated in the multi - stage approach . we split the interaction as in eq.([int - plit ] ) , @xmath177 . low - frequency noise is treated in the adiabatic and longitudinal approximation . in addition we limit the analysis to the spa . we denote with @xmath178 the eigenvalues of @xmath179 with @xmath180 . _ first stage : elimination of quantum noise _ quantum noise is traced out by solving the born - markov master equation for the reduced density matrix @xcite with eigenvalues parametrically dependent on the random fields @xmath181 . in the system eigenstate basis with @xmath180 ] and performing the secular approximation ( to be self - consistently checked ) it takes the standard form @xcite : @xmath183 where @xmath184 . the rates @xmath185 , @xmath186 depend on the real parts of the lesser and greater green s functions , describing emission ( absorption ) rates to ( from ) the quantum reservoirs @xcite . in addition , for white quantum noise energy shifts are vanishing and do not appear in eq . ( [ eq : coherences ] ) . because of the symmetry of ( [ eq : hamiltonian ] ) , dissipative transitions inside each - swap or z - subspace are forbidden . the allowed inelastic energy exchange processes are evidenced in fig.([splittings ] ) . the only independent emission rates are @xmath187 , @xmath188 . they read @xmath189 \\ \gamma_{20 } = \frac{1}{8 } \ , ( 1-\sin \varphi ) \ , [ c_{x_1}(\omega_{20 } ) + c_{x_2}(\omega_{20 } ) ] \ , , \end{array } \label{eq : ratesgeneral}\end{aligned}\ ] ] where the absorption rates , @xmath190 , are related to the spectrum of quantum noise , @xmath191 . emission rates have the same form with @xmath192 replaced by @xmath193 . for the considered initial condition , @xmath194 , the only non vanishing elements of the reduced density matrix in the eigenbasis of @xmath195 are the populations and the swap coherence , @xmath196 . for independent quantum noise sources acting on the two qubits , the swap decay rate reads @xmath197 = \frac{1}{2 } \ , [ \gamma_1 + \gamma_2 ] $ ] , where @xmath198 , @xmath199 , are relaxation rates of the swap levels . from eq.([eq : coherences ] ) we get @xmath200 the secular approximation is valid provided that @xmath201 . this condition is fulfilled , for instance , for white noise levels extrapolated from single charge - phase qubit experiments , as it can be evinced from the slow decay due to quantum noise reported in fig.([figure2 ] ) , gray line . equations ( [ eq : populations ] ) for the populations do not decouple even in the secular limit . general solutions are quite lengthy . here we report the approximate forms in the small temperature limit with respect to the uncoupled qubits splittings , @xmath202 . in this case level @xmath203 remains unpopulated and @xmath204 \quad , \quad \rho_{11}(t ) \approx \frac{1}{2 } \ , e^{- \gamma_{10 } t } \quad , \quad \rho_{22}(t ) \approx \frac{1}{2 } \ , e^{- \gamma_{20 } t } \ , . \label{populations}\end{aligned}\ ] ] _ second stage : elimination of adiabatic noise _ we now consider the effect of low - frequencies in the adiabatic , longitudinal and static path approximations . populations are unaffected by adiabatic noise , whereas the swap coherence , @xmath205 , has to be averaged over @xmath206 . here we disregard the negligible @xmath206 dependence of the rates @xmath207 via @xmath208 . under these simplifying assumption , we are left with the following average @xmath209 with @xmath210}$ ] . the swap splitting can be estimated by treating in perturbation theory @xmath211 , with @xmath180 , with respect to @xmath157 . this leads to @xmath212 the average in eq . ( [ eq : staticint ] ) can be evaluated in analytic form and gives @xcite @xmath213 \label{eq : spa - swap}\ ] ] where @xmath214 , and @xmath215 $ ] is the k - bessel function of order zero @xcite . we considered equal standard deviations for noise sources acting on both qubits , @xmath216 . inserting ( [ populations ] ) and ( [ eq : spa - swap ] ) in ( [ eq : psw1 ] ) and ( [ eq : psw2 ] ) we obtain the switching probabilities in the multi - stage approach . out of phase oscillations signals two - qubit states anti - correlations and follows from @xmath217 $ ] , with @xmath218 + \cos^2\left ( \frac{\varphi}{2}\right ) \approx - \frac{1}{4 } \ , \cos \varphi \ , \left [ e^{- \gamma_{10}t } + e^{- \gamma_{20}t } \right ] + g^2/4 $ ] . the efficiency of the gate results from the interplay of quantum and adiabatic noise . high - frequency noise levels expected from single qubit experiments @xcite weakly affect the switching probability , whose decay is mainly due to low - frequency noise , fig . ( [ figure2 ] ) . remarkably , considerable recovery of short - times oscillation amplitude may be achieved by an optimal choice of the coupling strength , @xmath219 @xcite . this regime is illustrated in fig.([figure3 ] ) . in this article we presented a road - map to treat broadband noise typical of solid state nanodevices . the introduced multi - stage approach allows to obtain reasonable approximations by systematically including only the relevant information on the complex environment , out of the huge parametrization which would be required for a microscopic description . since the environment is in general long - time correlated , the required information depends on the specific protocol . the predictions obtained with the present approach are in agreement with observations in various single qubit jj implementations and in different protocols @xcite . we extended the procedure to deal with complex solid - state architectures . this is a required step in order to predict efficiency and possibly appropriately design of nanodevices for quantum information processing . both because of the complexity of architectures and of the unavoidable broadband nature of solid state noise , theoretical tools allowing systematic and controlled approximations are particularly valuable . as an illustrative case , we performed a simplified analysis of the effects and interplay of low and high frequency noise components in a two - qubit gate in a fixed coupling scheme . our results points out that efficient operations in the solid state require an accurate preliminary characterization of the noise spectral characteristics and tuning appropriately the device working point . we acknowledge support from the eu - eurosqip ( ist-3 - 015708-ip ) . 99 m. nielsen , i. chuang , `` quantum computation and quantum information '' , cambridge univ . press , 2005 . t. hyashi _ et al . _ , lett . * 91 * , 226804 ( 2003 ) ; j. r. petta _ et al . lett . * 93 * , 186802 ( 2004 ) ; j.r . et al . _ , science * 309 * , 2180 ( 2005 ) ; j. gorman , d.g . hasko , and d.a . williams , phys . lett . * 95 * , 090502 ( 2005 ) ; f. h. l. koppens _ et al . _ , nature * 442 * , 766 ( 2006 ) . y. nakamura _ et al . _ , nature * 398 * , 786 ( 1999 ) ; y. yu _ et al . _ , science * 296 * , 889 ( 2002 ) ; j.m . martinis _ et al . _ , lett . * 89 * , 117901 ( 2002 ) ; i. chiorescu _ et al . _ , science , * 299 * , 1869 , ( 2003 ) ; t. yamamoto _ et al . _ , nature * 425 * , 941 ( 2003 ) ; s. saito _ et al . * 93 * , 037001 ( 2004 ) ; j. johansson _ et al . _ , ibid . * 96 * , 127006 ( 2006 ) ; f. deppe _ et al . _ , nat . * 4 * , 686 ( 2008 ) ; d. vion _ et al . _ , science * 296 * , 886 ( 2002 ) . j. koch et al . a * 76 * , 042319 ( 2007 ) ; j. a. schreier _ et al . _ , b * 77 * , 180502(r ) ( 2008 ) . d. i. schuster , _ et al . _ nature * 431 * , 162 ( 2004 ) ; d. i. schuster _ nature * 445 * , 515 ( 2007 ) ; m. gppl _ nature * 454 * , 315 ( 2008 ) ; lev s. bishop _ et al . _ nature physics * 5 * , 105 ( 2008 ) . y. nakamura _ et . al _ , phys . lett . * 88 * , 047901 ( 2002 ) . g. falci , e. paladino , r. fazio , in _ quantum phenomena of mesoscopic systems _ , b. l. altshuler and v. tognetti eds . , proc . of the international school `` enrico fermi '' , varenna 2002 , ios press ( 2003 ) , cond - mat/0312550 . lett . * 94 * , 167002 ( 2005 ) . c. p. slichter _ principles of magnetic resonance _ , springer - verlag , berlin ( 1996 ) s. kafanof _ et al . b , * 78 * , 125411 ( 2008 ) . o. astafiev _ lett . * 93 * , 267007 ( 2004 ) . g. ithier _ et al . _ , b * 72 * , 134519 ( 2005 ) . r.w . et al . _ , phys . lett . * 93 * , 077003 ( 2004 ) ; k.b . cooper _ et al . _ , ibid . * 93 * , 180401 ( 2004 ) . j. eroms _ et al . . lett . * 89 * , 122516 ( 2006 ) . c. h. van der wal , _ et al . _ , science 290 , 773777 ( 2000 ) . t. duty _ et al . _ , phys . rev . b * 69 * , 140503(r ) ( 2004 ) ; y. m. galperin _ 96 * , 097009 ( 2006 ) ; j bergli , y m galperin and b l altshuler , new j. phys . * 11 * , 025002 ( 2009 ) . f. chiarello private communication 2008 . experiment based on the setup of s.poletto _ et al . _ new j. phys . * 11 * , 013009 ( 2008 ) . r. fazio , g. m. palma , and j. siewert phys . 83 * , 5385 ( 1999 ) . , rev . * 60 * , 537 ( 1988 ) . j. claudon _ et al . _ , b * 76 * , 024508 ( 2007 ) ; l. tian and r.w . simmonds , phys . 99 * , 137002 ( 2007 ) ; f. deppe _ b * 76 * , 214503 ( 2007 ) ; z. kim , _ b * 78 * , 144506 ( 2008 ) ; a. lupascu _ _ arxiv:0810.0590 . a. m. zagoskin _ lett . * 97 * , 077001 ( 2006 ) . m. neeley , _ et al . _ nature physics * 4 * , 523 ( 2008 ) . e. paladino _ b * 77 * , 041303(r)(2008 ) . s. ashab _ _ , new j. phys . * 8 * , 103 ( 2006 ) ; n. p oxtoby _ et al . _ new j. phys . * 11 * , 063028 ( 2009 ) . c. cohen - tannoudji , j. dupont - roc and g. grynberg _ atom - photon interactions _ , wiley - interscience ( 1993 ) . e. paladino _ et al . _ , * 88 * , 228304 ( 2002 ) . correlated noise and cross - talk effects have been addressed in a. darrigo _ et al . njp * 10 * , 115006 ( 2008 ) . g. falci _ _ in preparation . w. zurek , physics today * 44 * , 36 ( 1991 ) . g. m. palma , k.a . suominen , a. k. ekert , proc . london a , * 452 * , 567 ( 1996 ) . d. g. cory _ * 81 * , 2152 ( 1998 ) . c. isham , _ lectures on quantum theory : mathematical and structural foundations _ , imperial college press , london ( 1995 ) . y. makhlin , a. shnirman , phys . lett . * 92 * , 178301 ( 2004 ) . a.b . zorin _ et al . _ , b * 53 * , 13682 ( 1996 ) . ( [ ref : quadratic ] ) is also valid for ornstein - uhlenbeck processes , see @xcite . for random telegraph noise the discrete nature of the process modifies this result see e. paladino _ state phys . , * 43 * , 747 ( 2003 ) k. rabenstein , v.a . sverdlov , d.v . averin jetp letters * 79 * , 646 ( 2004 ) . et al . _ , nature * 398 * , 305 ( 1999 ) ; j. q. you _ et al . lett . * 89 * , 197902 ( 2002 ) . . a. pashkin _ et al . _ , nature * 421 * , 823 ( 2003 ) ; a. j. berkley _ et al . _ , science * 300 * , 1548 ( 2003 ) ; t. yamamoto _ et al . _ , nature * 425 * , 941 ( 2003 ) ; r. mcdermott _ et al . _ , science * 307 * , 1299 ( 2005 ) ; j. b. majer _ lett . * 94 * , 090501 ( 2005 ) ; m. steffen _ et al . _ , science * 313 * , 1423 ( 2006 ) ; j.h . et al . _ , nature * 447 * , 836 ( 2007 ) . m. a. sillanp , j. i. park and r. w. simmonds nature * 449 * , 438 ( 2007 ) . averin , c. bruder , phys . * 91 * , 057003 ( 2003 ) ; a. blais _ et al . _ , ibid . * 90 * , 127901 ( 2003 ) ; f. plastina , g. falci , phys . b * 67 * , 224514 ( 2003 ) ; b.plourde _ et al . _ , ibid . * 70 * , 140501(r ) ( 2004 ) ; a.o . niskanen , y. nakamura , j.s . tsai , ibid . * 73 * , 094506 ( 2006 ) ; p. bertet , c. j. harmans , j.e . mooij , ibid . * 73 * , 064512 ( 2006 ) ; y - d wang , a. kemp , k. semba , ibid . * 79 * , 024502 ( 2009 ) . c. rigetti , a. blais , and m. devoret , phys . * 94 * , 240502 ( 2005 ) ; yu - xi liu _ et al . _ , ibid . * 96 * , 067003 ( 2006 ) ; g. s. paraoanu , phys . rev . b 74 , 140504(r ) ( 2006 ) . t. v. filippov _ et al . _ , ieee trans . * 13 * , 1005 ( 2003 ) ; j. lantz _ et al . b * 70 * , 140507(r ) ( 2004 ) ; a. maassen van den brink _ et al . _ , new j. phys . * 7 * , 230 ( 2005 ) . a. izmalkov _ * 93 * , 037003 ( 2004 ) ; m. grajcar _ _ , prb * 72 * , 02503 ( 2005 ) ; t. hime _ et al . _ , science * 314 * , 1427 ( 2006 ) ; a. o. niskanen _ _ , science * 316 * 723 ( 2007 ) ; a. o. niskanen _ et al . _ , nature * 447 * , 386 ( 2007 ) ; j. majer _ _ , ibid . * 449 * , 443 ( 2007 ) ; s. h. w. van der ploeg _ et al . _ , * 98 * , 057004 ( 2007 ) ; a. fay _ * 100 * , 187003 ( 2008 ) ; a. o. niskanen _ et al . _ , b * 77 * , 064505 ( 2008 ) ; k. harrabi _ _ , phys . rev . b 79 , 020507 ( 2009 ) . e. paladino , a. mastellone , a. darrigo , g. falci , arxiv:0906.3115 u. weiss , `` quantum dissipative systems '' , third edition , world scientific 2008 . m. abramowitz , i. a. stegun `` handbook of mathematical functions '' , dover ( 1965 ) . e. paladino _ at al . _ physica e in press ( 2009 ) .

peer - to - peer ( p2p ) communication in the internet is provided through the sharing of widely distributed resources typically involving end users computers acting as both clients and servers . in an unstructured peer - to - peer network , such as _ bittorrent _ @xcite , a file is divided into many pieces . seeds , which hold all pieces , distribute pieces to peers . new peers continually arrive into the network ; they simultaneously download pieces from a seed or other peers and upload pieces to other peers . peers exit the system after they collect all pieces . determining whether a given p2p network is stable can be difficult . roughly speaking , the aggregate transfer capacity scales up in proportion to the number of peers in the network , but it has to be in the right places . many p2p systems have performed well in practice , and they incorporate a variety of mechanisms to help achieve stability . a broad problem , which we address in part , is to provide a better understanding of which mechanisms are the most effective under various network settings . these mechanisms include * _ rarest first piece selection policies , _ such as the one implemented in bittorrent , whereby peers determine which pieces are rarest among their neighbors and preferentially download such pieces . * _ tit - for - tat participation constraints , _ such as the one implemented in bittorrent , whereby peers are choked off from receiving pieces from other peers unless they upload pieces to those same peers . this mechanism provides an important incentive for peers to participate in uploading pieces , but it may also be beneficial in balancing the distribution of pieces . * _ peers dwelling in the network after completing download , _ to provide extra upload capacity . * _ network coding @xcite , _ whereby data pieces are combined to form coded pieces , giving peers numerous ways to collect enough information to recover the original data file . this paper determines what parameter values yield stability for a simple model of a p2p file sharing network . the main model does not include the enhancements mentioned in the previous paragraph , but extensions and discussion regarding the above mechanisms is given . the model includes a fixed seed in the network that uploads with a constant rate . new peers arrive according to a poisson process , and have no pieces at the time of arrival . _ random peer contact _ is assumed ; each peer contacts a randomly selected target peer periodically . _ random useful piece selection _ is also assumed ; each peer chooses which piece to download uniformly at random from the set of pieces that its selected target has and it itself does not have . as in the bittorrent system , we assume that new peers arrive with no pieces ; in effect a peer must first obtain a piece from another peer or the fixed seed before it can begin uploading to other peers . we also assume that peers depart as soon as they have completed their collection . in a p2p network , the last few pieces to be downloaded by a peer are often rare in the network , so it usually takes the peer a long time to finish downloading . this phenomenon has been referred to as the _ delay in endgame mode _ @xcite ( or _ last piece problem _ ) . we refer to the specific situation that there are many peers in the network and most of them are missing only one piece which is the same for all peers , as the _ missing piece syndrome_. in that situation , peers lucky enough to get the missing piece usually depart immediately after getting the piece , so their ability to spread the missing piece is limited . the main result in this paper is to show , as suggested by the missing piece syndrome , that the bottleneck for stability is the upload capacity of the seed . specifically , if the arrival rate of new peers is greater than the seed upload rate , the number of peers in the system converges to infinity almost surely ; if the arrival rate of new peers is less than the seed upload rate , the system is positive recurrent and the mean number of peers in the system in equilibrium is finite . the next section gives the precise problem formulation , simulation results illustrating the missing piece syndrome , and the main proposition . the proposition is proved in sections [ sec : instability ] and [ sec : stability ] , with the help of some lemmas given in the appendix . section [ sec : extensions ] provides extensions of the result , including consideration of the enhancement mechanisms mentioned above . in particular , it is shown that the region of network stability is not increased if rarest first piece selection policies , or network coding policies , are applied . section [ sec : extensions ] also provides a conjecture regarding a refinement of the main proposition for the borderline case when the arrival rate is equal to the seeding rate ; it is suggested that whether the system is stable then depends on the rate that peers contact each other . the model in this paper is similar to the flat case of the open system of massouli and vojnovi @xcite . the model in @xcite is slightly different in that , rather than having a fixed seed , it assumes that new peers each arrive with a randomly selected piece . a fluid model , based on the theory of density - dependent jump markov processes ( see @xcite ) , is derived and studied in @xcite . it is shown that there is a finite resting point of the fluid ordinary differential equation . the analysis in this paper is different and complementary . rather than appealing to fluid limits , we focus on direct stochastic analysis methods , namely using coupling to prove transience for some parameter values and the foster - lyapunov stability criterion to prove positive recurrence for complementary parameter values . furthermore , our work shows the importance of considering asymmetric sample paths even for symmetric system dynamics . forthcoming work described in @xcite provides analysis of p2p networks with peers having pieces upon arrival , as in @xcite , and with peers remaining for some time in the system after obtaining a complete collection . some other works related to stability and the missing piece syndrome are the following . the instability phenomenon identified in this paper was discovered independently by norros et al . norros et al . @xcite proved a version of our main proposition for a similar model , for the case of two pieces . in the model of @xcite a peer receives one piece on arrival , with the distribution of the piece number ( either one or two ) being determined by sampling uniformly from the group consisting of a fixed seed and the population of peers already in the system . menasch et al @xcite pointed out that in their simulation studies , their smooth download assumption " and swarm sustainability " break down if the seed upload rate is not sufficiently large . leskel et al . @xcite investigate stability conditions for a single piece file , or a two piece file when the pieces are obtained sequentially , when peers remain in the system for some time after obtaining the piece . the earliest papers to analytically study unstructured peer - to - peer files systems with arrivals of new peers are @xcite . these papers provide simple models in which a two dimensional differential equation is used that does not take into account the stages of service as peers gain more pieces . the model in this paper is a composite of models in @xcite . it incorporates poisson arrivals , fixed seed , random uniform contacts , and random useful piece selection , as follows . the parameters of the model are an integer @xmath0 and strictly positive constants @xmath1 and @xmath2 * there are @xmath3 pieces and @xmath4 so that @xmath5 indexes all the pieces . * the set of proper subsets of @xmath5 is denoted by @xmath6 * a peer with set of pieces @xmath7 for some @xmath8 is called a _ type @xmath9 peer_. * a type @xmath9 peer becomes a type @xmath10 peer if it downloads piece @xmath11 for some @xmath12 * a markov state is @xmath13 with @xmath14 denoting the number of type @xmath9 peers , @xmath15 denoting the number of peers in the system , and @xmath16 denoting the state space of the system . * peers arrive exogenously one at a time with no pieces ; the times of arrival form a rate @xmath17 poisson process . * each peer contacts other peers , chosen uniformly at random from among all peers , for opportunities to download a piece ( i.e. pull ) from the other peers , according to a poisson process of rate @xmath18 mathematically , an equivalent assumption is the following . each peer contacts other peers , chosen uniformly at random from among all peers , for opportunities to upload a piece ( i.e. push ) to the other peers , according to a poisson process of rate @xmath18 * downloads are modeled as being instantaneous . this assumption is reasonable in the context of the previous assumption . * random useful piece selection is used , meaning that when a peer of type @xmath9 has an opportunity to download a piece from a peer of type @xmath19 the opportunity results in no change of state if @xmath20 otherwise , the type @xmath9 peer downloads one piece selected at random from @xmath21 , with all @xmath22 possibilities having equal probability . * there is one fixed seed , which at each time in a sequence of times forming a poisson process of rate @xmath23 selects a peer at random and uploads a random useful piece to the selected peer . * peers leave immediately after obtaining a complete collection . given a state @xmath24 , let @xmath25 denote the new state resulting from the arrival of a new peer . given @xmath8 @xmath26 such that @xmath27 and a state @xmath24 such that @xmath28 let @xmath29 denote the new state resulting from a type @xmath9 peer downloading piece @xmath30 the positive entries of the generator matrix @xmath31 of the markov process are given by : @xmath32 to provide some intuition , we present some simulation results . figure [ fig.peers_sim ] shows simulations of the system for @xmath33 and @xmath34 pieces . ( dashed ) and @xmath35 ( solid ) , and the second is for @xmath36 ( dashed ) and @xmath37 ( solid ) . [ fig.peers_sim ] , title="fig:",width=226 ] ( dashed ) and @xmath35 ( solid ) , and the second is for @xmath36 ( dashed ) and @xmath37 ( solid ) . [ fig.peers_sim ] , title="fig:",width=226 ] the first plot shows apparently stable behavior . after an initial spike , the number of peers in the system seems to hover around 30 ( for @xmath38 ) or 45 ( for @xmath35 ) , which by little s law is consistent with a mean time in system around 50 to 60 time units ( or about 25% to 50% larger than the sum of the download times ) . however , the second plot shows that for @xmath36 or @xmath39 the number of peers in the system does not appear to stabilize , but rather to grow linearly . the explanation for this instability is indicated in figure [ fig.pieces_sim ] , which shows the time - averaged number of peers that held each given piece during the simulations , for @xmath38 in the first plot and for @xmath37 in the second plot . the first plot shows that the 40 pieces had nearly equal presence in the peers , with piece 7 being the least represented . the second plot shows that 39 pieces had nearly equal presence and most of the peers had these pieces most of the time , but only a small number of peers held piece 3 . and the second is for @xmath40 the dashed lines indicate time - average number of peers in system . [ fig.pieces_sim ] , title="fig:",width=377 ] and the second is for @xmath40 the dashed lines indicate time - average number of peers in system . [ fig.pieces_sim ] , title="fig:",width=377 ] the following proposition , which is the main result of this paper , confirms that the intuition behind the simulation results is correct . [ prop.seed_only ] ( i ) if @xmath41 then the markov process is transient , and the number of peers in the system converges to infinity with probability one . ( ii ) if @xmath42 the markov process with generator @xmath43 is positive recurrent , and the equilibrium distribution @xmath44 is such that @xmath45 in the remainder of this section , we give an intuitive explanation for the proposition , which also guides the proof . we first give an intuitive justification of proposition [ prop.seed_only](i ) , so assume @xmath46 under this condition , eventually , due to random fluctuations , there will be many peers in the system that are all missing the same piece . while any of the @xmath3 pieces could be the missing one , to be definite we focus on the case that the peers are missing piece one . a peer is said to be in the _ one club _ , or to be a one - club peer , if it has all pieces except piece one . we consider the system starting from an initial state in which there are many peers in the system , and all of them are in the one club . the system then evolves as shown in figure [ fig.oneclub ] . , title="fig:",width=264 ] the large size of the box showing the one club indicates that most peers are one club peers . a peer not in the one club is said to be a _ young peer _ , and a young peer is said to be _ normal _ if it does not have piece one and _ infected _ if it does have piece one . since there are so many one club peers to download from , a peer does nt stay young very long ; most of the young peers join the one club soon after arrival . however , due to the fixed seed uploading pieces , some of the normal young peers become infected peers . those infected peers can infect yet more young peers , thereby forming a branching process . but typically the infected young peers do not infect other young peers , so that the branching process is highly subcritical . therefore , the rate of departures from the one club due to uploads of piece one from infected peers is small . therefore , most peers eventually enter the one club , and the main way that peers leave the one club is to receive piece one directly from the fixed seed . so the long term arrival rate at the one club is close to @xmath17 and the departure rate from the one club is close to @xmath2 therefore , the one club can grow at rate close to @xmath47 while the number of young peers will stay about constant . these ideas are made precise in the proof . to understand why the system is stable for @xmath48 the rough idea is to show that whenever there are many peers in the system , no matter what the distribution of pieces they hold , the system moves towards emptying out . if there are many peers in the system , one of the following two cases holds . the first case is that most of the peers have the same number , say @xmath49 , of pieces . intuitively , the worst case would be for all peers with @xmath49 pieces to have identical collections of pieces , in which case no peer with @xmath49 pieces would be useful to another . however , if @xmath48 such a state ca nt persist , because peers with @xmath49 pieces get additional pieces from the fixed seed at an aggregate rate near @xmath23 while the long term rate that new peers with exactly @xmath49 pieces can appear is less than or equal to @xmath50 if the system is not in the first case just described , then there are at least two sizeable groups of peers , so that all the peers in the first group have one number of pieces and all peers in the second group have some larger number of pieces . then all peers in the second group can be helpful to any peer in the first group , so that there will be a large rate of downloads . thus , if there are many peers in the system , no distribution of the pieces they hold can persist . to prove stability , it is still necessary to show that the state ca nt spiral out to ever increasing loads through some quasi - periodic behavior . this is achieved through the use of a potential function and the foster - lyapunov stability criterion . proposition [ prop.seed_only](i ) is proved in this section ; it can be read independently of the proof of proposition [ prop.seed_only](ii ) in the next section . the proof follows along the lines of the intuitive explanation given just after the statement of the proposition in section [ sec : formulation ] , and an additional explanation of the proof is provided in a remark at the end of the section . assume @xmath46 if @xmath51 the system reduces to an @xmath52 queueing system with arrival rate @xmath17 and departure rate @xmath53 , in which case the number of peers in the system converges to infinity with probability one . so for the remainder of this proof assume @xmath54 to begin : * select @xmath55 so that @xmath56 * select @xmath57 so that @xmath58 and @xmath59 it follows from that @xmath60 * select @xmath61 small enough that @xmath62 * select @xmath63 large enough that @xmath64}2^{-b}}{1 - 2^{-\epsilon_o}}\leq 0.1 , \label{eq.bmminfty } \\ \frac{64k^2 \xi u_s}{2b(\epsilon-4k\xi u_s ) } \leq 0.1 , \label{eq.bmm1 } \\ \frac{\lambda}{2b\epsilon } \leq 0.1,~~\mbox{and}~~\frac{u_s}{2b\epsilon } \leq 0.1 . \label{eq.bpoi}\end{aligned}\ ] ] * select @xmath65 large enough that @xmath66 . we shall use the notions of one club , young peer , and infected young peer , as described in the paragraph after proposition [ prop.seed_only ] . for a given time @xmath67 define the following random variables : * @xmath68 cumulative number of arrivals , up to time @xmath69 * @xmath70 number of peers at time @xmath69 * @xmath71 number of young peers at time @xmath69 * @xmath72 cumulative number of uploads of piece one by infected peers , up to time @xmath69 * @xmath73 cumulative number of uploads of piece one by the fixed seed , up to time @xmath69 the system is modeled by an irreducible , countable - state markov process . a property of such random processes is that either all states are transient , or no state is transient . therefore , to prove proposition [ prop.seed_only](i ) , it is sufficient to prove that some particular state is transient . with that in mind , we assume that the initial state is the one with @xmath65 peers , and all of them are one - club peers . let @xmath74 be the extended stopping time defined by @xmath75 with the usual convention that @xmath76 if @xmath77 for all @xmath78 it suffices to prove that @xmath79 the equation depends on the transition rates of the system out of states such that @xmath80 thus , we can and will prove instead for an alternative system , that has the same initial state , and the same out - going transition rates for all states such that @xmath81 as the original system . the alternative system is defined by modifying the original system by letting the rate of downloads from the set of one - club peers by each young peer be @xmath82 , and the aggregate rate of downloads from the fixed seed to the set of young peers be @xmath83 note that the rates used for this definition are equal to the original ones on the states such that @xmath84 as required . the alternative system has the following two properties : 1 . each young peer receives opportunities to download from one - club peers at rate greater than or equal to @xmath85 2 . the fixed seed contacts the entire population of young peers at aggregate rate less than or equal to @xmath86 _ for the remainder of this proof we consider the alternative system , but for brevity of notation , use the same notation for it as for the original system , and refer to it as the original system . _ the following four inequalities will be established , for @xmath87 and @xmath65 satisfying the conditions given near the beginning of the section . @xmath88 let @xmath89 be the intersection of the four events on the left sides of - . since @xmath90 is greater than or equal to the number of peers in the system that do nt have piece one , on @xmath89 , + @xmath91 for all @xmath92 therefore , on @xmath89 , for any @xmath67 @xmath93 thus , @xmath89 is a subset of the event in . therefore , if - hold , @xmath94 and is implied . so to complete the proof , it remains to prove - . the process @xmath95 is a poisson process with rate @xmath17 , and @xmath96 is stochastically dominated by a poisson process with rate @xmath53 . thus , both and follow from kingman s moment bound ( see lemma [ lemma : kingman ] in the appendix ) and the conditions in on @xmath97 turning next to the proof of , we shall use the following observation about _ stochastic domination _ ( the notion of stochastic domination is reviewed in the appendix ) . the observation is a mathematical version of the statement that the number of young peers remains roughly bounded because peers do nt stay young for long . [ lemma.mg1compare ] the process @xmath98 is stochastically dominated by the number of customers in an @xmath99 queueing system with initial state zero , arrival rate @xmath100 and service times having the gamma distribution with parameters @xmath101 and @xmath85 the idea of the proof is to show how , with a possible enlargement of the underlying probability space , an @xmath99 system can be constructed on the same probability space as the original system , so that for any time @xmath69 , @xmath102 is less than or equal to the number of peers in the @xmath99 system . let the @xmath99 system have the same arrival process as the original system it is a poisson process of rate @xmath50 for any young peer , the intensity of downloads from the one club ( i.e. from any peer in the one club ) is always greater than or equal to @xmath103 for the original system , where we use the fact @xmath104 which is true by and the assumption @xmath54 we can thus suppose that each young peer has an internal poisson clock , which generates ticks at rate @xmath103 , and is such that whenever the internal clock of a young peer ticks , that young peer downloads a piece from the one club . we declare that a peer remains in the @xmath99 system until its internal clock ticks @xmath101 times . this gives the correct service time distribution , and the service times of different peers in the @xmath99 system are independent , as required . a young peer can possibly leave the original system sooner than it leaves the @xmath99 system , because a young peer in the original system can possibly download pieces at times when its internal clock does nt tick . but if a young peer is still in the original system , it is in the @xmath99 system . given this lemma , follows from lemma [ lemma.mginfty ] with @xmath105 in the lemma equal to @xmath106 , and @xmath107 in the lemma equal to @xmath108 and . it remains to prove . + consider the following construction of a stochastic system that is similar to the original one , with random variables that have similar interpretations , but with different joint distributions . we call it the _ comparison system . _ it focuses on the infected peers and the uploads by infected peers , and it is specified in table [ table.compare ] . .specification of comparison system [ cols="<,<",options="header " , ] it should be clear to the reader that both the original system and the comparison system can be constructed on the same underlying probability space such that any infected peer in the original system at a given time is also in the comparison system . when such a peer becomes infected in the original system , we require that it also arrives to the comparison system , it discards all pieces it may have downloaded before becoming infected , and it subsequently ignores all opportunities to download except those occurring at the times its internal clock ( described in the proof of lemma [ lemma.mg1compare ] ) ticks . because infected young peers possibly stay longer in the comparison system than in the original system , some of the peers in the comparison system correspond to peers that already departed from the original system . there can also be some infected peers in the comparison system that never existed in the original system because the arrival rate of infected peers to the comparison system is greater than the arrival rate for the original system . but whenever there is an infected peer in the original system , that peer is also in the comparison system , and the following property holds . whenever any one of the following events happens in the original system , it also happens in the comparison system : * the fixed seed creates an infected peer . * an infected peer creates an infected peer * an infected peer uploads piece one to a one - club peer events of the second and third type just listed correspond to the two possible ways that infected peers can upload piece one . therefore , the property implies the following lemma , where @xmath109 is the cumulative number of uploads of piece one by infected peers , up to time @xmath69 , in the comparison system . [ lemma.onecompare ] the process @xmath110 is stochastically dominated by @xmath111 we can identify two kinds of infected peers in the comparison system the _ root peers _ , which are those created by the fixed seed , and the infected peers created by other infected peers . we can imagine that each root peer affixes its unique signature on the copy of piece one that it receives from the fixed seed . the signature is inherited by all copies of piece one subsequently generated from that piece through all generations of the replication process , in which infected peers upload piece one when creating new infected peers . in this way , any upload of piece one by an infected peer can be traced back to a unique root peer . in summary , the jumps of @xmath109 can be partitioned according to which root peer generated them . of course , the jumps of @xmath109 associated with a root peer happen after the root peer arrives . let @xmath112 denote a new process which results when all of the uploads of piece one generated by a root peer ( in the comparison system ) are counted at the arrival time of the root peer . since @xmath113 counts the same events as @xmath114 but does so earlier , @xmath115 for all @xmath92 in view of this and lemma [ lemma.onecompare ] , it is sufficient to prove with @xmath116 replaced by @xmath117 the random process @xmath113 is a compound poisson process . jumps occur at the arrival times of root peers in the comparison system , which form a poisson process of rate @xmath86 let @xmath118 denote the size of the jump of @xmath113 associated with a typical root peer . the distribution of @xmath118 can be described by referring to an @xmath119 queueing system with arrival rate @xmath120 and service times having the distribution of a random variable @xmath121 which has the gamma distribution with parameters @xmath101 and @xmath85 note that @xmath122 in is the usual load factor for the reference queueing system : @xmath123.$ ] the reference queueing system is similar to the number of infected peers in the comparison system , except that the customers in the @xmath119 queueing system are served one at a time . we have @xmath124 where * @xmath125 is the number of infected peers that are descendants of the root peer ( not counting the root peer itself . ) that includes peers directly created by the root peer , peers created by peers created by the root peer , and so on , for all generations . @xmath125 has the same distribution as the number of customers in a busy period of the reference queueing system , not counting the customer that started the busy period . * @xmath126 is the number of uploads of piece one to one - club peers by either the root peer or any of the descendants of the root peer . the sum of all the times that the root peer and its descendants are in the comparison system is the same as the duration , @xmath127 of a busy period of the reference queueing system . while in the comparison system , those peers upload piece one to the one club with intensity @xmath128 so @xmath129=\mu e[l]$ ] and @xmath130=\mu^2 e[l]^2 + \mu e[l].$ ] using this stochastic description , the formulas for the busy period in an @xmath119 queueing system ( and in the appendix ) , and the facts @xmath131 @xmath132=2(k-1)/\mu,$ ] and @xmath133 yields @xmath134 & = & e[j_1]+e[j_2 ] = \frac{1+\mu e[\widehat { x}]}{1-\rho } -1\\ & \leq & 2[1 + 2(k-1 ) ] \leq 4k\end{aligned}\ ] ] and @xmath135 \leq e[(j_1 + 1)^2 ] = \frac{1+(\xi u_s)^2 \mbox{var}(\widehat { x})}{(1-\rho)^3 } \leq \frac{1+\rho^2}{(1-\rho)^3}\ ] ] @xmath136 & = & e [ e[j_2 ^ 2 | l ] ] = \mu e[l ] + \mu^2 e[l^2 ] \\ & = & \frac{\mu e[\widehat { x}]}{1-\rho } + \frac{\mu^2 e[\widehat { x}^2 ] } { ( 1-\rho)^3}\end{aligned}\ ] ] @xmath137}\\ & = & e[(j_1 + j_2)^2 ] \leq 2\ { e[j_1 ^ 2 ] + e[j_2 ^ 2 ] \ } \\ & \leq & 16\ { 2 + \mu e[\widehat { x } ] + \mu^2 e[\widehat { x}^2 ] \ } \\ & = & 16\left\ { 2 + 2(k-1 ) + 4(k-1 ) + 4(k-1)^2 \right\ } \\ & = & 16\left\ { 4k^2 - 2k \right \ } \leq 64k^2 \end{aligned}\ ] ] thus , @xmath113 is a compound poisson process with arrival rate of batches equal to @xmath138 and batch sizes with first and second moments of the batch sizes bounded by @xmath139 and @xmath140 respectively . hence , with @xmath116 replaced by @xmath113 follows from corollary [ cor : compoundkingman ] and . the proof of proposition [ prop.seed_only](i ) is complete . we briefly explain why the comparison system was introduced in the above proof , to provide a better understanding of the proof technique . the intuitive idea behind the definition of the comparison system is that it is based on worst case assumptions regarding the number of peers that are infected by the fixed seed ( i.e. the number of root peers ) and the number of uploads of piece one that can be caused by each root peer . the advantage is then that the arrivals of root peers form a poisson process and the total number of uploads of piece one that can be traced back to different root peers are independent in the comparison system , so that kingman s bound for compound poisson processes , which is a form of the law of large numbers , can be applied . proposition [ prop.seed_only](ii ) is proved in this section , using the version of the foster - lyapunov stability criterion given in the appendix , and the intuition given in the last paragraph of section [ sec : formulation ] . if @xmath141 is a function on the state space @xmath142 then @xmath143 is the corresponding drift function , defined by @xmath144 . $ ] if , as usual , the diagonal entries of @xmath43 are defined to make the row sums zero , then the drift function is also given by matrix - vector multiplication : @xmath145 suppose @xmath146 given a state @xmath24 , let @xmath147 that is , @xmath148 is the number of peers with precisely @xmath11 pieces . when the dependence on @xmath24 is clear , we write @xmath149 instead of @xmath150 we shall use the foster - lyapunov criteria with the following potential function : @xmath151 where @xmath152 are positive constants and @xmath153 the following two lemmas and their proofs correspond to the two cases described in the intuitive description given in the last paragraph of section [ sec : formulation ] . [ lem : d_lf_k ] there exist positive values @xmath170 and @xmath171 so that @xmath172 whenever : @xmath173 and , for some @xmath11 , @xmath174 [ lem : d_lf_nk ] let @xmath175 be as in lemma [ lem : d_lf_k ] . there exist positive values @xmath176 and @xmath177 so that @xmath178 whenever : @xmath179 and , for all @xmath11 , @xmath180 lemmas [ lem : d_lf_k ] and [ lem : d_lf_nk ] imply that @xmath181 whenever @xmath182 , so that @xmath43 and @xmath141 satisfy the conditions of proposition [ cor.fostercompcontinuous ] with @xmath183 and @xmath184 where @xmath185 therefore , to complete the proof of proposition [ prop.seed_only](ii ) it remains to prove lemmas [ lem : d_lf_k ] and [ lem : d_lf_nk ] . ( proof of lemma [ lem : d_lf_k ] . ) it suffices to prove the lemma for an arbitrary choice of @xmath30 so fix @xmath186 and consider a state @xmath24 such that @xmath187 ( and , in particular , @xmath188 ) . then for any @xmath189 , @xmath190 use and and an interchange of summation ( @xmath191 ) to get @xmath192 notice that according to , @xmath193 and @xmath194 thus , if @xmath175 is small enough and @xmath195 is large enough , the quantity within braces in is negative . therefore , if @xmath175 and @xmath107 are small enough , and @xmath171 is large enough , @xmath196 under the conditions of the lemma , whenever @xmath197 lemma [ lem : d_lf_k ] is proved . ( proof of lemma [ lem : d_lf_nk ] . ) let @xmath175 be given by lemma [ lem : d_lf_k ] , and consider a state @xmath24 such that @xmath198 for all @xmath11 . it follows that there exists @xmath199 and @xmath200 with @xmath201 such that @xmath202 and @xmath203 then @xmath204 the conclusion of the lemma follows because of the term in that is quadratic in @xmath205 a piece selection policy is used by a peer to choose which piece to download whenever it contacts another peer . the random useful piece selection policy is assumed above , but the results extend to a large class of piece selection policies . essentially the only restriction needed is that if the contacted peer has a useful piece for the contacting peer , then a useful piece must be downloaded . this restriction is similar to a work conserving restriction in the theory of service systems . in particular , the results hold for a broad class of rarest first piece selection policies . peers can estimate which pieces are more rare in a distributed way , by exchanging information with the peers they contact . even more general policies would allow the piece selection to depend in an arbitrary way on the piece collections of all peers . interestingly enough , the results extend even to seemingly bad piece selection policies . for example , it includes the sequential piece selection policy , in which peers obtain the pieces in order , beginning with piece one . the sequential policy can be viewed as a _ most abundant first _ useful piece selection policy , or just the opposite of rarest piece first . to be specific , consider the following family @xmath206 of piece selection policies . each policy in @xmath206 corresponds to a mapping @xmath207 from @xmath208 to the set of probability distributions on @xmath209 satisfying the usefulness constraint : @xmath210 with the following meaning of @xmath207 : * when a type @xmath95 peer selects a piece to download from a type @xmath63 peer and the state of the entire network is @xmath211 piece @xmath11 is selected with probability @xmath212 for @xmath213 * when the fixed seed selects a piece to upload to a type @xmath95 peer and the state of the entire network is @xmath211 piece @xmath11 is selected with probability @xmath214 for @xmath213 the piece selection policies noted above are included in @xmath215 reconsider the proof of transience in section [ sec : instability ] under a piece selection policy in @xmath206 . from any state it is possible to reach the empty state , and from the empty state it is possible to reach a state with one peer in the network having all pieces except some piece @xmath216 from that state , for any @xmath217 , it is possible to reach the state with @xmath65 peers missing only piece @xmath218 and no other peers in the network . it may be impossible for @xmath219 to equal one , but by renumbering the pieces if necessary , it can be assumed without loss of generality that @xmath219 is one . thus , whatever piece selection policy in @xmath206 is applied , beginning from any initial state , for any @xmath220 in a finite time with a positive probability , the system can arrive into the state where there are @xmath65 peers and all of them are one - club peers . thus , as in section [ sec : instability ] , to prove transience it suffices to show that from such an initial state , there is a positive probability that the number of peers converges to infinity . the arrival rate of new peers and the upload rate of the seed does not depend on the piece selection policy , so and are valid for any piece selection policies in @xmath206 . moreover , lemma [ lemma.mg1compare ] and lemma [ lemma.onecompare ] are valid for any piece selection policies in @xmath206 because the two lemmas depend on the properties that peer selection is uniformly random and the piece selection is useful if a useful piece is available . therefore and are also valid for any piece selection policy in @xmath206 . thus , we conclude that the proof of proposition [ prop.seed_only](i ) in section [ sec : instability ] works for any piece selection policy in @xmath215 reconsider next the proof of positive recurrence in section [ sec : stability ] , but for an arbitrary piece selection policy in @xmath215 the inequalities developed for the proofs of lemmas [ lem : d_lf_k ] and [ lem : d_lf_nk ] hold with the same lyapunov function ; useful piece selection suffices . thus , if @xmath48 it can be shown that the lyapunov stability condition , namely @xmath221 for @xmath15 sufficiently large , still holds . the final conclusion has to be modified , however , because under some policies in @xmath222 the markov process might no longer be irreducible . for example , with the sequential useful piece selection policy , the set of states such that every peer holds a set of pieces of the form @xmath223 for some @xmath118 with @xmath224 is a closed subset of states , in the terminology of classification of states of discrete - state markov processes . in general , the set of all states that are reachable from the empty state is the unique minimal closed set of states , and the process restricted to that set of states is irreducible . by a minor variation of the foster - lyapunov stability proposition , the lyapunov stability condition implies that the markov process restricted to that closed set of states is positive recurrent , and the mean time to reach the empty state beginning from an arbitrary initial state is finite . we summarize the discussion of the previous two paragraphs as a proposition . ( stability conditions for general useful piece selection policies ) suppose a useful piece selection policy from @xmath206 is used , for a network with random peer contacts and parameters @xmath3 , @xmath17 , @xmath23 and @xmath225 as in section [ sec : formulation ] . there is a single class of closed states containing the empty state , and all other states are transient . ( i ) if @xmath41 then the markov process is transient , and the number of peers in the system converges to infinity with probability one . ( ii ) if @xmath42 the markov process with generator @xmath43 restricted to the closed set of states is positive recurrent , the mean time to reach the empty state from any initial state has finite mean , and the equilibrium distribution @xmath44 is such that @xmath45 thus , with the exception of the borderline case @xmath226 rarest first piece selection does not increase the region of stability , nor does most abundant first piece selection decrease the region of stability . network coding , introduced by ahlswede , cai , and yeung , @xcite , can be naturally incorporated into p2p distribution networks , as noted in @xcite . the related work @xcite considers all to all exchange of pieces among a fixed population of peers through random contacts and network coding . the method can be described as follows . the file to be transmitted is divided into @xmath3 data pieces , @xmath227 the data pieces are taken to be vectors of some fixed length @xmath228 over a finite field @xmath229 with @xmath230 elements , where @xmath230 is some power of a prime number . if the piece size is @xmath231 bits , this can be done by viewing each message as an @xmath232 dimensional vector over @xmath233 any coded piece @xmath234 is a linear combination of the original @xmath3 data pieces : @xmath235 the vector of coefficients @xmath236 is called the _ coding vector _ of the coded piece ; the coding vector is included whenever a coded piece is sent . the fixed seed uploads coded pieces to peers , and peers exchange coded pieces . in this context , the type of a peer @xmath95 is the subspace @xmath237 of @xmath238 spanned by the coding vectors of the coded pieces it has received . once the dimension of @xmath237 reaches @xmath3 , peer @xmath95 can recover the original message . when peer a contacts peer b , suppose peer b sends peer a a random linear combination of its coded pieces , where the coefficients are independent and uniformly distributed over @xmath233 equivalently , the coding vector of the coded piece sent from @xmath63 is uniformly distributed over @xmath239 the coded piece is considered useful to @xmath95 if adding it to @xmath95 s collection of coded pieces increases the dimension of @xmath240 equivalently , the piece from @xmath63 is useful to @xmath95 if its coding vector is not in the subspace @xmath241 the probability the piece is useful to @xmath95 is therefore given by @xmath242 if peer @xmath63 can possibly help peer @xmath95 , meaning @xmath243 ( true , for example , if @xmath244 ) , the probability that a random coded piece from b is helpful to a is greater than or equal to @xmath245 the probability a random coded piece from the seed is useful to a peer @xmath95 with @xmath246 is precisely @xmath245 therefore , when all peers have the same state and the common state has dimension @xmath101 , the departure rate from the network is @xmath247 the network state @xmath24 specifies the number of peers in the network of each type . there are only finitely many types , so the overall state space is still countably infinite . moreover , the markov process is easily seen to be irreducible . reconsider the proof of transience in section [ sec : instability ] , but now under network coding . fix any subspace @xmath248 of @xmath238 with dimension @xmath249 call a peer a one - club peer if its state is @xmath250 for any @xmath217 , it is possible to reach the state with @xmath65 one - club peers and no other peers in the network . as before , call a peer a young peer if it is not a one - club peer . in the case of network coding , call a peer infected if its state is not a subspace of @xmath250 the only way a peer can become infected is by downloading a piece either from the seed or from an infected peer . lemmas [ eq.aineq ] and [ eq.zineq ] are valid for network coding , if the condition @xmath41 is replaced by @xmath251 moreover , lemma [ lemma.mg1compare ] and lemma [ lemma.onecompare ] are valid for network coding because the two lemmas depend on the properties that peer selection is uniformly random and the rate useful pieces are delivered by the seed to one - club peers is arbitrarily close to @xmath252 thus , we conclude that proposition [ prop.seed_only](i ) in section [ sec : instability ] , with @xmath53 replaced by @xmath253 extends to the case of network coding . reconsider the proof of positive recurrence in section [ sec : stability ] , but with random useful piece selection replaced by network coding as described , and @xmath53 replaced by @xmath254 suppose the same lyapunov function is used , except the new meaning of @xmath255 or @xmath149 for short , is the number of peers @xmath95 with @xmath256 lemmas [ lem : d_lf_k ] and [ lem : d_lf_nk ] are valid for network coding , if the condition @xmath42 is replaced by @xmath257 thus , if @xmath258 it can be shown that the lyapunov stability condition , namely @xmath221 for @xmath15 sufficiently large , still holds , and the foster - lyapunov stability criterion applies . we summarize the discussion of the previous two paragraphs as a proposition . ( stability conditions for network coding based system ) suppose random linear network coding with vectors over @xmath238 is used , with random peer contacts and parameters @xmath3 , @xmath17 , @xmath23 and @xmath225 as in section [ sec : formulation ] . ( i ) if @xmath259 then the markov process is transient , and the number of peers in the system converges to infinity with probability one . ( ii ) if @xmath260 the markov process is positive recurrent , and the equilibrium distribution @xmath44 is such that @xmath45 thus , as @xmath261 the stability region for the system with network coding converges to that for useful piece selection . network coding has the advantage that no exchange of state information among peers is needed because there is no need to identify useful pieces . in many unstructured peer - to - peer systems , such as bittorrent , peers often dwell in the network awhile after they have collected all the pieces . in effect , these peers temporarily become seeds , called peer seeds . the uploading provided by peer seeds is able to mitigate the missing piece syndrome and enlarge the stability region . intuitively , if every peer can upload , on average , just one more piece after collecting all pieces , then every peer can help one one - club peer to depart , so the missing piece syndrome would not persist . this is explored for the case of @xmath262 and @xmath263 ( for the sequential piece selection policy ) in @xcite and for random useful piece selection with arbitrary @xmath0 in @xcite . another way to overcome the missing piece syndrome relies on peer selection policies . for instance , if young peers contact infected peers preferentially , or if the seed uploads to young peers preferentially , the network can be stabilized by the resulting increase in the number of infected peers . so some sort of coordination policy , providing the identification of rare pieces and young peers , and the transmission of the rare pieces to the young peers , can counter the missing piece syndrome . a mechanism built into bittorrent , called tit - for - tat operation , may alter the peer selection policy enough to yield stability for any choice of @xmath1 and @xmath2 under tit - for - tat operation , peers upload almost exclusively to peers from which they can simultaneously download . an obvious benefit of tit - for - tat is to give peers incentive to upload , thereby helping other peers , but it also may be effective against the missing piece syndrome . specifically , tit - for - tat encourages one - club peers to reduce their rate of download to the young peers , because the young peers have nothing to upload to the one - club members . this increases the amount of time that peers remain young , giving them a greater chance to obtain a rare piece from the fixed seed . also , infected peers would preferentially send to young peers , because often a normal young peer and an infected young peer would be able to help each other . while it is thus clear that tit - for - tat operation helps combat the missing piece syndrome , we leave open the problem of quantifying the effect for a specific model . we have shown that , for any @xmath264 the system is stable if @xmath42 and unstable if @xmath265 and this result is insensitive to the value of @xmath225 and to the piece selection policy , as long as a useful piece is selected whenvever possible . while it may not be interesting from a practical point of view , we comment on the case @xmath266 first , we give a precise result for a limiting case of the original system , and then we offer a conjecture . a simpler network model results by taking a limit as @xmath267 call a state _ slow _ if all peers in the system have the same type , which includes the state such that there are no peers in the system . otherwise , call a state _ fast_. the total rate of transition out of any slow state does not depend on @xmath268 and the total rate out of any fast state is greater than or equal to @xmath85 for very large values of @xmath268 the process spends most of its time in slow states . the original markov process can be transformed into a new one by _ watching _ the original process while it is in the set of slow states . this means removing the portions of each sample path during which the process is in fast states , and time - shifting the remaining parts of the sample path to leave no gaps in time . the limiting markov process , which we call the @xmath269 process , is the weak limit ( defined as usual for probability measures on the space of cdlg sample paths equipped with the skorohod topology ) of the original process watched in the set of slow states , as @xmath267 by symmetry of the model , the state space of the @xmath269 process can be reduced further , to @xmath270 where a state @xmath271 corresponds to @xmath272 peers in the system which all possess the same set of @xmath273 pieces . the positive transition rates of the @xmath269 process are given by : @xmath274 and the transition rate diagram is pictured in figure [ fig.mu_infinity ] for @xmath275 system for @xmath275,title="fig:",width=226 ] the top layer of states consists of those for which all peers have @xmath101 pieces . these states correspond to all peers being in the one club , or all missing some other piece . from any state the process reaches the top layer in mean time less than or equal to @xmath276 and within the top layer the process behaves like a birth - death process with birth rate @xmath17 and death rate @xmath2 since such birth - death processes are null - recurrent if @xmath277 it follows that _ the @xmath269 model is null - recurrent if @xmath266 _ consider the original process with @xmath278 and finite @xmath128 suppose the process is in a state with a very large one club which includes all or nearly all the peers ; let @xmath272 be the number of peers in the one club . new young peers arrive at rate @xmath17 , and they are in the system for approximately @xmath279 time units while they are holding exactly @xmath273 pieces for @xmath280 . thus , over the short term , the mean number of young peers in the system holding @xmath273 pieces is near @xmath281 for @xmath280 the average fraction of peers that are young peers holding @xmath273 pieces is thus approximately @xmath282 for @xmath280 the average total rate that young peers holding @xmath273 pieces become infected is dominated by the rate the fixed seed downloads piece one to them and is thus approximately @xmath283 where the factor @xmath284 comes the assumption of uniform random useful piece selection for downloads from the seed . a young peer that becomes infected when it has @xmath273 pieces will eventually release , on average , about @xmath285 other peers from the one club . thus , to a first order approximation , for large @xmath272 , the number of peers in the system behaves like a birth - death process with arrival rate @xmath17 and state dependent departure rate @xmath286 where @xmath287 the elementary theory of birth - death processes shows that a birth - death process with constant birth rate @xmath17 and state - dependent death rate @xmath288 is positive recurrent if @xmath289 and null - recurrent if @xmath290 this strongly suggests the following to be true : if @xmath278 , the process is positive recurrent if @xmath291 and is null recurrent if @xmath292 we also expect similar results to be true for other piece selection policies , but the value of @xmath293 would depend on the piece selection policy . suppose @xmath294 and @xmath295 are two random processes , either both discrete - time random processes , or both continuous time , cdlg random processes . then @xmath95 is _ stochastically dominated _ by @xmath63 if there is a single probability space @xmath296 , and two random processes @xmath297 and @xmath298 on @xmath296 , such that + let @xmath302 be a random process with stationary , independent increments with @xmath303 suppose the sample paths are cdlg ( i.e. right - continuous with finite left limits ) . suppose @xmath304 $ ] is finite , so there are finite constants @xmath225 and @xmath305 such that @xmath306= \mu t$ ] and @xmath307 for all @xmath92 let @xmath308 for each integer @xmath313 , let @xmath314 denote the random walk process @xmath315 let @xmath316 by kingman s moment bound for discrete time processes , @xmath317 \leq \frac{\mbox{var}(s_1^n)}{-2e[s_1^n ] } = \frac{\sigma^2}{-2\mu}\ ] ] since @xmath318 is nondecreasing in @xmath272 and converges a.s . to @xmath319 the first conclusion of the lemma follows . the second conclusion follows from the first by markov s inequality . [ cor : compoundkingman ] let @xmath320 be a compound poisson process with @xmath321 , with jump times given by a poisson process of rate @xmath322 , and jump sizes having mean @xmath323 and mean square value @xmath324 then for all @xmath325 and @xmath326 @xmath327 [ lemma.mginfty ] let @xmath231 denote the number of customers in an @xmath99 queueing system , with arrival rate @xmath17 and mean service time @xmath334 suppose that @xmath335 then for @xmath336 , @xmath337 our idea is to find another @xmath99 system whose number of customers sampled at integer times can be used to bound @xmath338 suppose we let every customer for the original process stay in the system for one extra unit time after they have been served . let @xmath339 be the number of customers in this new @xmath99 system at time @xmath69 . note that @xmath340 is also the number in an @xmath99 system , with arrival rate @xmath17 and mean service time @xmath341 by a well - known property of @xmath99 systems , for any time @xmath69 , @xmath339 is a poisson random variable . since the initial state is zero , the mean number in the system at any time @xmath69 is less than @xmath342 which is the mean number in the system in equilibrium . if @xmath343 represents a poisson random variable with mean @xmath225 , then the chernoff inequality yields @xmath344 , and taking @xmath345 yields @xmath346 for any integer @xmath347 , if @xmath348,$ ] then @xmath349 therefore , @xmath350 \ } \\ & \leq & \sum_{i=1}^\infty p\ { m^\sharp_i \geq b+\epsilon ( i-1 ) \ } \\ & \leq & \sum_{i=1}^\infty e^{\lambda(m+1 ) } 2^{-(b+\epsilon ( i-1 ) ) } \\ & = & \frac { e^{\lambda(m+1 ) } 2^{-b } } { 1 - 2^{-\epsilon}}\end{aligned}\ ] ] consider an @xmath119 queue with arrival rate @xmath50 let @xmath351 denote the number of customers served in a busy period , let @xmath171 denote the length of a busy period , and let @xmath329 denote the service time of a typical customer . let @xmath352.$ ] it @xmath353 then @xmath354=\frac{1}{1-\rho } ~~~~~~~~ e[n^2]=\frac{1+\lambda^2 \mbox{var}(x)}{(1-\rho)^3 } \label{eq.n } \\ e[l]=\frac{e[x]}{1-\rho } ~~~ e[l^2 ] = \frac{e[x^2]}{(1-\rho)^3 } ~~~~ \label{eq.l } \\ \mbox{cov}(n , l ) = \frac{\lambda e[x^2]}{(1-\rho)^3}~~~~~~~~~~~~~~~\end{aligned}\ ] ] the lemma can be proved by the well - known branching process method . let @xmath329 denote the service time of a customer starting a new busy period . let @xmath98 denote the number of arrivals while the first customer is being served . then , given @xmath355 , the conditional distribution of @xmath98 is poisson with mean @xmath356 view any customer in the busy period that arrives after the first customer , to be the offspring of the customer in the server at the time of arrival . this gives the well known representation for @xmath351 and @xmath171 : @xmath357 where @xmath358 is a sequence of independent random 2-vectors such that for each @xmath11 , @xmath359 has the same distribution as @xmath360 using wald s identity , these equations can be used to prove the lemma . [ cor.fostercompcontinuous ] _ combined foster - lyapunov stability criterion and moment bound continuous time _ ( see @xcite . ) suppose @xmath329 is a continuous - time , irreducible markov process on a countable state space @xmath361 with generator matrix @xmath362 suppose @xmath141 , @xmath363 , and @xmath364 are nonnegative functions on @xmath365 such that @xmath366 for all @xmath367 , and , for some @xmath368 , the set @xmath320 defined by @xmath369 is finite . suppose also that @xmath370 is finite for all @xmath3 . then @xmath329 is positive recurrent and , if @xmath44 denotes the equilibrium distribution , @xmath371 . kurtz , t. g. ( 1981 ) . _ approximation of population processes_. cbms - nsf regional conference series in applied mathematics , vol . * 36*. society for industrial and applied mathematics ( siam ) , philadelphia , pa . massouli , l. and vojnovi , m. ( 2005 ) . coupon replication systems . in _ sigmetrics 05 : proceedings of the 2005 acm sigmetrics international conference on measurement and modeling of computer systems_. acm , new york , ny , usa , 213 . menasch , d. s. , de arago rocha , a. a. , de souza e silva , e. , leo , r. m. m. , towsley , d. f. , and venkataramani , a. ( 2010 ) . estimating self - sustainability in peer - to - peer swarming systems .

the modern effective field theory ( eft ) analysis of the nn interaction using chiral symmetry as a constraint has a recent but prolific history @xcite ( for comprehensive reviews see e.g. ref . most theoretical setups are invariably based on a perturbative determination of the chiral potential @xcite and the subsequent solution of the scattering problem @xcite . actually , the theory encounters many problems in the low partial waves and in particular in the @xmath2waves ( see however ref . @xcite for a more optimistic view ) . indeed , there is at present an ongoing debate on how an eft program should be sensibly implemented within the nn context and so far no consensus has been achieved ( see e.g. @xcite ) . the discussion is concerned with the issue of renormalization vs. finite cutoffs , a priori ( power counting ) vs. a posteriori error estimates or the applicability of perturbation theory both on a purely short distance theory or around some non - perturbative distorted waves . at the moment , it seems fairly clear that an eft scheme with a cutoff - independent and systematic perturbative power counting for the s - matrix ( the so - called ksw counting ) fails @xcite . on the other hand , the original eft inspired scheme @xcite ( the so - called weinberg counting ) has recently been shown to produce many results which turn out to be strongly cutoff dependent @xcite and hence to be incompatible with renormalizability . thus , some acceptable compromise must be made . actually , the chosen approach between this dichotomy depends strongly on the pursued goals and it is fair to say that any choice has both advantages and disadvantages . in any case , the reason for both failures can be traced back to the nature of the long - distance chiral potentials ; while pion - exchange potentials fall off exponentially at long distances they include strong power - law singularities at short distances . those singularities become significant already at distances comparable with the smallest de broglie wavelength probed in nn scattering below pion production threshold . obviously , any development of nn interactions based on chiral dynamics will presumably require a deeper understanding and proper interpretation of the peculiarities of these highly singular chiral potentials . although singular potentials where first analyzed many years ago @xcite ( for an early review see e.g. @xcite and for a more updated view within an eft context see @xcite ) , their short distance singular character within the nn interaction has seriously been faced more recently within a renormalization context for the one - pion exchange ( ope ) potential @xcite and the two - pion exchange ( tpe ) potential @xcite . in the np scattering problem the @xmath0 channel is very special since the scattering length is unnaturally large as compared to the range of the strong interaction , @xmath3 . in fact , even at zero energy the wave function probes relatively short distance components of the chiral potential @xcite \ ] ] where @xmath4 is the zero energy wave function , fulfilling the asymptotic condition @xmath5 . most of the integrand is located in the region around @xmath6 which is in between ope and tpe ranges . moreover , the low energy theorem of ref . @xcite allows to write @xmath7 which in the extreme limit @xmath8 yields @xmath9 . numerically it is found that @xmath10 is far more dependent than @xmath11 and @xmath12 when being evaluated at lo , nlo and nnlo . ] . higher energies become even more sensitive to short distance interactions . consequently , this channel looks like an ideal place to learn about the size of the most relevant short range corrections to the nn force in the elastic scattering region . actually , in the @xmath0 channel , most eft inspired schemes yield at leading order ( lo ) ( which consists of ope plus a nonderivative counterterm ) an almost constant phase of about @xmath13 around @xmath14 , and an effective range of @xmath15 . on the other hand , most determinations from partial wave analysis @xcite and high quality potential models @xcite yield an almost vanishing phase at this center - of - mass ( cm ) momentum while the experimental effective range is about twice the ope value , @xmath16 . . ] it is quite unbelievable that such large changes can be reliably accommodated by perturbation theory starting from this lo result despite previous unsuccessful attempts treating ope and tpe perturbatively @xcite . actually , for the singlet channel case , short distance components are enhanced due to the large value of the scattering length and the weakness of the ope interaction in this channel . this is why tpe contributions have been treated with more success in a non - perturbative fashion @xcite . despite phenomenological agreement with the data , the inclusion of finite cutoffs suggests that there might be some regulator dependence in those calculations . in a series of recent papers @xcite , two of us ( m.p.v . and e.r.a . ) have proposed not only to iterate but also to renormalize to all orders the nn chiral potential within a long distance expansion . by allowing the minimal number of counterterms to yield a finite result , long - distance regulator - independent correlations are established . in practice , the potential must be computed within some power counting scheme . while the potential is used within weinberg s power counting to lo , next - to - leading order ( nlo ) , and next - to - next - to - leading order ( n2lo ) , we only allow for those counterterms which yield a finite and unique scattering amplitude . in the @xmath0 np channel a single energy and momentum independent counterterm @xmath17 is considered which is determined by adjusting the physical scattering length . in the derivation of this result the mathematical requirements of completeness and self - adjointness for the renormalized quantum mechanical problem for a local chiral potential play a decisive role . this surprising result is in contrast to the standard weinberg counting where an additional counterterm @xmath18 is included already at nlo . this @xmath18 counterterm could , in fact , be determined by fitting the experimental value of the effective range ; the physics of @xmath18 is to provide a short distance contribution to the effective range in addition to the contribution from the known long distance chiral potential . within this context it is remarkable that according to ref . @xcite , where such a short distance contribution vanishes ( or equivalently @xmath19 when the cut - off is removed ) , rather accurate values are _ predicted _ yielding @xmath20 and @xmath21 _ after renormalization_. this latter value is less than @xmath22 larger than the experimentally accepted value and it suggests that most of the effective range is saturated by n2lo tpe contributions and calls for pinning down the remaining discrepancy . this trend to convergence and agreement is also shared by higher order slope parameters in the effective range expansion without strong need of specific counterterms although there is still room for improvement . motivated by this encouraging result , one goal of the present paper is to analyze the size of the next - to - next - to - next - to - leading order ( n3lo ) corrections to the results found in @xcite . the calculations in ref . @xcite exploit explicitly the local character of the chiral potential by conducting the calculations in coordinate space which makes the analysis more transparent . many results , in particular the conditions under which a renormalized limit exists , can be established a priori analytically . moreover , the highly oscillatory character of wave functions at short distances is treated numerically using efficient adaptive - step differential equations techniques . this situation contrasts with momentum space calculations where , with the exception of the pion - less theory , there is a paucity of analytical results , and one must mostly rely on numerical methods . moreover , the existence of a renormalized limit is not obvious a priori and one may have to resort to some trial and error to search for counterterms . finally , renormalization conditions are most naturally formulated at zero energy for which the momentum space treatment may be challenging , at times . of course , besides these technical issues , there is no fundamental difference between proceeding in momentum or coordinate space , particularly _ after _ renormalization , provided the same renormalization conditions are specified , since disparate regulators stemming from either space are effectively removed . indeed , we will check agreement for the phase - shifts determined in different spaces whenever such a comparison becomes possible . this equivalence is in itself a good motivation for renormalization . however , at n3lo some unavoidable non - localities appear in the chiral long distance potential . although they could be treated in configuration space , we adopt here a momentum space treatment . this will also allow us to answer an intriguing question which was left open in the coordinate space analysis of previous works @xcite , namely , the role played by the conventional momentum - polynomial representations of the short distance interaction used in most calculations @xcite in the renormalization problem . more specifically , ref . @xcite showed that taking @xmath19 was consistent , and a regularization scheme exists where a fixed @xmath18 was irrelevant , but could not discriminate whether @xmath23 was inconsistent as far as it was readjusted to the effective range parameter for any cut - off value . , becomes irrelevant in the limit @xmath24 ( see below ) . ] in this regard , the present paper yields a definite answer making the surprising agreement of the effective range found in ref . @xcite an inevitable mathematical consequence of renormalization . the paper is organized as follows . the problem is stated in sec . [ sec : renorm - prob ] where a general overview of the renormalization problem is given both in momentum as well as coordinate space . in sec . [ sec : mom - space ] , we particularize the momentum space formulation of the scattering problem with counterterms for the @xmath0 channel within a sharp three - momentum cutoff scheme . likewise , in sec . [ sec : coor - space ] we proceed similarly in the coordinate space formulation within a boundary condition regularization with a short distance cutoff @xmath25 . in sec . [ sec : coor - mom ] , we discuss some features of both coordinate and momentum space formulations in the pion - less theory and try to connect the high momentum cutoff @xmath26 with the short distance radial cutoff @xmath25 . this allows a one - to - one mapping of counterterms in both spaces which will prove useful later on in the pion - full theory . the identification between the sharp momentum cutoff and the short distance radius found in the pion - less theory is discussed further in appendix [ sec : nyquist ] in the presence of a long distance potential in the light of the nyquist theorem . in sec . [ sec : n3lo ] , we come to the central discussion on n3lo corrections to the phase shifts when the scattering amplitude is renormalized with only one short distance counterterm . a wider perspective is achieved by further considering the role of explicit @xmath1-excitations in intermediate states and the subsequent one - counterterm renormalization of the scattering amplitude . we also discuss the role of three pion exchange as well as how the results depend on the renormalization scheme used to compute the potential based on cut - off independent counterterms . the mathematical justification for using just one counterterm is provided in sec . [ sec : weinberg ] where the standard weinberg scheme is pursued both in momentum and coordinate space at nlo and n2lo and shown to potentially have some problems . finally , in sec . [ sec : conc ] , we summarize our main points . let us define the scope and goals of the present work . the standard non - perturbative formulation of the renormalization problem starts with an effective lagrangian or hamiltonian ( see e.g. @xcite and @xcite and references therein ) , from which a certain set of irreducible feynman diagrams ( usually up to a certain order ) is calculated . these irreducible diagrams are defined to represent a potential @xmath27 . the potential is then inserted into a scattering equation where it is iterated infinitely many times or , in other words , re - summed non - perturbatively . in the cm frame , where the np kinetic energy is given by @xmath28 , with @xmath29 , the scattering process is governed by the lippmann - schwinger equation @xmath30 with @xmath27 the potential operator and @xmath31 the resolvent of the free hamiltonian . the outgoing boundary condition corresponds to @xmath32 . using the normalization @xmath33 one has @xmath34 here @xmath26 means a generic regulator and represents the scale below which all physical effects are taken into account _ explicitly_. the degrees of freedom which are above @xmath26 are taken into account _ implicitly _ by including a suitable cutoff dependence in the potential . the precise equation governing this cutoff dependence was described and studied in some detail in ref . @xcite with particular emphasis on infrared fixed points . we will analyze the cutoff dependence below focusing on the ultraviolet aspects of the interaction . motivated by the low energy nature of the effective theory , the potential is usually separated into short and long distance components in an _ additive _ form @xmath35 where the long distance contribution is usually given by successive pion exchanges @xmath36 and the short distance component is characterized by a power series expansion in momentum @xmath37 where for simplicity we assume a spin singlet channel . note that @xmath17 and @xmath18 contribute to s - waves while @xmath38 contributes to p - waves , and so on . one should face the fact that , although the decomposition given by eq . ( [ eq : long - short ] ) is perturbatively motivated and seems quite natural , the additivity between short and long range forces is actually an _ assumption _ which has important consequences , as we will see . ) does not foresee for instance terms of the form @xmath39 , i.e. terms which are not polynomial but influence the renormalization process . of course , once additivity is relaxed there are many possible representations , in particular for the short distance components . ] in principle , for a given regularization scheme characterized by a cutoff @xmath26 , the counterterms @xmath17 , @xmath38 , @xmath18 and so on are determined by fixing some observables . one naturally expects that the number of renormalization conditions coincides with the number of counterterms in a way that all renormalization conditions are fully uncorrelated . the statement of uv - renormalizability is that such a procedure becomes always possible when the cutoff @xmath26 is removed by taking the limit @xmath24 . this may not be the case as there may appear redundant contributions ( see the discussion below in sec . [ sec : coor - mom ] ) meaning that one counterterm or counterterm combination can take any value . another possible situation is just the opposite ; one may want to impose more renormalization conditions than possible . in this case some counterterms or counterterm combinations are forbidden . rather than being intricate mathematical pastimes , these features have already been investigated recently @xcite for large cutoffs casting some doubt on the regulator independence of the original proposal @xcite . even if one admits generically eq . ( [ eq : long - short ] ) as well as eq . ( [ eq : long ] ) and eq . ( [ eq : short ] ) , it is not obvious how many terms should be considered and whether there is a clear way of defining a convergence criterium or identifying a convergence pattern . it is fairly clear that , on physical grounds , one should consider an expansion of the potential that starts at long distance and decreases in range as the number of exchanged particles increases . note , however , that , while ope is well - defined , the general form of the tpe potential is not uniquely determined . within this context , one of the main attractive features of the eft approach has been the definition of a power counting scheme which provides a hierarchy and an _ a priori _ correlation between long and short range physics . in the present paper , we will assume weinberg power counting @xcite ( see below for a more precise definition ) , where the long distance potential is determined in a dimensional power expansion . the long distance component @xmath40 is obtained by particle exchanges and , in some simple cases , depends only on the momentum transfer channel . ] . in such a case , if we formally take a fourier transformation of the long distance potential @xmath41 and take the limit @xmath42 one has the standard schrdinger equation in coordinate space , @xmath43 where the coordinate space potential is @xmath44 + \dots \ , \ , . \label{eq : pot - coor}\end{aligned}\ ] ] the whole discussion , which has been carried on for years now , concerns the precise meaning of these delta and derivatives of delta interactions , particularly when a long distance potential is added to the short distance one . a crucial finding of the present paper based on a direct analysis in momentum space is that _ non - perturbative renormalization imposes restrictions _ on the number of terms and form of the short distance potential which depend also on the particular long distance potential . some of these restrictions were discussed in previous works @xcite . remarkably these new renormalizability restrictions were conjectured in coordinate space in ref . @xcite for the @xmath0 channel based on self - adjointness and completeness of states and apply to the tpe potential ; a single @xmath45 counterterm is allowed while two counterterms , @xmath45 and @xmath46 , are forbidden . in the @xmath0 channel the scattering process is governed by the lippmann - schwinger equation @xmath47 where @xmath48 and @xmath49 are the scattering amplitude and the potential matrix elements , respectively , between off - shell momentum states @xmath50 and @xmath51 in that channel and the sharp three - momentum cutoff @xmath26 represents the scale below which all physical effects are taken into account _ explicitly_. from the on - shell scattering amplitude the phase shift can be readily obtained @xmath52 the short range character of the nuclear force implies that at low energies one has the effective range expansion ( ere ) @xmath53 \nonumber \\ & = & -\frac1{\alpha_0 } + \frac12 r_0 p^2 + v_2 p^4 + v_3 p^6 + \dots \label{eq : ere}\end{aligned}\ ] ] where @xmath54 is the scattering length , @xmath55 the effective range and @xmath56 , @xmath57 etc . are slope parameters . in the @xmath0 channel , the potential is decomposed as the sum of short and long range pieces @xmath58 in the standard weinberg counting , the short distance contribution is written as follows @xmath59 where the counting is related to the order of the momentum which appears explicitly . the long distance component of the potential is taken to be the sum of explicit pion exchanges @xmath60 where @xcite @xmath61 using dimensional power counting . ideally , one should determine the physically relevant long range regulator independent correlations , i.e. , long distance effects of similar range . this would amount to consider _ all _ @xmath62 exchange effects on the same footing , since they yield a long distance suppression @xmath63 modulo power corrections . at present , the only way how these long distance potentials can be systematically computed is by dimensional power counting in perturbation theory , as represented schematically in eqs . ( [ eq : potpi ] ) and ( [ eq:1pi2pi3pi ] ) . in the standard weinberg counting one has @xmath64 note that this counting involves both unknown short - distance physics and chiral long - distance physics in a _ uncorrelated way_. note also that there is no first order contribution and that there is no third order contribution to the short distance potential . regarding eq . ( [ eq : long - short-1s0 ] ) one should stress that the separation between long and short range contributions to the potential is not unique . in fact , there is a polynomial ambiguity in the long range part which can freely be transferred to the short distance contribution . however , the non - polynomial part is unambiguous as it is directly related to the left hand cut of the partial wave amplitude which for @xmath65 exchange is located at @xmath66 but presumably becomes incomplete for @xmath67 . for any non - vanishing radius ( see e.g. the discussion in ref . @xcite . ) ] in coordinate space , the problem in the @xmath0 channel is formulated as follows @xcite . assuming a _ local _ long distance potential @xmath68 one has to solve the schrdinger equation @xmath69 where @xmath70 is the reduced potential ( in fact , the fourier transformation of @xmath71 ) and @xmath72 the reduced wave function for an s - wave state . here @xmath25 is the short distance cutoff and the reduced wave function is subject to the boundary condition at @xmath73 and the standard long distance free particle behaviour @xmath74 where @xmath75 is the short distance phase - shift encoding the physics for @xmath76 . in the case of a vanishing long range potential @xmath77 the phase shift is given by @xmath78 . on the other hand , if we take @xmath79 we get a standard problem with a hard core boundary condition , @xmath80 which for @xmath81 becomes the standard regular solution at the origin . at low energies both the _ full _ phase - shift @xmath82 and the _ short distance _ phase - shift @xmath83 can be described by some low energy approximation , like e.g. , an effective range expansion , @xmath84 where @xmath85 is the short range scattering length , @xmath86 the short range effective range , and @xmath54 and @xmath55 the full ones for a further discussion on this . ] . if we also make an expansion at low energies of the reduced wave function @xmath87 we get the hierarchy of equations @xmath88 and @xmath89 and so on . the standard way to proceed would be to integrate the equations for @xmath90 , @xmath91 , etc . from infinity downwards , with a known value of @xmath92 , using eq . ( [ eq : u0 ] ) to obtain @xmath93 and then one can use eq.([eq : sch_k ] ) together with eq . ( [ eq : bc_sch ] ) and eq . ( [ eq : ere_short ] ) to compute @xmath94 for any energy with a given truncated boundary condition . this procedure provides by construction the low energy parameters we started with and takes into account that the long range potential determines the form of the wave function at long distances . the only parameter in the procedure is the short distance radius @xmath25 , which is eventually removed by taking the limit @xmath95 . one should mention at this point that the coordinate space is particularly suited for the case of local long distance potentials , but the renormalization with an arbitrary number of countertersm requires an _ energy dependent _ but real boundary condition at short distances which eventually violates self - adjointness . on the other hand , the momentum space formulation allows the discussion of nonlocal long distance potentials and the renormalization is done in terms of a _ momentum dependent _ short distance polynomial potential . although this looks like a self - adjoint problem , we will see that in this formulation the counterterms may in fact become complex . the renormalization of the pion - less theory , i.e. , a set of pure contact interactions , has been treated with great detail in the literature @xcite although without much consideration on how this problem might be embedded into the wider and certainly more realistic situation where the finite range and short distance singular chiral nn potentials are present . in fact , much of the understanding of non - perturbative renormalization within the modern nn context has been tailored after those and further studies based on the non - singular ope singlet @xmath0 potential @xcite plus the standard perturbative experience . in previous @xcite and in the present work , we pursue exactly the opposite goal : we will only consider renormalization procedures which can directly be implemented in the _ presence _ of long distance potentials since , after all , contact nn interactions are always assumed to approximate truly finite range interactions in the long wavelength limit . thus , it is useful to review here those developments with an eye on the new ingredients which appear in the non - perturbative renormalization of singular pion exchange potentials as analyzed in later sections . in addition , the deduced running of the counterterms in the contact theory in the infrared domain serves as a useful starting point when the long distance pion exchange potential is switched on . finally , we will also discuss the size of finite cutoff corrections to the renormalized result depending both on the particular regularization as well as the corresponding representation of the short distance physics . although the previously described momentum space framework has extensively been used in the past to describe successfully the data @xcite with a _ finite cutoff _ @xmath96 it is worth emphasizing some puzzling features regarding the off - shell ambiguities of the short distance potential when finite range corrections , encoded in the @xmath18 , @xmath97 etc . counterterms , are included . in momentum space , the pion - less theory corresponds to taking @xmath98 . in such a case the lippmann - schwinger equation reduces to a simple algebraic equation @xcite . at very small values of the cutoff @xmath99 , the long range part of the potential may be neglected since they scale with powers of momentum and a simple contact theory of the form of eq . ( [ eq : vs(k , k ) ] ) may be used . for instance , when @xmath100 , the lippmann - schwinger equation ( lse ) may be directly solved . using the basic integral corresponds to making the replacement @xmath101 . in dimensional regularization ( minimal subtraction scheme ) the integral is just the unitarity piece , @xmath102 . ] @xmath103 for a sharp momentum cutoff @xmath26 and @xmath104 , one obtains for the phase shifts @xmath105 at zero energy , @xmath106 and , thus , the running of @xmath17 is given by @xmath107 in this case , the phase shift is given by @xmath108 which corresponds to an ere with @xmath109 in the limit @xmath24 . note that finite cutoff corrections scale as @xmath110 . this indicates a relatively slow convergence towards the infinite cutoff limit and hence that finite cutoff effects are quantitatively important and might even become a parameter of the theory . actually , one might determine @xmath26 by fixing the effective range from the first of eq . ( [ eq : pcotdmom0 ] ) , @xmath111 , yielding the accurate numerical value @xmath112 . in this case , this particular three - momentum regularization method becomes itself a model , since we have no control on the remainder . in any case , it is straightforward to check that for any finite cutoff there is no off - shellness : @xmath113 . the running given by eq . ( [ eq : c0-short ] ) must be used for _ any _ cutoff @xmath26 if we want to renormalize in the end . however , thinking of the more general case where finite range corrections are relevant such a running is only reliable for very small cutoffs @xmath114 . if we consider also a @xmath115 coefficient in the potential , the corresponding lse can be solved with the ansatz @xmath116 which yields a set of three linear equations for @xmath117 , @xmath118 and @xmath119 . after some algebraic manipulation , the final result for the phase shift can then be written in the form @xmath120 matching at low energies to the ere , eq . ( [ eq : ere ] ) , we get the running of @xmath17 and @xmath18 @xmath121 the first equation allows to eliminate uniquely @xmath17 in favour of @xmath54 and @xmath18 , but as we see there are two branches for the solutions . however , we choose the branch for which @xmath18 decouples in the infrared domain , i.e. fulfills @xmath122 for @xmath123 . in fact , at small cutoffs , one gets for this branch @xmath124 the factor @xmath125 appearing in the small cutoff expansion for @xmath17 differs already from the coefficient in the case @xmath19 . eliminating @xmath17 and @xmath18 in favour of @xmath54 and @xmath55 , the phase shift becomes @xmath126 note that the finite cutoff corrections are , after fixing @xmath55 , again @xmath127 . so , fixing more low energy constants in the contact theory does not necessarily imply a stronger short distance insensitivity , as one might have naturally expected . in other words , the inclusion of a higher dimensional operator such as @xmath18 does not improve the ultraviolet limit , at least in the polynomial representation given by eq . ( [ eq : vs(k , k ) ] ) . in sec . [ sec : n3lo ] we will show , however , that with just one counterterm @xmath17 the inclusion of pion exchange long distance contributions generates a much faster convergence towards the renormalized limit as anticipated in refs . @xcite ( see also ref . @xcite for a quantitative estimate ) . in sec . [ sec : weinberg ] we will also show that when a @xmath128 counterterm is added this scaling behaviour is not only broken but also the phase shift fails to converge in the limit @xmath129 . thus , we see that one can establish a one - to - one mapping between the counterterms @xmath17 , @xmath18 and the threshold parameters @xmath54 and @xmath55 . nevertheless , this is done at the expense of operator mixing , i.e. , both @xmath17 and @xmath18 are intertwined to determine both the scattering length and the effective range . in other words the cutoff dependence of @xmath130 is different depending on the presence of @xmath18 . as we have seen this is not a problem since for small cutoffs we expect the running of @xmath17 to be fully independent of @xmath18 and hence on @xmath55 . however , unlike the one counterterm case , where @xmath19 , the solutions of eq . ( [ eq : c0-c2 ] ) may become complex when @xmath131 for the physical @xmath0 threshold parameters this happens already for @xmath132 ( the other two roots are complex ) . above this critical value the potential violates self - adjointness . for @xmath133 one has @xmath134 . thus , the cutoff can only be fully removed with a self - adjoint short distance potential if @xmath135 . this is consistent with the violation of the wigner causality condition reported in @xcite . note that the violation of self - adjointness is very peculiar since once @xmath17 and @xmath18 have been eliminated the phase - shift ( [ eq : ps - lambda ] ) remains real between off - shell momentum states , is violated , since the schwartz s reflection principle fails @xmath136^\dagger \neq t ( e-{\rm i } 0^+ ) $ ] . this would also have far reaching consequences for the three body problem , since three body unitarity rests on two - body off - shell unitarity and self - adjointness of three body forces . ] . one feature in the theory with two counterterms @xmath17 and @xmath18 is that the off - shell @xmath137matrix becomes on - shell only in the infinite cutoff limit , @xmath138 this is unlike the theory with one counterterm @xmath17 where there is no off - shellness at any cutoff . thus , finite cutoff effects are also a measure of the off - shellness in this particular problem . this will have important consequences in sec . [ sec : weinberg ] when attempting to extend the theory with two counterterms in the presence of the long distance pion exchange potentials since the off - shellness of the short distance contribution of the potential becomes an issue in the limit @xmath24 . the situation changes qualitatively when the fourth order corrections depending on two counterterms @xmath139 and @xmath140 are considered . obviously , we can not fix both @xmath139 and @xmath140 _ simultaneously _ by fixing the slope parameter @xmath56 of the effective range expansion , eq . ( [ eq : ere ] ) . clearly , one expects some parameter redundancy between @xmath139 and @xmath140 or else an inconsistency would arise since a sixth order parameter in the effective range expansion @xmath57 should be fixed . the situation worsens if higher orders in the momentum expansion are considered due to a rapid proliferation of counterterms while there is only one more threshold parameter for each additional order in the expansion . this required parameter redundancy is actually a necessary condition for consistency which is manifestly fulfilled within dimensional regularization but not in the three - momentum cutoff method . moreover , it was realized some time ago @xcite that the finite cutoff regularization and dimensional regularizarion in the minimal subtraction scheme yielded _ different _ renormalized amplitudes for a _ truncated _ potential . this non - uniqueness in the result due to a different regularization happens when a non - vanishing @xmath18 counterterm is considered . in any case , the dimensional regularization scheme has never been extended to include the long range part of the tpe potential which usually appear in the present nn context . thus , for the momentum space cutoffs which have been implemented in practice the short distance representation is somewhat inconsistent at least for a finite value of the cutoff @xmath26 . alternatively , one may choose an energy dependent representation of the short distance physics as @xmath141 in this case the correspondence between counterterms and threshold parameters @xmath54,@xmath55 , @xmath56 , etc . is exactly one - to - one , and the parameter redundancy is manifest , since the on shell @xmath142-matrix depends only on the on - shell potential . actually , under dimensional regularization the representations of the potential eq . ( [ eq : vs(k , k ) ] ) and eq . ( [ eq : vs(p ) ] ) yield the same scattering amplitude . although this on - shell equivalence is certainly desirable it is also unnatural , if the long distance potential is energy independent . we will nevertheless analyze such a situation in the next subsection in coordinate space . the previous discussion highlights the kind of undesirable but inherent off - shell ambiguities which arise when finite range corrections are included in the short distance potential is considered . it vanishes on the mass shell @xmath143 but nonetheless generates non trivial on shell scattering for the three - momentum cutoff . ] . in our view these are unphysical ambiguities which have nothing to do with the unambiguous off - shell dependence of the long distance potential . of course , one way to get rid of the ambiguities is to take the limit @xmath129 which corresponds to the case where a truly zero range theory is approached . however , even for a finite cutoff there is a case where one is free from the ambiguities , namely when the short distance potential is _ both _ energy and momentum independent for s - wave scattering @xmath144 the key point is that we allow only this counterterm to be cutoff dependent and real , as required by self - adjointness . of course , the discussion above for the contact theory suggests the benefits of using just one @xmath17 counterterm but does not exactly provide a proof that one _ must _ take further counterterms such as @xmath18 to zero . the extension of this analysis to the case of singular chiral potentials in sec . [ sec : weinberg ] will yield the definite conclusion that renormalizability is indeed equivalent to take @xmath19 . the previous renormalization scheme is the momentum space version corresponding to the coordinate space renormalization adopted in a previous work by two of us ( mpv and era ) @xcite . actually , in the pure contact theory , we can relate the renormalization constant with the momentum space wave function explicitly . at large values of the short distance cutoff @xmath25 , the zero energy wave function reads , @xmath145 thus , the following relation holds @xmath146 comparing with eq . ( [ eq : c0-short ] ) , we get @xmath147 where the momentum cutoff @xmath26 and the short distance cutoff @xmath25 are related by the equation @xmath148 which is nothing but an uncertainty principle relation between cutoffs . ] . note that for the standard regular solution @xmath149 one has a vanishing counterterm @xmath150 . in contrast , @xmath151 for the irregular solution . in the case of the singular attractive potentials the solution is regular but highly oscillatory and the @xmath130 takes all possible values for @xmath152 . a more detailed discussion on these issues can be seen in refs . of course , strictly speaking both eq . ( [ eq : c0-equiv ] ) and eq . ( [ eq : lambda - rc ] ) are based on a zero energy state , and in the finite energy case we will assume these relations having the limit @xmath152 or @xmath42 in mind . let us now deal with finite energy scattering states . since there is no potential , @xmath153 , for @xmath154 we have the free wave solution @xmath155 in the theory with one counterterm , we fix the scattering length @xmath54 by using the zero energy wave function and matching at @xmath73 so we get @xmath156 yielding @xmath157 and thus @xmath158 this is in qualitative agreement with the momentum space result when the cutoff is being removed , eq . ( [ eq : pcotdmom0 ] ) and , as we can see , the approach to the renormalized value is similar if the identification @xmath159 is made . note further that since the boundary condition is energy independent the problem is self - adjoint and hence orthogonality between different energy states is guaranteed . the theory with two counterterms where both @xmath54 and @xmath55 are fixed to their experimental values opens up a new possibility , already envisaged in ref . @xcite , related to the non - uniqueness of the result both for a finite cutoff as well as for the renormalized phase - shift . as pointed out above , this non - uniqueness was noted first in momentum space ref . @xcite when using a finite three dimensional cutoff or dimensional regularizarion ( minimal subtraction ) . remarkably , within the boundary condition regularization we will be able to identify both cases as different short distance representations . actually , when fixing @xmath54 and @xmath55 we are led to @xmath160 where @xmath161 and @xmath162 are defined in sec . [ sec : coor - space ] . note that now self - adjointness is violated from the beginning due to the energy dependence of the boudary condition . within the second order approximation in the energy the neglected terms are @xmath163 , so _ any _ representation compatible to this order might in principle be considered as equally suitable . the close similarity to a pad approximant suggests to compare the following three possibilities for illustration purposes @xmath164 \ , , \nonumber \\ d_p^{\rm iii } & = & \frac{u_0 ' ( r_c)^2}{u_0 ( r_c ) u_0'(r_c ) + p^2 \left [ u_2 ( r_c)u_0'(r_c ) - u_0 ( r_c ) u_2 ' ( r_c ) \right ] } \ , , \nonumber \\ \label{eq : d_p's}\end{aligned}\ ] ] and study what happens as the cutoff is removed , @xmath152 . note that all three cases possess _ by construction _ the same scattering length @xmath54 and effective range @xmath55 and no potential for @xmath165 . straightforward calculation yields @xmath166 as we see , all three representations provide the _ same _ threshold parameters , but _ do not _ yield identical renormalized amplitude for finite energy . actually , cases i and ii coincide with the three - dimensional cutoff regularization method ( see sec . [ sec : mom - pion - less ] ) , whereas case iii corresponds to dimensional regularization ( ms ) . moreover , the finite cutoff corrections to the renormalized result are , generally , @xmath167 while the rational representation yields corrections @xmath168 . these observations survive at higher orders when @xmath56 , @xmath57 , etc . threshold parameters are further taken into account . this indicates that not all short distance representations are equally `` soft '' in the uv - cutoff . the generalization of these results to the case of singular tpe chiral potentials was studied in ref . @xcite and will be also re - analyzed in sec . [ sec : weinberg ] while discussing the consistency of the standard weinberg s power counting . in any case , when finite range corrections are considered within the boundary condition regularization , there are two possible renormalized solutions depending on the particular parameterization of short distance physics . a nice feature of this regularization is that they can be identified with similar results found already in the momentum space analysis of ref . @xcite when confronting three - momentum cut - off and dimensional regularization . we note also here that no ambiguity arises when the boundary condition is assumed to be energy independent , in which case self - adjointness is guaranteed . the study of the contact theory in sec . [ sec : coor - mom ] provides suggestive arguments why it is highly desirable to carry out a regularization with a single counterterm in the @xmath0 channel by adjusting it to the physical scattering length for any cutoff value . in this section we want to extend that study when the long distance chiral potential organized according to weinberg power counting enters the game and the cutoff is removed . by taking the cutoff to infinity , we are actually assuming that all degrees of freedom not included in the present calculation become infinitely heavy . this way we expect to learn about missing physics in a _ model _ and _ regularization _ independent fashion . the traditional strategy of adjusting an increasing number of counterterms may obscure the analysis . in other words , by using this minimal number of counterterms , we try not to mock up what might be still missing in the long range description . in this context there is of course the question of convergence or cutoff insensitivity of the phase shift , when @xmath42 , provided we keep at any rate the scattering length @xmath54 to its physical value by suitably adjusting the unique counterterm @xmath169 . in coordinate space this is a rather trivial matter if the long distance potential is local @xcite , as it happens in the lo , nlo , and n2lo weinberg counting . the analysis in momentum space involves detailed large momenta behaviour of the lippmann - schwinger equation and one must resort to trial and error . indeed , the n3lo case analyzed below includes nonlocalities and it _ turns out _ to provide convergent results . for numerical calculations , we take the values for the @xmath170 and @xmath171 parameters appearing in the pion exchange potential @xmath172 used in ref . @xcite , which do a good job for peripheral waves , where re - scattering effects are suppressed and where one is , thus , rather insensitive to cutoff effects . we will only consider tpe contributions to the n3lo potential . to determine the running of the counterterm we start from low cutoffs @xmath173 since the long range part of the potential is suppressed and eq . ( [ eq : c0-short ] ) may be used . actually , the analytical result is well reproduced by the numerical method used to solve the lippmann - schwinger equation in the pure contact theory with @xmath17 . once this identification has been done , the value of the cutoff is increased steadily so that the scattering length is always fixed to the experimental value , @xmath174 . this adiabatic switching on of the long range physics guarantees that we are always sitting on the correct branch which smoothly goes into the contact theory at low cutoffs fitting @xmath54 . they are physically unacceptable unless they behave as @xmath175 for @xmath176 since they do not evolve into the theory where the long range components are decoupled . ] . the running of the counterterm @xmath177 is depicted for the lo , nlo , n2lo , and n3lo potentials in fig . [ fig : c0 - 1s0.eps ] in the low cutoff region . as expected , the deviations from the simple result of the pure contact theory , eq . ( [ eq : c0-short ] ) , start at @xmath178 for lo because of the @xmath179 exchange potential . for higher cutoffs the nlo , n2lo , and n3lo counterterm displays a cycle structure very similar to what has been observed in coordinate space @xcite . the convergence of the @xmath180 phase shift for fixed values of the lab energy , @xmath181 is displayed in fig . [ fig:1s0-cutoff.eps ] . of course , one observes a faster convergence for small energies . for the maximal value of @xmath182 , cutoff values @xmath183 are needed to change the phase shift by less than @xmath184 . in agreement with the analytical estimates of ref . @xcite , the convergence of the regulated phase - shifts towards their renormalized values follows a computable power like pattern , @xmath185 . the more singular the potential at large momenta the faster the convergence . thus , the expected increased insensitivity at short distances is indeed confirmed . finally , the renormalized @xmath0 phase shift is presented in fig . [ fig : ps1s0.eps ] for lo , nlo , n2lo , and n3lo . as a check of the present calculation in momentum space , let us mention that we reproduce the coordinate space renormalized phase shifts of @xcite at lo , nlo , and n2lo . for instance , at the maximal cm momentum of @xmath186 , the maximal discrepancy between the coordinate space and momentum space phase shifts is less than half a degree when @xmath187 and @xmath188 , respectively . the clear converging pattern can be observed all over the elastic scattering region , actually at @xmath189 , one has @xmath190 , @xmath191 , @xmath192 , and @xmath193 . however , there is still a discrepancy with the nijmegen pwa result which at this energy is @xmath194 for np scattering . to have an idea on the uncertainty of the calculated phase shift , we vary the scattering length @xmath195 , the value of , @xmath196 and @xmath197 . actually the error in the latter is correlated through the goldberger - treiman relation , so we will quote both as an error in @xmath198 only . in addition , for the chiral constants we take the central values used in @xcite which provided a good description for the peripheral waves and for the uncertainties we assume as an educated guess those from the @xmath199 study @xcite . the only exception is @xmath200 , for which the error from @xmath199 @xcite is much smaller than the systematic discrepancy with the nn determination @xcite . so it is more realistic to take the systematic discrepancy as the error . thus we take @xmath201 , @xmath202 , @xmath203 , @xmath204 , @xmath205,@xmath206 , @xmath207 . the results for the particular variations are presented in table [ tab : table1 ] and as we clearly see , the uncertainty in @xmath208 dominates the total error . at @xmath189 , one has @xmath209 , @xmath210 , @xmath211 , @xmath212 . as we see the statistical uncertainty stemming from the input parameters increases with the order . if we take the difference @xmath213 as an estimate of the systematic error , adding in quadrature statistical and systematic errors we have , @xmath214 , still a smaller quantity than the discrepancy to the nijmegen phase shift . this may suggest that _ after renormalization _ there are some physical effects missing even at the n3lo level beyond pure tpe . [ cols="^,^,^,^,^,^,^,^,^,^,^,^",options="header " , ] in the previous section we have seen that all tpe effects included to n3lo display a convergent pattern after renormalization , but there is still some missing physics . note that while at lo and nlo the only parameters are @xmath198 , @xmath215 , @xmath216 and @xmath217 , at n2lo and n3lo there appear new low energy constants ( the @xmath170 and @xmath171 , respectively ) which can be related to @xmath199 scattering and encode short range physics not considered explicitly . the @xmath1 resonance is an outstanding feature of @xmath218 scattering and explains a great deal of the low energy constants . thus , it is interesting to analyze the role of explicit @xmath1 excitations as intermediate states in the @xmath219 potential . the importance of explicit @xmath1 degrees of freedom has been emphasized on power counting grounds in several previous works with finite cutoffs where the @xmath220 splitting is regarded as a small parameter @xmath221 @xcite . the crucial role played in the renormalization problem has been stressed in ref . @xcite . in this section , we analyze the nn potential with the nlo terms of @xcite together with the @xmath222 and @xmath223 in the box diagrams as computed in ref . one advantage of such an approach is that , as compared to the standard @xmath1-less theory , there only appears the @xmath220 splitting as a parameter . the renormalization of the @xmath0 channel proceeds along the lines discussed in the @xmath1-less theory . the results for lo , nlo , nlo+@xmath222 , and nlo+@xmath224 renormalized phase shifts in the @xmath0 channel are plotted and compared to the nijmegen partial wave analysis @xcite as a function of the cm momentum in fig . [ fig : ps1s0-del.eps ] . the most striking result is the very strong resemblance between the n2lo @xmath1-less vs nlo+@xmath222 and the n3lo @xmath1-less vs nlo+@xmath224 . these results sustain the treatment of the @xmath225 splitting as a small parameter @xmath226 corresponding to the @xmath1-counting nlo@xmath227 = nlo + @xmath222 + @xmath224 . the fact that all the contributions fall off at large distances as @xmath228 suggests that despite the ability to mimic higher order corrections in the weinberg counting for the long distance potential in the @xmath0 channel there is still missing shorter range physics beyond tpe . at n3lo , three - pion exchange occurs for the first time . these contributions have been calculated by kaiser @xcite and found to be small , which is why present n3lo nn potentials omit these contributions when renormalization is not implemented . however , it should be noted that , for small distances , the @xmath229 diagrams are proportional to @xmath230 and , thus , will ultimately dominate at short distances . a rough estimate of the results published in refs . @xcite suggests that the sum of all @xmath229 graphs is attractive . thus , we infer from there that the scattering length will still be a free input parameter . on the other hand the @xmath231-exchange contribution falls off as @xmath232 at long distances so it becomes active at rather short distances , and so the effect is expected to be small because of the short distance suppression ( modulo oscillations ) of the wave function @xmath233 typical of potentials with a short distance power like singularity @xcite . this agrees with the rule that the more singular the potential the more convergent is the calculation , as we have extensively discussed above . therefore , in a complete and renormalized n3lo calculation of the @xmath0 phase shifts , we expect the @xmath229 effects not to be large , although the predictions may be closer to the empirical values than in fig . [ fig : ps1s0.eps ] . an accurate investigation of the impact of @xmath229 exchange at n3lo on nn phase shifts represents an interesting and challenging project for the future . as we have mentioned , the very definition of the potential is ambiguous as it requires fixing the polynomial terms in the momentum . in the calculations presented above , we have taken the renormalization scheme for the potential where the fixed and cut - off independent choice @xmath19 for the potential is made . we have also analyzed the situation when a different renormalization scheme is taken , namely a non - vanishing arbitrary @xmath18 coefficient which does not run with @xmath26 . however , we allow @xmath177 to run in a way that the scattering length @xmath92 is reproduced . this generates a different renormalization trajectory for @xmath234 as compared to the case @xmath19 . we find by actual calculations that this fixed @xmath18 coefficient is irrelevant , i.e. the renormalized phase shift does not depend on this fixed value in the limit @xmath129 . roughly speaking , the reason is that while the polynomial combination @xmath235 is large at large momenta , the pion exchange part behaves as @xmath236 with an additional logarithm and thus dominates for fixed @xmath18 . this irrelevance of @xmath18 was highlighted in the coordinate space analysis of ref . @xcite where the regularization based on a radial cut - off @xmath25 would provide a compact support not sensing the details of the distributional contributions in eq . ( [ eq : pot - coor ] ) , regardless on how many derivatives of delta s are included . a quite different situation arises when @xmath18 depends on the cut - off in a way that the effective range is fixed as we discuss in the next section [ sec : weinberg ] . there , it will be shown that if @xmath237 is relevant then the phase shift is not convergent . in previous sections , we have seen that the lo , nlo , n2lo , and n3lo chiral potentials can be renormalized when one counterterm @xmath169 is determined for any value of the cutoff @xmath26 by fixing the scattering length to its experimental value , @xmath238 and the limit @xmath42 is subsequently taken . this agrees with the observation of refs . @xcite that attractive singular potentials can be renormalized with a single counterterm . however , weinberg counting requires further counterterms in the short distance potential , see eq . ( [ eq : vs(k , k ) ] ) . for instance , both at nlo and n2lo a @xmath128 counterterm should be included . in this section , we discuss whether the @xmath0 scattering amplitude is renormalizable , i.e. whether the scattering amplitude has a well defined limit , when both the @xmath177 and @xmath239 counterterms are included and fixed by fitting the scattering length @xmath240 as well as the effective range @xmath241 for any value of the cutoff @xmath26 and the limit @xmath24 is pursued . actually , when the nlo ( n2lo ) long distance potential , eq . ( [ eq:1pi2pi3pi ] ) is used , this way of proceeding corresponds to renormalizing the nlo ( n2lo ) approximation of the @xmath0 channel in the standard weinberg counting . the method we follow in practice is a straightforward extension of the case with only just one counterterm @xmath17 . in the case when @xmath242 , we get a finite ( renormalized ) effective range @xmath243 , which is close but still differs significantly from the experimental value . at n2lo with @xmath19 one gets @xmath21 . this last value is so close to the experimental one that one would not expect big changes when a @xmath18 is added to exactly fit the experimental @xmath55 . thus , it makes sense to investigate what would happen if one uses a non - vanishing @xmath18 to account for the missing , and in principle tiny , contribution to the effective range . the surprissing result , to be discussed below in detail , is that trying to fit the discrepancy in the effective range with a @xmath18 counterterm is incompatible with renormalizability . in the contact theory we found that for @xmath244 the counterterms @xmath17 and @xmath18 diverge before becoming complex numbers . thus , the corresponding hamiltonian becomes non - self - adjoint , in harmony with the wigner causality bound violations unveiled in ref . when the nlo contribution to the potential is included , the critical value of the cutoff is slightly shifted towards the higher values @xmath245 ; above those values the short distance contribution to the potential becomes non - self - adjoint and causality bounds are violated . nevertheless , and similarly to the pion - less theory , the phase shifts remain real beyond this critical cutoff . in fig . [ fig:1s0-ps - c0-c2-mom.eps ] the value of the phase shift for both nlo and n2lo approximations is depicted for several cutoff values . as we see , at nlo one observes large variations when the cutoff is changed from @xmath246 to @xmath247 . in fig . [ fig:1s0-ps - c0-c2-mom.300.eps ] , we plot the value at a fixed lab energies as a function of the cutoff for both nlo and n2lo . as we see the phase shift does not seem to converge to any particular value when @xmath26 is increased ; in fact large variations can be clearly seen at moderate cutoff values , @xmath248 for @xmath249 and @xmath250 for @xmath251 . on the other hand , let us note that a plateau region not always appears , and when it does the residual contribution from pion - exchange effects is less important and the counterterms may dominate the calculation . in other words , if the cutoff was too small , one would be driven back to an effective range expansion with no visible contribution from chiral potentials whatsoever . thus , any finite cutoff calculation within a higher cutoff regime turns out to be strongly cutoff dependent , or else the cutoff must be fine tuned to intermediate energy data , hence becoming an esssential , and not an auxiliary , parameter of the theory . note that all the problems are triggered by _ insisting _ on fitting the effective range parameter to the experimental value by introducing the @xmath18 counterterm required by weinberg power counting on the short distance interaction . in contrast , if @xmath19 , as advocated in refs . @xcite and sec . [ sec : n3lo ] above , not only is the phase shift convergent in the limit @xmath42 ( in practice @xmath252 ) but also most of the effective range is saturated by the chiral potential and under the assumption of van ver waals dominance one gets at n2lo the analytical result ( see ref . @xcite for details ) , @xmath253^\frac14\ ] ] which yields @xmath254 , a surprissingly good approximation . further studies exploiting the chiral van der waals correlations can be found in ref . @xcite . ] . although we have checked that these results are numerically robust within the lse by increasing the number of grid points , there is always a reasonable doubt , since these are demanding calculations . thus , it would be nice to understand the lack of convergence from a different perspective , as we do next . let us compare the previous findings in momentum space with related investigations in coordinate space @xcite . a thorough study in coordinate space has revealed that the very existence of the @xmath42 limit may actually depend on the specific representation of the short distance physics . this analytical result has been verified by numerical calculations and will become extremely helpful in analyzing the momentum space calculations presented above . thus , for the sake of completeness it is worth reviewing the emerging pattern from ref . @xcite . as we have discussed in sec . [ sec : coor - space ] , within the boundary condition regularization the _ unknown _ short distance interaction is represented by the logarithmic derivative of the wave function at the short distance cutoff radius @xmath25 , @xmath255 . in an energy expansion of the wave function at short distances , one has @xmath256 and its logarithmic derivative for which a continuity condition is required . this introduces an energy dependence which eludes the wigner causality bound discussed in refs . @xcite , since self - adjointness is broken from the start . as we discussed previously in the contact theory described in sec . [ sec : coor - pion - less ] , at second order in the energy the neglected terms are @xmath163 and _ any _ of the three representations displayed by eq . ( [ eq : d_p s ] ) might be equally acceptable . actually , we found that renormalized amplitudes fall into two classes and that not all of them approach the renormalized limit in the same way , see eq . ( [ eq : coor - pion - renor ] ) . here , we will extend that study by inquiring what happens when the same three representations displayed by eq . ( [ eq : d_p s ] ) are used and the long distance potential @xmath257 is taken to be the nlo and n2lo for @xmath154 as the cutoff is removed , @xmath152 . note that all three cases possess _ by construction _ the same scattering length @xmath54 and effective range @xmath55 and the same long distance potential for @xmath154 . thus , any difference is clearly attributable to the different short distance representation . the results are displayed in fig . [ fig:1s0-ps - c0-c2-coor.eps ] for fixed short distance cutoff values @xmath25 as a function of the cm momentum @xmath258 . for finite values of @xmath25 we see a difference which can naturally be explained by the different off - shell behaviour of the short distance physics . as we see , the difference persists as the cutoff is being removed and in fact is magnified . as was already pointed out in a previous work @xcite , only the case @xmath259 representing a rational function turns out to yield a unique and well defined finite value for the phase shift ) . ] . this fact is clearly seen from inspection of fig . [ fig:1s0-ps - c0-c2-coor.300.eps ] where we plot the phase shifts for a fixed value of the cm momentum @xmath260 as a function of the equivalent sharp momentum cutoff @xmath261 derived for the pion - less theory in sec . [ sec : coor - mom ] . the striking similarity between the momentum space calculation presented in fig . [ fig:1s0-ps - c0-c2-mom.300.eps ] and the coordinate space calculation displayed in fig . [ fig:1s0-ps - c0-c2-coor.300.eps ] for _ finite cutoffs _ is noteworthy although not completely surprising in the light of the analysis of appendix [ sec : nyquist ] . there , the finite cutoff equivalence , @xmath262 , deduced in the contact theory is shown to hold also in the presence of a local potential . of course , all these features depend on the singular character of the interaction at short distances and do not depend on the specific form of the potential , so we expect them to hold also when @xmath1 degrees of freedom are explicitly taken into account . although we can not prove it analytically in momentum space , the remarkable similarity of the coordinate space analysis with the present momentum space calculations strongly suggests that the standard polynomial representation of the short distance potential is _ incompatible _ with renormalization . of course , this does not preclude the possible existence of a suitable potential representation of the short distance interaction , most likely with complex counterterms , such that both low energy parameters can be fixed and the cutoff can at the same time be removed . we leave such an interesting investigation for the future . in the present paper , we have analyzed the renormalization of the singlet np phase shift in the @xmath0 channel incorporating one- and two - pion exchange effects . for the long distance potential , the standard weinberg scheme based on dimensional power counting is adopted to n3lo . however , the short distance physics is parameterized in terms of one unique energy and momentum independent counterterm whose cutoff dependence is determined by adjusting the scattering length to its physical value . the present analysis is carried out in momentum space in a somewhat complementary manner as previously done in coordinate space @xcite . actually , we have analyzed and reproduced those results directly in the more popular momentum space after the cutoff has been effectively removed . in order to stress the equivalence of both approaches , we have displayed many results in parallel . this is not just a stylistic matter of presentation ; besides numerical simplicity , much understanding of the renormalization problem has been achieved by a direct analysis in coordinate space at least for local potentials . the present work provides a further example in this respect . the momentum space framework allows a direct extension to include n3lo contributions which include non - local pieces in the long distance chiral potential . the main outcome of such a calculation is that the n3lo chiral potential induces rather small corrections as compared to the n2lo results in all of the elastic scattering region . as anticipated previously @xcite , this happens to be so even when the n3lo potential is more singular than the n2lo one at short distances . moreover , the analytical scaling behaviour for large cutoffs @xmath26 of the scattering amplitude predicted in a previous coordinate space analysis @xcite is confirmed qualitatively by the present momentum space calculations ; increasing the order in the expansion of the long distance potential suppresses more the finite cutoff dependence . in summary , the scheme is convergent , but differs from the expected phase shifts obtained from partial wave analyses . an error analysis of the results shows that although there are some uncertainties induced by the input parameters , the corresponding error bands are not large enough to be compatible with the partial wave analysis of the nijmegen group . the n3lo approximation to the long distance chiral potential is computed within a heavy baryon expansion and assuming only explicit nucleon degrees of freedom in the nn potential . this implies in particular that the @xmath220 splitting is considered a non - small parameter . however , this number is about twice the pion mass , so it is not clear whether such an assumption is fully justified . therefore , we have analyzed the nn scattering problem when explicit @xmath1 intermediate state excitations are included in the potential . the net result is that , after renormalization , the @xmath222 and @xmath224 box diagrams contributions mimic extremely accurately the @xmath1-less n2lo and n3lo potentials , respectively , and thus fail to describe the higher energy region of the @xmath0 phase shift . given the fact that these contributions also fall off at large distances as @xmath263 , this result reinforces the conclusion that there is some shorter range missing physics beyond that provided by tpe . further , it is interesting to compare the present renormalized results with those obtained for a two pion exchange potential computed within a relativistic baryon framework which sums up all heavy nucleon components @xcite thus contains a full determination of tpe contributions assuming all other degrees of freedom ( including explicit @xmath1 s ) are infinitely heavy . there , all np partial waves with @xmath264 are analyzed and for the particular case of the @xmath0 channel the results are rather similar to those found here . this , again , suppports the present conclusion that , presumably , the components of the nn potential with a shorter range than tpe might finally provide the missing repulsion needed to reduce the @xmath22 overshooting of the @xmath0 effective range as well as the too low value of the phase shift in the region with cm momenta @xmath265 . we have also investigated the convergence of the standard weinberg counting for both the short distance as well as the long distance potential . at nlo and n2lo this corresponds to including a polynomial momentum dependence in the short distance interaction by means of two counterterms which may then be fixed by adjusting the scattering length and the effective range to their experimental values . we find that , with these two renormalization conditions , the counterterms turn out to be complex for not too large cutoffs ( around @xmath266 ) , signaling the breakdown of self - adjointness of the potential , and more generally suggesting a violation of wigner s causality condition . nonetheless , despite phase shifts remaining real , a unique renormalized limit does not exist . this lack of convergence agrees with similar findings in coordinate space anticipated in ref . the physical understanding of the situation is as follows . on the one hand the long range physics is fixed , thus , the off - shell propagation within the long range region is unambiguous . on the other hand , the short range potential must be adjusted to provide the threshold parameters such as the scattering length , the effective range etc . if the short range potential is just determined from fixing the scattering length only , this is at zero energy and becomes zero range when the cutoff is removed , so there is no off - shell ambiguity . in contrast , fixing further the effective range requires non - zero energy and there are in fact infinitely many ways of parameterizing this , with different off - shell behaviour even when the cutoff is removed . in other words , fixing a finite range and removing the cutoff depends on details of the method . these conflicts between off - shellness and finiteness are not new in field theory . green s functions which are renormalized on - shell do not necessarily provide finite off - shell amplitudes . the standard weinberg parameterization as a polynomial in momenta is one choice which may not turn out to be consistent with renormalizability . in this regard , it is interesting to mention that coordinate space detailed studies based on renormalization group properties @xcite suggest the existence of a suitable short distance representation yielding convergent results in theories with more than one counterterm . the generalization of those results to momentum space would require an in - depth study of the renormalization group in the presence of eventually nonlocal but singular potentials . we note that the energy dependence of the coordinate space solution violates self - adjointness explicitly and we expect that most likely an energy independent non - polynomial momentum space solution would spoil self - adjointness as well . it is important to note that with a finite cutoff and four counterterms the @xmath0 phase shift has been successfully described within the standard weinberg counting @xcite . in this regard , it is natural to question the usefulness of taking the infinite cutoff limit and to carry out a renormalization process . from a physical point of view the limit @xmath42 corresponds to consider other degrees of freedom than those not considered explicitly to be infinitely heavy . thus , the aim of the renormalization program is far more stringent than previous finite cutoff calculations . within such a framework any failure can be unambiguously attributed to missing physical information on the long distance potential , and the renormalization process highlights this in a rather vivid manner as one clearly sees when going from ope to tpe potentials . from a mathematical perspective the renormalizability requirement imposes tight constraints on the admissible forms of the short distance physics for some given long distance interactions . these are powerful conditions which have traditionally been the great strength of the renormalization ideas to avoid unnecessary proliferation of interactions . since the power counting of long distance potentials is not uniquely determined yet ( one may e.g. include or not explicit @xmath1 s ) , it would be very helpful to see what kind of short distance constraints on those potentials might be imposed as well on the basis of renormalizability or other principles . the problems with the weinberg counting for the short distance polynomial form in the momenta of the interaction with renormalization found in this paper for nlo and n2lo is to be added to the other ones noted in previous works @xcite . on the mathematical side , it is noticeable that the coordinate space analysis of ref . @xcite conjectured this result by exploiting the compelling requirement of completeness and self - adjointness for the renormalized quantum mechanical problem for a local chiral potential which at first sight may seem completely germane concepts to the eft machinery . on the phenomenological side , it should be noted that this inconsistency result does not explain why the renormalized phase shift with just one counterterm comes out reasonably close to the accepted ones , but certainly makes this unique and finite prediction more credible and inevitable from a theoretical perspective if self - adjointness is maintained , and provides further confidence on the virtues of a renormalization principle within the chiral approach to the nn interaction . one of us ( e.r.a . ) thanks diego pablo ruiz padillo ( departamento de fsica aplicada , universidad de granada ) for patient instruction on the virtues of nyquist theorem some years ago . we also thank evgeny epelpaum for a critical reading of the manuscript and daniel phillips for remarks on the paper . the work of d. r. e. has been partially funded by the ministerio de ciencia y tecnologa under contract no . fpa2004 - 05616 , the junta de castilla y len under contract no . the work of e.r.a . is supported in part by funds provided by the spanish dgi and feder funds with grant no . fis2005 - 00810 , junta de andaluca grants no . fqm225 - 05 , eu integrated infrastructure initiative hadron physics project contract no . rii3-ct-2004 - 506078 . m. p. v. has been funded by the deutsche forschungsgemeinschaft ( sfb / tr 16 ) , helmholtz association ( contract number vh - ng-222 ) . r. m. has been supported in part by the u. s. national science foundation under grant no . in this appendix we discuss further the relation between momentum and coordinate space for _ finite cutoffs _ and local potentials . we will show that solving the lippmann - schwinger equation for a long range potential with a sharp cutoff @xmath26 is equivalent to the solution of the schrdinger equation for the same potential in a discretized grid with @xmath267 . this equivalence reminds the previous identification between sharp momentum cutoff @xmath26 and the short distance cutoff @xmath25 found in the contact theory in sec . [ sec : coor - mom ] , @xmath268 . we can choose the initial grid point at @xmath269 , thus recovering the previously mentioned equivalence . actually , by invoking nyquist theorem on the cut - off ls equation , we deduce a discretized version of the variable phase equation @xcite which allows to discuss both the renormalization as well as the decimation problem of the nn - force based on chiral potentials . let us consider as a starting point the lippmann - schwinger equation for @xmath270 wave scattering , written in the form @xmath271 where @xmath272 are the matrix elements of the @xmath137 matrix between initial and final cm momentum states @xmath50 and @xmath51 respectively and the corresponding potential matrix element is given by @xmath273 for a local potential . since we are integrating the intermediate momentum up to infinity , we are implicitly assuming that @xmath49 is a regular potential . the singular potential case will be discussed later . now , if we cutoff the potential in momentum space ) . ] , we get the regularized potential @xmath274 where @xmath275 is the heaviside step function . in coordinate space the cutoff potential becomes @xmath276 which is obviously nonlocal and becomes local only when @xmath129 ( second line ) . the cutoff lse , where @xmath277 , can be solved by standard means . in the spirit of an eft , based on the idea that low energy dynamics does not depend on the details at short distances , it may perhaps be appropriate to proceed a bit differently . actually , if high momentum states are cutoff from the theory the same idea should apply to small _ resolution _ scales , i.e. regardless whether they are short or long . that means that using _ too much _ information on the potential @xmath278 even if it is exactly known point - wise may be illusory at wave - lengths longer than a given resolution , @xmath279 , is supplemented by highly oscillatory ripples at short resolution scales , @xmath280 the phase shifts should be insensitive to them at @xmath281 . ] so that one can sample @xmath278 with _ some _ resolution @xmath282 . this obviously reduces the number of coordinate mesh points in the integration . in the present context , nyquist theorem @xcite is remarkably useful because it provides an optimal way of sampling signals which have a bandwidth in fourier space , i.e. , functions for which @xmath283 then , for the original function we have @xmath284 if we define the sampling function @xmath285 of the function @xmath286 at the equidistant points @xmath287 , with the optimal @xmath288 separation , @xmath289 and compute its fourier transform we get @xmath290 note that for the optimal sampling @xmath291 is bandwidth limited to the @xmath292 $ ] interval , it can be expressed as a fourier sum @xmath293 where @xmath294 . thus it is trivial to see that @xmath295 when the sampling is done with @xmath296 . ] . inverting the fourier transform we get @xmath297}{{\pi } ( x - x_n ) } \delta x \ , .\end{aligned}\ ] ] thus , if we sample the function according to eq . ( [ eq : sampl ] ) , the following identity holds at the sampling points @xmath298 hence , there is no loss of information on the sampling points if the sampling is done equidistantly with the optimal nyquist frequency , @xmath299 . in particular , it does not really make sense to sample the function with smaller @xmath300 . in the next section we apply this sampling principle to the potential . nyquist theorem also applies for the special case of the lippmann - schwinger equation in s - wave scattering , although the derivation is different in some details to the one presented in the previous section . for clarity , we present here these details . we will consider first the general case of a non - local potential , for which the schrdinger equation in s - wave reads @xmath301 we can sample this non - local potential as @xmath302 for the momentum space representation of the sampled potential we get @xmath303 where @xmath304 . as a consequence of the cut - off @xmath26 , the potential @xmath305 can be expressed as a sum of spherical bessel functions @xmath306 where @xmath307 . then by taking @xmath267 , the sampled potential recovers the original one , i.e. @xmath308 , for @xmath309 . by fourier - transforming @xmath310 back to coordinate space , it can be checked that it reproduces the original sampling points , i.e. @xmath311 for the nyquist sampling frequency . in the case of a local potential , the one which interests us most , we sample the following way @xmath312 after double fourier - transforming , we get @xmath313 which makes the potential local for the grid points . it is in fact a finite cut - off version of @xmath314 once we notice that @xmath315 @xmath316 . for a lippmann schwinger equation with a finite cut - off @xmath26 all the matrix elements become a linear combination of separable terms . thus the ls equation becomes a linear matrix equation which can be solved by standard techniques by writing @xmath317 and defining propagator matrix elements @xmath318 therefore we get @xmath319 this equation can be reduced to a finite - dimensional @xmath320 linear algebra problem by cutting the sums to @xmath321 . the effect of this simplification can be seen by sampling @xmath322 in coordinate space ( see eqs . ( [ eq : sampl ] ) and ( [ eq : sampl - local ] ) ) @xmath323 from which we can check that cutting the sum is equivalent to introducing the ( harmless ) infrared cut - off @xmath324 . note that for local potentials , the sampled potential matrix elements at the grid points are diagonal @xmath325 thus eq . ( [ eq : ls_grid ] ) becomes @xmath326 which looks like a multiple scattering equation , with on - shell propagation between delta - shell scatterers . after matrix inversion , the on - shell solution is then given by @xmath327 in the previous section we have seen that for the lippmann - schwinger equation with a cut - off @xmath26 , we can either use the original momentum space potential @xmath305 and solve by standard means , or expand this potential , i.e. use the sampled potential @xmath310 , and solve as a linear algebra problem ( as they are both the same potential for momenta below the cut - off ) . alternatively we can directly solve the schrdinger equation for the sampling potential @xmath328 . this procedure will give an excellent approximation to the solution of the ls equation ( although not the exact solution , as explained at the end of this section ) but at a much smaller computational cost . for this purpose we make the replacement @xmath329 which is a superposition of equally spaced delta - shell potentials . it should be noted that the point @xmath330 does not contribute , as it lies at the integration boundary of the schrdinger equation . then , we have @xmath331 thus , for a potential of size @xmath332 where we do not want to describe energies higher than @xmath26 we should do with an infrared cut - off larger than the potential s length scale , @xmath333 , or equivalently @xmath334 . we can solve the schrdinger equation piecewise , @xmath335 where @xmath336 is the amplitude and @xmath337 can be understood as the accumulated phase - shift due to adding a new delta shell at @xmath338 chosen to be located at the midpoint @xmath339 ( for reasons to become clear soon ) . note that with this choice the lowest possible location of the phase - shift , @xmath340 corresponds to the point @xmath341 which we may identify with a short distance ( ultraviolet ) cutoff . this identification between the momentum - space cutoff agrees with the one obtained for the pure short range theory by solving the lse without any discretization ( see sec . [ sec : coor - mom ] ) . it should be noted that the previous method for solving the schrdinger equation does not really generate the exact solution of the ls equation with a cut - off , but only a close approximation . although @xmath328 is the best sampling function for @xmath342 , it is not the same quantum mechanical potential . the solution to the ls equation with @xmath342 can be exactly reproduced by solving the non - local schrdinger equation , eq . ( [ eq : schroedinger_nonlocal ] ) , with the potential @xmath343 which comes from inverse fourier transforming @xmath310 . the schrdinger equation should be solved with trivial initial conditions , @xmath344 , as all the physically relevant information is included in @xmath343 . on the contrary , if we solve @xmath328 , a sum of delta shells , we must include a non trivial initial condition sampling point . ] , even in the absence of any short range physics . if we solve the discretized schrdinger equation with a trivial initial boundary condition , @xmath345 , then a certain error will be included in the final solution . in the worst case we can expect an error of order @xmath279 . by means of the variable phase equation @xcite we can perform a better assessment of the error , the variable phase equation will yield @xmath346 as a curiosity , we can see that @xmath340 will scale as an inverse power of @xmath279 for a singular potential , thus signaling the need of a counterterm . ] , yielding , for example , an @xmath347 error for a yukawa potential or @xmath348 for a square well . from the previous discussion , it is apparent how to include a counterterm in the computation . it enters throught the initial sampling point of @xmath349 , which maps onto a constant term in @xmath310 counterterm when projected to the s - wave takes the form @xmath350 thus , if we write the sampled potential @xmath351 , and take into account eq . ( [ eq : v_ij_nyquist ] ) , we see that @xmath17 maps only onto @xmath352 , giving a constant contribution in momentum space , as expected . ] . thus it is ignored when solving the schrdinger equation for @xmath349 , and it enters through the initial condition @xmath340 . similar remarks can be made for singular potentials , in which a counterterm must be included in order to obtain a stable result for @xmath353 . the previous discussion can be elaborated further to reach interesting results . matching the wave functions at the points where the delta shells are located , @xmath354 , we simply get @xmath355 where @xmath356 is the reduced potential . this is a recursion relation for the phase - shift at the interval midpoint @xmath357 . unlike the matrix equation which has traditional storage limitations for large number of grid points @xmath358 , this equation does not posses this shortcoming allowing for rather large @xmath358 values . similar equations were deduced many years ago @xcite as a practical tool to attack the inverse scattering problem and to determine the nn potential on the grid points . defining the discretized effective range function , @xmath359 we get @xmath360 which can be rewritten as a continuous fraction . note that for the cutoff theory both eqs . ( [ eq : vf_dis ] ) and ( [ eq : m_dis ] ) are almost exact . the only approximation comes from the finiteness of the cutoff @xmath26 . an important property which will be used later on is the reflection property , namely the symmetry under the replacement @xmath361 which means that on the grid running the relation upwards or downwards are inverse operations of each other . obviously this property may fail in practice only due to accumulation of computer round - off errors over large evolution distances . this is also the reason why we choose the midpoint @xmath362 for the accumulated phase shift : in any other case we would lose the reflection property . the regular solution at the origin reads @xmath363 if we take the limit @xmath42 we can define @xmath364 , to get @xmath365 which is the variable phase equation @xcite up to finite grid corrections and can be interpreted as the change in the accumulated phase when a truncated potential of the parametric form @xmath366 is steadily switched on as a function of the variable @xmath367 . ( [ eq : vf_dis ] ) is thus a discretized variable phase equation , corresponding to a discretized potential sampled at the optimal nyquist frequency . this equation and its generalization to coupled channels has extensively been used to treat the renormalization problem in nn scattering in refs . @xcite . and similar equations for higher low energy parameters , @xmath370 . note that the potential only enters explicitly in the equation involving the discretized scattering length . one of the nice features of this equation is the way how it handles the case of singular points , when @xmath371 or other parameters diverge , since standard integration methods for the corresponding differential equation are based on the smoothness of the solution and hence fail . the way to proceed in practice is quite simple . for a given number of grid points @xmath358 we take the scattering length , @xmath372 , the effective range , @xmath373 and run eq . ( [ eq : alpha_dis ] ) and eq . ( [ eq : r_dis ] ) to determine @xmath374 and @xmath375 . then , we use , e.g. @xmath376 and run eq . ( [ eq : m_dis ] ) upwards to obtain @xmath377 . due to the reversibility of the algorithm one _ exactly _ has @xmath378 in the limit @xmath379 for any finite grid size @xmath380 ( modulo computer arithmetic round - off errors ) . in this way we can fix the initial boundary conditions to exactly reproduce any given scattering length , effective range , etc . this method was used successfully @xcite to extract low energy threshold parameters in all partial waves with @xmath381 from high quality potentials @xcite . on that case , we ran eq . ( [ eq : alpha_dis ] ) and eq . ( [ eq : r_dis ] ) with trivial initial boundary conditions @xmath382 , @xmath383 , etc , and obtain the potential s threshold parameters as @xmath384 , @xmath385 , etc . as we have said the discretized running parameters located at the lowest possible radius @xmath386 correspond to the short distance interactions . nowhere in the equations does the potential at the origin appears explicitly , since according to eq . ( [ eq : m_dis ] ) , one starts with @xmath387 . from this point of view the treatment of regular and singular potentials at the origin is on equal footing . the previous equations provide the accumulated phase shifts due to the addition of equidistant delta - shells sampling the original potential in coordinate space . we want to show that they actually provide a solution of the decimation problem of the lse where the low energy states are cutoff . assuming that we have the lse with a given cutoff @xmath26 related to potential , we ask how does the physical phase - shift change when we make the transformation @xmath388 , in the potential fixed . applying the process explained above based on nyquist theorem we see that this corresponds to double the grid resolution @xmath389 . obviously , by repeating the process we may effectively have @xmath390 ( being @xmath332 the range of the potential ) and hence the potential never contributes and @xmath391 . thus , for @xmath392 , the short range theory is recovered . in the opposite limit @xmath393 we get the full short range plus long range physics . the discretized representation of the lse is algebraically closed , since it corresponds to a separable interaction , but any term contains _ all _ powers in momentum . moreover , they also vanish for large @xmath50 , unlike any polynomial approximation to it . the numerical procedure explained above of computing the phase - shift from some low energy parameters like @xmath397,@xmath55 , etc . and the long distance potential @xmath278 can be also carried out by assuming that the potential at the lowest grid point , @xmath398 , is energy dependent yielding @xmath399 expanding in powers of momentum we can identify the terms @xmath400\end{aligned}\ ] ] with @xmath17 , @xmath18 , etc . note that the radius of convergence of the expansion in @xmath50 and @xmath401 is the whole complex plane , due to the meromorphic character of the spherical bessel functions .

with advancing quantum technologies @xcite it is becoming increasingly important to understand interaction of quantum systems with external degrees of freedom . evolution of a system coupled to environment can be described by a master equation . a particularly appealing type of a master equation is a lindblad equation a rather general setting that can describe any markovian evolution @xcite . while lindblad equations have been used extensively in the past to describe few - particle systems in nmr or quantum optics , recently increasing efforts are devoted to understand many - body systems in the lindblad setting . motivation comes from a wide range of fields where lindblad equations find their application , to name a few : as a computational resource in quantum information , to study e.g. transport properties of strongly - correlated condensed - matter systems , or to study nonequilibrium statistical physics of many - body systems . usually the object of most interest for open systems is a steady state , that is , a state to which any initial state converges after a long time . besides the steady state , dynamics is also of interest with one of the most important quantities being the relaxation time . in a finite system relaxation time is simply equal to the inverse gap of the liouvillian propagator generating evolution . in statistical physics the scaling power of the gap is called the dynamical exponent a critical exponent determining a universality class to which a model belongs . dynamical exponents have been extensively studied in classical exclusion models , see e.g. ref . @xcite or refs . @xcite for some more recent results . in a quantum domain much less is known about relaxation times of many - body systems . depending on a situation one might want the gap to be large or small . for instance , if dissipation is engineered in order to prepare a specific steady state one might want relaxation to be as fast as possible . on the other hand , if dissipation is unwanted , for instance , in a quantum memory device , relaxation should be as slow as possible . the value of the gap @xmath2 is important not just for the relaxation time itself @xcite , but can also carry information about the steady - state properties . namely , if the gap is finite ( so - called rapidly mixing systems ) one can show that this implies a clustering of correlations in the steady state @xcite , meaning that local observables are uncorrelated on a scale larger than @xmath3 . rapid mixing also implies the stability of steady state to local perturbations @xcite . if the gap on the other hand closes in the thermodynamic limit this can lead to a nonequilibrium phase transition @xcite and can result in a non - exponential relaxation @xcite towards a steady state . understanding how the gap scales with the system size is therefore of fundamental importance . there have been few scattered results in the literature calculating or bounding the gap either analytically or numerically for specific lindblad models . with our work we plan to extend these results , studying in more detail how the gap scales with the system size . one can distinguish grossly two different situations depending on the number of sites at which dissipation acts : in a lattice with @xmath4 sites dissipation acts on @xmath5 sites ( i.e. , on most sites ) , a setting we will call bulk dissipation , or , it can act only on a fixed number of sites ( number of sites does not grow with @xmath4 ) , a setting we will call boundary dissipation ( eventhough sites at which it acts need not be at a boundary ) . not surprisingly , as we shall see the gap can behave differently in the two cases . what is known so far is that for so - called davies generators , also called thermal reservoirs , that are an example of bulk dissipation , one is in certain cases able to rigorously prove that the gap is independent of the system length @xcite , @xmath6 . one can also show exponential relaxation in weakly coupled systems @xcite . similarly , one can show that the gap can be constant for some other systems with bulk dissipation @xcite , while in other systems it can also scale as @xmath7 @xcite . on the other hand , for open systems with boundary dissipation the observed gaps have so - far all been scaling as @xmath1 , or smaller , examples being the xy @xcite or the xxx @xcite model . for scaling of gaps in the redfield equation see ref . @xcite . in the present work we are going to study the scaling of the gap with system size in a number of spin chain models with bulk as well as with boundary dissipation . the aim is to get a better overview of possible gap scaling and changes in the scaling power as one varies parameters . such changes can be associated with possible phase transitions in the steady state as well as in decay modes . in the main part of the paper we shall organize sections according to different models studied , explaining for each different techniques used ( ranging from analytical to numerical ) to infer the scaling . we shall also pay attention to the validity of a weak - dissipation perturbation theory that can be used to calculate the gap , demonstrating its failure in a number of cases . if one is not interested in all the details of gap calculation , or just wants an overview of different gap scalings found , there are tables [ tab : bulk ] and [ tab : bound ] provided in the summary section at the end of the paper . the paper is organized as follows . in sec.ii . we briefly explain the setting of lindblad equations . in sec iii . we then study systems with bulk dissipation , while in sec . we study systems with boundary dissipation . each of these two sections is split into subsections describing different systems . in sec.iv . we also explain an argument that the gap in systems with boundary dissipation can not be larger than @xmath0 . finally , in sec.v . we conclude as well as present a summary of the results found . the lindblad equation is @xcite @xmath8+\cld(\rho ) , \label{eq : lin}\end{aligned}\ ] ] where @xmath9 is a linear superoperator called a dissipator that can be expressed in terms of traceless lindblad operators @xmath10 . denoting eigenvalues of the liouvillian @xmath11 by @xmath12 , and ordering them according to their real parts , with @xmath13 assuming to be nondegenerate , the gap is equal to a negative real part of the 2nd largest eigenvalue , @xmath14\ ] ] eigenvectors of @xmath11 corresponding to nonzero eigenvalues are called decay modes . the models that we are going to study will all be spin chains composed of @xmath4 lattice sites , each carrying a spin-@xmath15 particle . coupling in @xmath16 will always be local between nearest - neighbor sites only , and with open boundary conditions . dissipation will also be local , i.e. , each lindblad operator @xmath10 will act nontrivially only on a single site ( different @xmath10 though can act on different sites ) . two types of dissipation will be employed , both physically motivated . the first one is dephasing for which @xmath17 , and which tries to destroy off - diagonal matrix elements ( in the eigenbasis of @xmath18 ) . the second one is magnetization driving in which lindblad operators proportional to raising and lowering operators try to impose an imbalance in populations of spin - up and spin - down states . we mention that the many - body spin chain models studied here are within reach of present day cold - atom technology , e.g. refs . @xcite , with individual components like few qubit controlled dissipation @xcite or heisenberg spin chains @xcite already demonstrated . a canonical model that we shall study in both bulk- and boundary - driven cases is the anisotropic heisenberg model ( xxz model for short ) . for zero anisotropy the xxz model goes into the xx model , which is especially simple and even allows for an exact asymptotic solution . the hamiltonian of the xx model is @xmath19 dissipation @xmath20 is given by dephasing of strength @xmath21 acting independently on each site , which is described by a set of @xmath4 lindblad operators , @xmath22 we note that such liouvillian @xmath11 conserves the total magnetization @xmath23 , that is , if @xmath24 is an eigenstate of @xmath25 with eigenvalue @xmath26 and similarly @xmath27 is an eigenstate of @xmath25 with an eigenvalue @xmath28 , then @xmath29 is a superposition of terms in which all @xmath30 and all @xmath31 are again eigenstates of @xmath25 with eigenvalues @xmath26 and @xmath28 , respectively . more formally conservation means that @xmath32 , with @xmath33 being rotation around the @xmath34-axis . @xmath11 therefore has a block structure and we shall label each block by a magnetization difference @xmath34 and by a number of flipped spins @xmath35 , @xmath36 for an @xmath4-site chain the allowed values of @xmath34 are from @xmath37 to @xmath38 in steps of @xmath39 , while that of @xmath35 are @xmath40 . of special interest will be sectors with @xmath41 because they carry a steady state , i.e. , an eigenstate of @xmath11 with eigenvalue @xmath42 . a subspace with @xmath41 and some value of @xmath35 will be simply called an @xmath35-particle sector and has ( operator ) dimension @xmath43 . two most important ones are @xmath44 ( 1-particle sector ) , being the smallest nontrivial one , and the largest one with @xmath45 ( or @xmath46 for odd @xmath4 ) being called a half - filling sector because half of the spins are pointing up and half are pointing down . an important property of dissipative systems where @xmath16 is quadratic in fermionic operators ( via jordan - wigner transformation ) and lindblad operators are hermitian ( but not necessarily quadratic ) , and our xx model is an example of such a system , is that equations for correlation functions split into a hierarchy of equations according to their order , i.e. , the number of fermionic operators in the expectation value @xcite . this enables one to exactly calculate few - point expectation values in the steady state , for instance in the presence of an additional boundary driving @xcite , an incoherent bulk hopping @xcite , or special engineered dissipation @xcite . here we are not interested in the steady state but instead in the liouvillian gap , in particular in its scaling with the system size @xmath4 . due to magnetization conservation there are @xmath47 steady states , one in each invariant subspace with @xmath41 and a particular @xmath35 . one can also easily see that such a steady state is a uniform mixture off projectors to all @xmath48 diagonal basis states in the corresponding @xmath35-particle subspace , for instance , for @xmath49 in a 1-particle sector ( @xmath44 , @xmath41 ; subspace dimension @xmath50 ) the steady state is @xmath51 . for more details about such states , for instance their schmidt spectrum , see appendix [ app : a ] . subspaces with @xmath52 do not contain any zero eigenvalues ( i.e. , steady states ) because they are orthogonal to the identity operator which is , due to trace preservation , always a left eigenvector of @xmath11 with eigenvalue zero and is therefore non - orthogonal to all steady states . because we are interested in the gap we can limit our discussion to subspaces with @xmath41 as they are the only ones that contain steady states . subspaces with @xmath53 and @xmath54 are of dimension @xmath55 and contain only the steady state and are of no interest to us . all other @xmath56 subspaces contain a nontrivial gap . we shall be in particular interested in the smallest gap out of those , i.e. , the global one , giving the slowest relaxation rate in the system . because of a hierarchy of correlations we know that the eigenvalues , i.e. decay rates , of @xmath57-point correlation equations are equal to ( some ) eigenvalues of the liouvillian @xcite , however , one in general does not know whether they also give the gap . regardless of that they can be used as an upper bound on the global gap . for instance , for a model with an incoherent hopping the relaxation rate of @xmath39-point correlations scales as @xmath7 @xcite meaning that the global gap can not be larger . similarly , for a boundary - driven system numerical calculation of the global gap also resulted in @xmath7 scaling @xcite . by numerically diagonalizing our @xmath11 in each @xmath35-particle sector we observe that the gap is in fact the same in all sectors . this in particular means that to calculate the global gap it is enough to consider the @xmath55-particle sector which is of dimension @xmath58 ( or its symmetric @xmath59-particle partner that has exactly the same eigenvalues ) . this is due to a free nature of @xmath16 , causing that the spectrum of the 1-particle sector to be contained in higher-@xmath35 sectors . one could be tempted to conclude that the @xmath55-particle sector will be rather trivial this is certainly true for a @xmath55-particle sector of dimension @xmath4 in the hilbert space of states however , here we are dealing with a @xmath55-particle sector of dimension @xmath60 in the hilbert space of operators . as we shall see this allows for a rich behavior , among other things for a discontinuous transition in the scaling of the gap from a constant @xmath61 for small dephasing strength @xmath21 to @xmath7 for non - small @xmath21 . before going to the actual calculation of the gap let us pause for a moment and have a look at an alternative formulation of the eigenvalue problem for the whole @xmath11 . the liouvillian that acts on a @xmath62 dimensional operator space is a non - hermitian linear operator that can be written as a `` hamiltonian '' of a non - hermitian spin-@xmath15 ladder composed of @xmath4 rungs . each rung , spanned by @xmath63 pure states , takes care of one site of @xmath11 which is also @xmath63 dimensional ( e.g. , three pauli matrices plus an identity ) . the resulting ladder is for the xx chain with dephasing very simple : it is composed of two @xmath64-coupled chains along the ladder legs with an additional imaginary coupling along rungs , see fig . [ fig : lestev ] and appendix of ref . @xcite for the mapping details . interaction along upper and lower legs ( thin lines ) and an imaginary @xmath65 interaction due to dephasing along rungs ( double lines).,width=316 ] the steady state is an eigenvector corresponding to the eigenvalue zero , in other words a `` ground state '' of a non - hermitian ladder . because of dissipative coupling along rungs this ground state is always , regardless of the value of @xmath21 , a direct product of singlet states at each rung . the steady state is therefore always dominated by dephasing @xmath21 , forcing the steady state to be a singlet state . this is due to a very special structure of @xmath11 no interaction in @xmath16 ( which plays an inert role ) and a dephasing that kills all off - diagonal matrix elements . as we shall show , things are very different for the first decay mode determining the gap . there is a transition from a @xmath21-dominated phase to a different phase in which @xmath16 becomes important . in the ladder formulation of @xmath11 we can also see why the @xmath55-particle superoperator sector is nontrivial : it corresponds ( in an appropriate basis ) to states with one particle in the upper leg ( bra ) and one particle in the lower leg ( ket ) . therefore , a @xmath55-particle superoperator problem is like a problem of @xmath39 interacting particles on a ladder . it constitutes the simplest case of an interacting system . let us now calculate the gap in the 1-particle sector . essentially the same eigenvalue problem , apart from different boundary conditions @xcite , has been rigorously solved in ref . @xcite . here we shall present a different and approximate calculation of the gap , which is simpler but nevertheless accurate in the thermodynamic limit @xmath66 in which we are especially interested . of the xx model with dephasing @xmath67 . shown is the exact gap ( symbols ) , the full curve is the approximate formula ( [ eq : gap1mag ] ) , while the dotted curve shows the asymptotic expression @xmath68 . the approximate result ( [ eq : gap1mag ] ) is accurate for @xmath4 larger than the transition point @xmath69.,width=316 ] the basic idea is the following . we have seen that the steady state is always dominated by the dephasing and one can expect that in some range of dephasing strengths the first decay mode will also be of the same nature that is dominated by @xmath21 . if this is the case , then its eigenoperator will be close to diagonal because dephasing kills all off - diagonal elements . in the lowest approximation the unitary part @xmath70 couples diagonal elements @xmath71 , where @xmath72 denotes a state of @xmath4 spins with the @xmath73-th spin being flipped to @xmath74 while all others are in state @xmath75 , to two off - diagonal @xmath76 and @xmath77 . such a tri - diagonal approximation reduces the size of @xmath11 in the 1-particle sector from @xmath60 to @xmath78 . in addition , we can immediately see that on this three - diagonal subspace all @xmath59 states of the form @xmath79 are eigenstates with eigenvalue @xmath80 ( they are eigenstates of @xmath81 with eigenvalue @xmath42 because of different signs when @xmath16 acts from the left and from the right , while @xmath82 and similarly for @xmath83 ) . reduction of @xmath11 to @xmath84 dimensional subspace spanned by @xmath71 and @xmath85 has a block structure of form , @xmath86 basis ordering is such that the first diagonal block corresponds to @xmath4 states @xmath71 , and the second to @xmath59 states @xmath85 . writing the eigenvector as @xmath87 and the eigenvalues as @xmath88 , and eliminating @xmath89 , we get an eigenvalue equation @xmath90 , with @xmath91 being a tridiagonal matrix of size @xmath92 with matrix elements @xmath93 @xmath94 is nothing but a standard matrix appearing is a solution of harmonic oscillators or a tight - binding model , and has eigenvalues @xmath95 , leading to eigenvalues @xmath96 of @xmath97 being @xmath98 . the gap is determined by the largest eigenvalue @xmath99 , and is @xmath100 the gap @xmath2 gives the distance of the largest decay mode eigenvalue from the origin along a real axis , with the next largest decay mode being asymptotically at a distance @xmath101 . for @xmath16 with periodic boundary conditions the gap is @xmath63 times larger than the above @xmath2 ( [ eq : gap1mag ] ) . we can see that for sufficiently small @xmath21 the expression under the square root in eq.([eq : gap1mag ] ) can become negative . for large @xmath4 this happens for @xmath102 , where the critical dephasing is @xmath103 on dephasing and @xmath4 . full black line is the critical @xmath104 ( [ eq : gammac ] ) delimiting two phases . in ( b ) and ( c ) are shown expansion coefficients @xmath105 of the largest decay mode @xmath106 , @xmath107 , showing an almost diagonal dephasing - dominated phase with @xmath108 in ( b ) , and a delocalized phase with @xmath109 in ( c).,width=316 ] our approximate expression for the gap ( [ eq : gap1mag ] ) is accurate only for @xmath110 , where the decay mode is indeed governed by dephasing , see fig . [ fig : gap1mag ] . note though that in the thermodynamic limit @xmath111 and therefore eq . ( [ eq : gap1mag ] ) becomes exact for any @xmath21 . one can see already in fig . [ fig : gap1mag ] that below @xmath104 ( or @xmath112 , depending on which parameter is held fixed ) the gap becomes independent of @xmath4 and in the thermodynamic limit equal to @xmath113 ( in this regime the largest decay mode eigenvalue is complex and the gap @xmath2 is equal to its real part ) . therefore , for the largest decay mode there are two phases : for @xmath114 the decay mode is governed by @xmath16 with the gap being @xmath115 , while for @xmath110 the decay mode is governed by dephasing and the gap is @xmath116 . one can use perturbation theory to show that the gap is indeed independent of system size for sufficiently small dephasing @xcite . the convergence radius of weak - coupling perturbation series for small @xmath21 is @xmath104 ( [ eq : gammac ] ) and algebraically shrinks to zero in the thermodynamic limit . also , the thermodynamic limit @xmath66 and the weak coupling limit @xmath117 do no commute , which can be also seen in fig . [ fig : gap1mag_faze ] showing the phase diagram . for xx model with bulk dephasing , @xmath118 , @xmath119 , and 1-particle sector . for @xmath120 there are of order @xmath5 eigenvalues that are on the real axis and are separated from the bulk consisting of the remaining @xmath121 complex eigenvalues . real parts of the bulk eigenvalues are around @xmath80 with the width being @xmath122.,width=316 ] finally , let us briefly comment also on the overall spectrum of @xmath11 . numerically diagonalizing @xmath11 on the 1-particle subspace the following picture emerges , see fig . [ fig : eig1mag ] . most of the eigenvalues , @xmath121 in number , are within the bulk laying in a complex plane . as one increases @xmath4 eigenvalues `` evaporate '' from the bulk and join a bunch of real eigenvalues to the right of the bulk . the number of these separated real eigenvalues is proportional to @xmath4 . decay modes corresponding to the separated real eigenvalues are dictated by dephasing , whereas decay modes in the bulk are instead dictated by the hamiltonian . remember that in the absence of dephasing the spectrum of @xmath11 would be entirely on the imaginary axis . such distinct nature of two decay mode types is in turn reflected in two different phases of the largest decay mode , fig . [ fig : gap1mag_faze ] . indeed , if one decreases @xmath21 at fixed @xmath4 , the real eigenvalues get `` absorbed '' in the bulk , with the last one disappearing at @xmath123 , at which point a transition happens in the scaling of the gap with @xmath4 . approximate values of real eigenvalues can be obtained from the approximation with @xmath97 ( [ eq : gap1mag ] ) , resulting in a scaling of the @xmath73-th largest eigenvalue of @xmath11 as @xmath124 . in the thermodynamic limit real eigenvalues therefore cluster around the origin with their density there having a square - root singularity . it has been observed @xcite that if the gap @xmath2 closes in the thermodynamic limit there is a possibility for an algebraic relaxation instead of an exponential one occurring when @xmath2 is finite . such an algebraic decay can be explained @xcite by clustering of eigenvalues around @xmath42 . in our xx chain with dephasing this clustering is of the same kind as in boundary - driven free models studied in ref . @xcite and as a consequence , in the thermodynamic limit the relaxation will have a power - law form . in the xx model the bulk hamiltonian describes non - interacting particles . choosing the xxz hamiltonian instead , the additional coupling in the @xmath34-direction represents the interaction between fermions in the jordan - wigner picture . in this section we shall therefore consider the xxz spin chain with bulk dephasing . the hamiltonian of the xxz chain is @xmath125 with @xmath126 being an anisotropy parameter . dissipation is the same dephasing at each site as used already for the xx model ( [ eq : dephasing ] ) . magnetization is again conserved and there is one steady state in each sector with magnetization difference @xmath41 and @xmath35 particles , see discussion for the xx model with dephasing . using a jordan - wigner transformation the anisotropy part @xmath127 has a form @xmath128 and therefore represents interaction between nearest - neighbor fermions . analytical solution for eigenvalues of @xmath11 is not possible anymore and we will have to resort to various perturbation approaches and numerical calculation . numerically calculating the gap one sees that , as opposed to the xx model , the gap is this time different in different sectors . in particular , the global , i.e. , the smallest gap , is from the half - filling sector with @xmath45 ( or @xmath46 for odd @xmath4 ) and not from the 1-particle sector , @xmath44 . we shall nevertheless first discuss the 1-particle sector , where in the thermodynamic limit things are simple . because the 1-particle sector s size is @xmath60 one can numerically calculate the gap for systems of reasonable size . results are in fig . [ fig : xxz-1mag ] . in the 1-particle sector of the xxz chain with dephasing , @xmath118 . symbols show results of numerical diagonalization for anisotropies @xmath129 and @xmath130 , while the full line is @xmath131 , giving the asymptotic gap , which is equal to the one in the xx chain with dephasing ( [ eq : gap1mag]).,width=316 ] as we can see for large system sizes the gap is equal to @xmath132 and is therefore equal to the one for the xx chain with dephasing . in the 1-particle sector the gap is in the thermodynamic limit independent of interaction ( anisotropy ) @xmath126 . this can be qualitatively understood if one looks at the non - hermitian ladder formulation of @xmath11 ( for the xxz chain the only difference compared to the xx chain in fig . [ fig : lestev ] is that the hamiltonian along the two legs is the xxz hamiltonian , see ref . @xcite ) . because there are only two particles present in the ladder interaction is infrequent , it is in fact only a @xmath0 boundary effect @xcite , and is asymptotically irrelevant ( for finite @xmath4 correction to the asymptotic gap ( [ eq : xxz_1mag ] ) due to finite @xmath126 scales as @xmath133 ) . as mentioned , for the xxz chain with dephasing the smallest gap is from the so - called half - filling sector with @xmath45 particles , which is also the largest subspace of @xmath11 . here we shall study the gap in this sector . we are going to use perturbation theory for small @xmath21 as well as for large @xmath21 , while in - between numerical calculations will be used to infer a general form of @xmath2 . we are also going to show that the perturbation theory for large @xmath126 fails . let us start with small @xmath21 . for the unperturbed part of the liouvillian we take the whole unitary part , @xmath134 $ ] , generated by the xxz hamiltonian ( [ eq : xxz ] ) , while perturbation is the dephasing @xmath135 . because the unperturbed @xmath136 has a degenerate steady state subspace corresponding to eigenvalue @xmath42 , we have to use 1st order degenerate perturbation theory . the steady - state subspace is spanned by all projectors @xmath137 to eigenstates @xmath138 of @xmath16 ( we assume a non - degenerate spectrum of @xmath16 ) . let us denote a projection of @xmath139 to the steady - state subspace of @xmath136 by @xmath140 , with @xmath141 being of size @xmath48 . using the form of dephasing lindblad operators ( [ eq : dephasing ] ) we get @xmath142 the largest eigenvalue of @xmath143 is equal to @xmath42 and the next - largest one , denoted by @xmath144 ( similarly as for @xmath11 , all eigenvalues of @xmath143 have non - positive real parts ) , determines the gap of @xmath11 for small @xmath21 , @xmath145 we observe that , while on one hand dephasing can be derived as being due to a classical fluctuating magnetic field in the @xmath34-direction @xcite , we also see that the gap of @xmath11 for small dephasing is determined by eigenstate fluctuations of magnetization in the @xmath34-direction , as reflected in the form of @xmath143 ( [ eq : r ] ) . of matrix @xmath143 ( [ eq : r ] ) in the xxz chain with dephasing , determining the gap eq . ( [ eq : xxz - weak ] ) . full line is @xmath146 , suggesting the asymptotic dependence for @xmath147.,width=316 ] in fig . [ fig : xxz - weak ] we show results of exact numerical diagonalization of @xmath143 for different anisotropies and system sizes , all in the largest half - filling sector with @xmath45 . we can see that for @xmath148 the eigenvalue @xmath149 becomes asymptotically independent of @xmath4 , for instance , for @xmath150 it converges towards @xmath63 as already calculated in section about the xx model . for @xmath151 it on the other hand decays ; numerical data for available @xmath152 is consistent with a @xmath153 decay . for small @xmath154 this asymptotic decay starts for sizes larger than @xmath155 . it would be interesting if one would be able to directly calculate the largest nontrivial eigenvalue of @xmath143 via the bethe ansatz eigenstates @xmath156 . so far we have calculated the global gap for small dephasing @xmath21 without saying anything about the convergence radius , i.e. , the validity of such an approximation . to find out critical @xmath104 upto which perturbative gap ( [ eq : xxz - weak ] ) can be expected to hold , we are going to numerically calculate the gap and compare it to perturbative prediction . . insets show dependence of the critical @xmath104 that determines validity of perturbation series . left : gapless regime of @xmath129 for which @xmath104 depends on @xmath4 algebraically ( dashed curve in the inset is @xmath157 ) . right : gapped regime @xmath158 for which @xmath159 ( inset).,title="fig:",width=155 ] . insets show dependence of the critical @xmath104 that determines validity of perturbation series . left : gapless regime of @xmath129 for which @xmath104 depends on @xmath4 algebraically ( dashed curve in the inset is @xmath157 ) . right : gapped regime @xmath158 for which @xmath159 ( inset).,title="fig:",width=151 ] in addition to symmetries of @xmath11 already pointed out there is also a spatial reflection symmetry with respect to exchanging sites @xmath160 , resulting in even and odd decay modes and eigenvalues of @xmath11 . the largest nonzero eigenvalue in even and odd sector ( both for @xmath41 and half - filling @xmath45 ) is shown in fig . [ fig : xxz - pert ] for system size @xmath161 and two values of @xmath126 . the steady state a uniform combination of all pure - state projectors is always from the even sector . we can see that the largest non - zero eigenvalues cross at a critical @xmath104 . the decay mode that determines the gap in the half - filling sector ( red pluses in fig . [ fig : xxz - pert ] ) is from the even sector for @xmath102 and from the odd one for @xmath120 . for @xmath102 we see that the gap ( red pluses ) is proportional to @xmath21 ( horizontal line in fig . [ fig : xxz - pert ] ) and therefore perturbation series ( [ eq : xxz - weak ] ) holds . the convergence radius @xmath104 shrinks algebraically with @xmath4 in the gapless phase @xmath162 , at @xmath129 it decays as @xmath163 , although we can not exclude asymptotic @xmath0 scaling , while it is exponentially small in the gapped phase @xmath151 . for the xxz chain with bulk dephasing the convergence radius of perturbation series in the dephasing therefore shrinks to zero in the thermodynamic limit . we also note that the two eigenvalues shown in fig . [ fig : xxz - pert ] are both real , except the odd one for @xmath102 and @xmath129 which forms a complex pair . for non - small @xmath21 the xxz chain with bulk dephasing is diffusive @xcite and therefore one in general expects the gap to scale as @xmath7 in the thermodynamic limit . namely , for diffusive systems evolution of macroscopic observables does not depend independently on time @xmath164 and spatial coordinate @xmath106 but instead only on one scaling variable @xmath165 and therefore time should scale with a size squared . we numerically calculated @xcite the gap for fixed non - small @xmath21 and indeed confirmed the @xmath108 scaling , see fig . [ fig : xxz - gap ] as well as ref . @xcite . . full lines denote asymptotic @xmath7 scaling for all @xmath126.,width=316 ] exact dependence of @xmath2 on @xmath21 and @xmath126 is more complicated ; a crude approximation that seems to work for large @xmath4 ( at least for @xmath118 ) is @xmath166 the straight lines plotted in fig . [ fig : xxz - gap ] is in fact this expression . when some parameters in the system are large one can again use perturbation theory . one such possibility is the case of large dephasing @xmath21 , where the unperturbed liouvillian @xmath136 contains @xmath167 and dephasing dissipation @xmath20 while the perturbation @xmath139 consists of hopping given by @xmath168 . this has been done in ref . @xcite , obtaining that the gap for large @xmath21 scales as @xmath169 . we shall redo calculations of ref . @xcite , deriving also subleading terms , in order to be able to comment on the validity of different limits . first , @xmath136 has a degenerate steady - state subspace and one has to use degenerate perturbation theory . first - order perturbation on the steady - state subspace of dimension @xmath170 is always zero because all steady states are diagonal , while the hopping @xmath139 transports one spin in the bra or ket , and so at least two applications of @xmath139 are needed to again get a diagonal state and with it a nonzero matrix elements of @xmath139 . second - order perturbation of the steady state manifold is determined by , @xmath171 where @xmath172 is a projection operator to the steady state manifold . the steady state manifold of @xmath136 consists of all diagonal density matrices , @xmath173 , where @xmath174 are @xmath170 ( basis ) states in the standard @xmath18-eigenbasis . one can write matrix elements of @xmath175 as @xmath176_{j , k}:=-\tr{\ket{\psi_j}\bra{\psi_j } \cl_{\rm eff}(\ket{\psi_k}\bra{\psi_k})}$ ] , where matrix @xmath177 can in turn be written in terms of pauli spin variables , obtaining @xmath178 + \nonumber \\ & & \!\!\!\!\!\!\ ! + \frac{\delta^2/\gamma}{2(\delta^2+\gamma^2 ) } \sum_{j=1}^{l-3 } ( \sz_j\sz_{j+3}-\1)(\1-\vec{\sigma}_{j+1}\cdot \vec{\sigma}_{j+2 } ) , \label{eq : heff}\end{aligned}\ ] ] where @xmath179 . such @xmath177 is hermitian , has @xmath47 zero eigenvalues ( one for each invariant magnetization sector ) , while all other eigenvalues are positive . the smallest non - zero eigenvalue is equal to the gap @xmath2 of @xmath11 in the limit when @xmath180 . in ref . @xcite the above result ( [ eq : heff ] ) has been derived in the leading order in @xmath181 , i.e. , @xmath182 , from which due to a quadratic low - energy dispersion of @xmath183 one gets the asymptotic gap @xmath116 . this perturbative result is also valid in the thermodynamic limit @xcite because the convergence radius does not shrink to zero . with the exact expression for @xmath177 we can also explore the case of large @xmath126 , for which one would be tempted to think that perturbative series will again work because @xmath139 is again small compared to @xmath136 that contains a large parameter @xmath126 . such reasoning though is in fact wrong and one can not use perturbation in @xmath139 if only @xmath126 is large . the reason for the failure is rather instructive and we are going to explain why it happens . as opposed to the limit @xmath184 , for @xmath185 the two terms that appear in addition to @xmath183 in eq . ( [ eq : heff ] ) can not be neglected because the two prefactors in front of them scale for large @xmath126 as @xmath181 and are therefore of the same order as the @xmath183 term . there is also an additional subtlety : it would be tempting to retain only the leading order expansion of the two prefactors , the already mentioned @xmath181 and @xmath186 , however , in that case the ground state of @xmath177 is highly degenerate and the gap would therefore remain zero ( one of the reasons for the degeneracy is that the boundary term in the first line of eq . ( [ eq : heff ] ) with a @xmath181 prefactor exactly cancels the boundary terms from @xmath183 , leaving the first and the last spin uncoupled ) . one has to retain at least the 1st order expansion of the prefactors , resulting in terms that are proportional to @xmath187 . this means that the gap of @xmath177 , and with it also the liouvillian gap @xmath2 , will scale as @xmath188 for large @xmath126 ( this conclusion remains true despite the failure of 2nd order perturbation expansion ) . calculating the gap of the full @xmath177 ( [ eq : heff ] ) and comparing it with the exact gap of @xmath11 , one gets data in fig . [ fig : heff ] . as one can see the error does not decrease to zero as one increases @xmath126 ( this residual error as well increases with @xmath4 ) . the reason is that for large @xmath126 and fixed dissipation @xmath21 not all non - zero eigenvalues of @xmath136 are large some are of order @xmath21 and therefore a pseudoinverse @xmath189 in perturbation series ( [ eq : pert2 ] ) is not necessarily small . this occurs because the spectrum of @xmath190 is highly degenerate with eigenstates being product states in the eigenbasis of @xmath18 , which is also the eigenbasis of @xmath20 . if @xmath24 and @xmath27 are two such degenerate eigenstates we will have @xmath191=0 $ ] while for dephasing @xmath192 . as a consequence , consecutive orders in perturbation series do not necessarily uniformly decrease . this can be compared to perturbation series for fixed @xmath126 and large @xmath21 , for which all non - zero eigenvalues of @xmath136 are large and of order @xmath21 and therefore @xmath189 is small . one can in fact draw a general rule : if a large perturbation parameter is in global dissipation ( strong coupling ) a perturbative series will be well behaved , whereas if a large parameter is only in @xmath190 one must be careful if @xmath190 has degeneracies . similar complications can occur if ( strong ) dissipation acts only on a few sites . let us summarize our findings for the xxz chain with bulk dephasing and the half - filling sector : for small dephasing @xmath102 the gap is @xmath193 for @xmath162 , while it scales as @xmath153 for @xmath194 . critical dephasing @xmath104 decays algebraically with @xmath4 for @xmath162 while it is exponentially small for @xmath195 , see also table [ tab : bulk ] . perturbation series in small dephasing therefore fails in the thermodynamic limit . for non - small dephasing the gap is @xmath7 regardless of @xmath126 , as one would expect for a diffusive system . perturbation series for large @xmath21 works , while it fails if @xmath126 is the only large parameter . we have seen in the xx and xxz models that the gap can be constant for a sufficiently weak bulk dissipation . problem with these two cases is that the critical dissipation goes to zero in the thermodynamic limit . a natural question is : can one have a constant gap also for non - small dissipation ? the answer is yes and we are going to give a simple example . known are examples with only dissipation and no hamiltonian @xcite . an example we are going to present has a nonzero hamiltonian and nonzero dissipation . it is a xx chain with an incoherent `` hopping '' given by lindblad operator @xmath196 at each site @xmath73 , @xmath197 . the liouvillian @xmath11 again conserves magnetization difference @xmath34 and particle number @xmath35 . numerically diagonalizing @xmath11 in a @xmath44 particle sector ( @xmath41 ) one gets gaps shown in fig . [ fig : spsm ] . in the @xmath198particle sector of the xx chain with an incoherent one - way hopping ( [ eq : oneway ] ) . the kink at @xmath199 is because the eigenvalue responsible for @xmath2 goes from being complex to real.,width=307 ] we can see that asymptotically the gap is independent of @xmath4 . because in the 1-particle sector interaction asymptotically does not matter , the same asymptotic gap would be obtained also in the xxz chain with an incoherent hopping . beware that if we would take the xx chain with an incoherent hopping in both directions ( i.e. , adding additional lindblad operators @xmath200 ) the gap in the 1-particle sector would asymptotically decay as @xmath201 , see also ref . @xcite . another way of having a constant gap is to use a so - called thermal dissipators ( also known as davies generators @xcite ) that can be derived in the limit of weak coupling to thermal reservoirs . the stationary state of such a master equation is thermal , though with lindblad operators that are not local . for such models one can in certain cases rigorously prove that the gap is constant @xcite . in this section we are going to study the gap in open systems with boundary dissipation , that is , with @xmath20 acting in the thermodynamic limit nontrivially only on a finite number of sites . in all our cases @xmath16 will always be a spin@xmath202 chain and dissipation will act on the leftmost and rightmost lattice sites . as mentioned , for boundary - driven open systems the gaps observed in the literature are all @xmath1 or smaller . examples are @xmath1 in the xy chain with boundary magnetization driving @xcite ( or @xmath203 at nonequilibrium phase transition points @xcite ; or even @xmath204 on the so - called resonances @xcite ) ; @xmath1 is the scaling also for the magnetization - driven xxx model @xcite , or for the xxz model with an incoherent hopping as a driving @xcite . considering these results one can ask whether the gap in a boundary - driven open system is perhaps always @xmath1 or smaller ? as we will see the answer is no . but let us first present an argument that the gap can not be larger than @xmath0 . let us have a one - dimensional system described by a local liouvillian , @xmath205 , which is a sum of local term @xmath206 , each of which acts nontrivially only on a fixed number of consecutive sites around site @xmath73 ( hilbert space dimension of a single site is finite , implying that @xmath206 are bounded , and the steady state is assumed to be unique ) . in addition , let all but a fixed number of @xmath206 be purely unitary , i.e. , dissipation is present only in a fixed number of @xmath206 . in the thermodynamic limit the number of consecutive sites without any dissipation therefore grows linearly with the system length @xmath4 . the question we want to ask is : what is the fastest possible relaxation time , i.e. , the largest possible gap , in such an open system ? we will show that the gap can not be larger than @xmath0 , or , in other words , relaxation can not happen in a time that grows with @xmath4 slower than linearly . the argument is actually very simple . because there are sites that are of distance @xmath5 away from the nearest site with dissipation , a `` disturbance '' at that site can not dissipate in a time smaller than @xmath207 , i.e. , the time a disturbance needs to get to a site with dissipation . we can also use a transport argument : for unitary evolution local conservation of energy holds and therefore , because the local energy current is a bounded operator , it will take at least a time @xmath207 for the energy of the initial state to be dissipated , if we choose an initial state having total energy proportional to @xmath4 . one could also rigorously formulate the above argument by e.g. using the lieb - robinson bound for open systems @xcite . the lieb - robinson bound essentially formalizes a statement that there is a finite propagation speed in bounded locally - coupled systems . one consequence is that connected correlations of local operators get exponentially suppressed outside of a light cone , or , that the heisenberg picture @xmath208 of the initial local operator can be approximated by a part of @xmath11 with support inside the light cone ( outside the light cone one has @xmath209 ) . taking a local @xmath210 with support on sites that are a distance @xmath5 away from dissipation , one immediately sees that , because one has @xmath211 , relaxation time can not be smaller than @xmath5 . in next two subsections we are going to study by now familiar xx and xxz models , showing that the lieb - robinson bound @xmath212 is never reached , and then in the last subsection show some examples for which one does get @xmath212 . we take the xx chain ( [ eq : xx ] ) with dephasing on the first and the last sites , that is , with only two lindblad operators @xmath213 and @xmath214 from eq . ( [ eq : dephasing ] ) . similarly as in the xx chain with dephasing on all sites , the gap is again the same in all @xmath35-particle sectors and we can limit our discussion to the 1-particle sector . in the 1-particle sector for the xx chain with boundary dephasing of strength @xmath215 . three crosses are eigenvalues that have the largest real part and give @xmath2 ( [ eq : boundaryxxsmallg ] ) . right : numerical @xmath216 ( symbols ) for @xmath217 and prediction by eq.([eq : boundaryxx ] ) ( full curve).,title="fig:",width=158 ] in the 1-particle sector for the xx chain with boundary dephasing of strength @xmath215 . three crosses are eigenvalues that have the largest real part and give @xmath2 ( [ eq : boundaryxxsmallg ] ) . right : numerical @xmath216 ( symbols ) for @xmath217 and prediction by eq.([eq : boundaryxx ] ) ( full curve).,title="fig:",width=148 ] we are first going to use perturbation theory for small @xmath21 . an example of a spectrum of @xmath11 in the 1-particle sector is in fig . [ fig : eigboundxx ] . one observes that for small @xmath21 the largest eigenvalue , determining the gap , is complex and has the largest absolute value of the imaginary part ( two crosses in left fig . [ fig : eigboundxx ] ) . this pair of complex eigenvalues determines the gap all the way upto the maximum of @xmath216 ( which for @xmath217 shown in right frame of fig . [ fig : eigboundxx ] happens around @xmath218 ) . for larger @xmath21 the gap is given by a real eigenvalue ( the cross on a real line in fig . [ fig : eigboundxx ] ) which though has in the leading order the same real part as the complex pair . this means that one can use simple nondegenerate perturbation theory to get the gap for small @xmath21 . the eigenvalue with the largest ( or smallest ) imaginary part of @xmath81 is @xmath219 , where we denote eigenvalues of @xmath16 by @xmath220 , where @xmath221 , and eigenstates by @xmath222 , where @xmath72 denotes a state with all spins down apart from the @xmath73-th spin . the eigenvalue correction will then be given by @xmath223 , where dephasing @xmath20 is a sum of nontrivial parts acting on the first and the last sites . @xmath20 gives a non - zero value ( equal to @xmath224 ) only when acting on a non - diagonal operator on the first / last site , and therefore @xmath225 is a sum of terms that are either @xmath226 or @xmath227 ( or their hermitian adjoint ) , i.e. , @xmath228 , at the end resulting in @xmath229 , \label{eq : sum}\end{aligned}\ ] ] where @xmath230 . for large @xmath4 the term in the first line of eq . ( [ eq : sum ] ) is subdominant , while the second line gives a large-@xmath4 expression for @xmath231 and therefore also the gap , resulting in @xmath232 it is instructive to qualitatively understand how the @xmath1 scaling comes about : we see that @xmath233 is just a product of eigenstate expansion coefficients @xmath234 and that the sum scales as @xmath235 , which in turn is the smallest overlap probability at the first site , giving one @xmath236 due to normalization and an additional @xmath237 due to the longest - wavelength eigenstate which is the slowest to relax . such a scenario has already been observed @xcite in the xx chain with boundary injection of particles , i.e. , a model with lindblad operators @xmath238 at the boundaries instead of our @xmath239 , making it exactly solvable . one can in fact argue that such @xmath1 scaling is generic for boundary driven models with integrable hamiltonian @xmath190 having a plane - wave like longest - wavelength eigenstates . for large @xmath21 one on the other hand expects the gap to decay with dephasing strength as @xmath240 because of an effective decoupling of the system and as a consequence increasing relaxation time . one can extend small-@xmath21 expression ( [ eq : boundaryxxsmallg ] ) to also have the correct large-@xmath21 behavior by writing , @xmath241 one can see in right fig . [ fig : eigboundxx ] that this expression indeed fits well numerical results ( small discrepancy seen is due to a subleading @xmath133 correction to eq . ( [ eq : boundaryxx ] ) ) . observe that for the boundary driven xx system , as opposed to the xx with bulk dephasing , there is no transition in the scaling of the gap as one varies @xmath21 . for the xxz model with boundary dephasing we take a standard xxz hamiltonian ( [ eq : xxz ] ) and the same two lindblad operators @xmath213 and @xmath214 ( [ eq : dephasing ] ) . we begin by discussing weak dephasing . for weak dephasing one observes that the gap comes from real eigenvalue of @xmath11 and one therefore has to use degenerate perturbation theory in an appropriate @xmath35-particle subspace ( again , always limiting to @xmath41 ) . first order already gives a non - zero contribution and so the procedure is very similar to the one already used for the xxz chain with bulk dephasing , leading to eq . ( [ eq : r ] ) . repeating the same steps now for @xmath20 that acts only on the boundary two sites , one gets @xmath242 where @xmath138 are eigenstates of the xxz chain in an appropriate @xmath35-particle sector . eigenvalues of @xmath143 are for small @xmath21 equal to the largest real eigenvalues ( divided by @xmath21 ) of @xmath11 . doing the calculation in the 1-particle sector one gets @xmath243 for small @xmath215 and the theory ( full curves , eq . ( [ eq : boundxxzweak1mag ] ) ) . ( b ) @xmath216 always scales as @xmath1 ; symbols are numerical values for two different sizes and @xmath126 , full curves are eq . ( [ eq : boundxxz1mag]).,title="fig:",width=158 ] for small @xmath215 and the theory ( full curves , eq . ( [ eq : boundxxzweak1mag ] ) ) . ( b ) @xmath216 always scales as @xmath1 ; symbols are numerical values for two different sizes and @xmath126 , full curves are eq . ( [ eq : boundxxz1mag]).,title="fig:",width=148 ] for small @xmath21 and in the 1-particle sector the gap therefore scales as @xmath1 irrespective of the value of @xmath126 . for a non - small @xmath21 expression that fits numerics well and has a correct weak - dephasing limit ( [ eq : boundxxzweak1mag ] ) is @xmath244 we see in fig . [ fig : boundxxz]b that the perturbative result in eq . ( [ eq : boundxxzweak1mag ] ) , i.e. , horizontal dependence for small @xmath21 in the figure , holds up - to an @xmath4-independent @xmath21 . we now move to the half - filling sector . in fig . [ fig : boundxxzhalf ] we show the results . for large @xmath4 the gap scales as @xmath245 in the gapless phase @xmath148 , while it is exponentially small , @xmath246 , in the gapped phase of @xmath194 @xcite . comparing gaps in the 1-particle and the half - filling sector one also observes ( data not shown ) that for @xmath148 the gap in the 1-particle sector is the smallest of the two , while for @xmath194 the smallest gap is from the half - filling sector . for arbitrary @xmath21 the global gap therefore scales as @xmath1 for @xmath148 and @xmath247 for @xmath194 , see also table [ tab : bound ] . an exponentially small gap can be most easily understood for small @xmath21 via matrix @xmath143 . without loss of generality we limit to even @xmath4 and focus on the half - filling sector . for matrix elements of @xmath143 eigenstates of @xmath16 and matrix elements of lindblad operators is what matters . in the spectrum of @xmath16 there are two almost degenerate eigenstates @xmath248 which are the most important for the gap , i.e. , for the largest nontrivial eigenvalue of @xmath143 . for large @xmath126 they are a symmetric and antisymmetric combination of a domain wall in the middle of the chain , @xmath249 and @xmath250 , where we denoted @xmath251 and @xmath252 ( @xmath253 is a state with the left - half of spins down , and @xmath254 a state with the right - most half of spins down ) . for finite @xmath126 domain wall states @xmath255 differ from a perfect wall only within a localization length @xmath256 of the middle spin , see e.g. the appendix in ref . @xcite . for @xmath194 the two states @xmath248 are therefore `` localized '' around the site @xmath257 , their energy is almost degenerate , they are the highest energy states ( in the half - filling sector ) , and they are gapped by @xmath258 from the rest of the spectrum . they have an important property that @xmath259 and @xmath260 , i.e. , a subspace spanned by @xmath248 is invariant under @xmath261 . as a consequence , overlaps @xmath262 are exponentially small in @xmath4 and in the matrix @xmath143 a @xmath263 block corresponding to these two eigenmodes is decoupled from the rest . this @xmath263 block is of the form @xmath264 resulting in two eigenvalues @xmath265 and @xmath266 , with the corresponding liouvillian decay eigenmodes @xmath267 for eigenvalue @xmath266 , and @xmath268 for a symmetric partner at @xmath265 ( remember that the steady state is a uniform mixture of all projectors , @xmath269 ) . the gap is thus equal to @xmath270 which is in turn determined by the localization length of two localized eigenmodes . the rate @xmath271 in @xmath246 is therefore proportional to the inverse localization length , resulting in @xmath272 . exponentially slow relaxation found for @xmath194 could be interesting for instance for quantum memory we see that the domain wall can support a one - qubit quantum memory formed out of @xmath248 which is exponentially resilient to dephasing . unfortunately , as we shall see in the next magnetization - driven model , this resilience is lost for boundary dissipation in a transverse direction . we note that one can also get exponentially small gap because of localization due to disorder , an example being a boundary driven xy chain with a disordered magnetic field @xcite . , eq . ( [ eq : boundaryr ] ) , determining the gap for small @xmath21 ( [ eq : xxz - weak ] ) . asymptotic behavior is denoted by dotted line ( @xmath273 ) and full curve ( @xmath274 ) . ( b ) the gap for non - small @xmath118 . asymptotically the scaling is the same as for small @xmath21 , namely , the dotted line is @xmath275 ( @xmath129 ) and the full curve is @xmath274 ( @xmath158).,title="fig:",width=153 ] , eq . ( [ eq : boundaryr ] ) , determining the gap for small @xmath21 ( [ eq : xxz - weak ] ) . asymptotic behavior is denoted by dotted line ( @xmath273 ) and full curve ( @xmath274 ) . ( b ) the gap for non - small @xmath118 . asymptotically the scaling is the same as for small @xmath21 , namely , the dotted line is @xmath275 ( @xmath129 ) and the full curve is @xmath274 ( @xmath158).,title="fig:",width=153 ] so - far we have used dephasing as our canonical example of dissipation because it eased up theoretical analysis . steady states are in those cases rather simple , namely uniform mixtures of all projectors in a respective symmetry subspace ( appendix [ app : a ] ) . much more interesting steady states arise if one has boundary driving that breaks symmetries , for instance imposing a left - right asymmetry by having different driving at chain ends . in such a case the steady state will be a genuinely nonequilibrium state with nonzero currents flowing through the system . in this subsection we are going to study one such example , on one hand to contrast the obtained scaling with that of a boundary dephasing case and on the other hand to calculate the gap for a physically much relevant open system . we shall take the same xxz model as in previous sections ( [ eq : xxz ] ) , without dephasing , but instead with a boundary magnetization driving described by 2 lindblad operators at each end , @xmath276 driving parameter @xmath277 parametrizes the asymmetry between injection and absorption of particles at the boundary , trying to impose expectation value of @xmath18 equal to @xmath278 at chain ends , while @xmath279 is the coupling strength . nonzero @xmath277 therefore causes a nonzero magnetization gradient along the chain . we are going to use @xmath280 , however , the scalings observed ( e.g. , table [ tab : bound ] ) are independent of the precise value of @xmath277 as long as @xmath277 is not too close to @xmath281 for which one gets a blockade of transport due a step - like magnetization profile resulting also in an exponentially small gap leading to exponentially slow relaxation @xcite . the liouvillian of such magnetization - driven xxz model still conserves @xmath34 , but not anymore @xmath35 ( [ eq : const ] ) . note that @xmath279 plays the role of dissipation strength , similarly to @xmath21 in the case of dephasing dissipation small @xmath279 means weak dissipation , e.g. , weak external coupling . dissipation strength in the paper is therefore determined either by @xmath21 for dephasing ( [ eq : dephasing ] ) , or by @xmath279 for `` magnetization '' driving in eq . ( [ eq : mu ] ) . for the xxz model and boundary driving ( [ eq : mu ] ) with @xmath280 and @xmath282 . straight dashed line suggests asymptotic scaling @xmath283.,width=297 ] numerically calculated gaps are shown in fig . [ fig : drivxxz ] . the gap @xmath2 looks to have a nice @xmath1 scaling regardless of the anisotropy @xmath126 . we can compare the gap obtained here to the one obtained in previous subsection for the xxz model driven with boundary dephasing ( fig . [ fig : boundxxzhalf ] ) : in the gapless phase of @xmath162 the scaling is in both cases @xmath1 , while in the gapped phase of @xmath151 the gap is here @xmath1 while it was exponentially small for the boundary dephasing case . we see that the scaling of the gap can change already by changing boundary terms only . analyzing the gap for small coupling @xmath279 , i.e. , small dissipation , again requires application of a degenerate perturbation theory . using lindblad operators ( [ eq : mu ] ) the perturbation matrix @xmath143 is in this case @xmath284 the spectrum of such @xmath143 for the xxz chain is in fact independent of @xmath277 . for @xmath285 the above @xmath143 can be further simplified , taking into account also that @xmath16 is real , results in @xmath286 such @xmath143 is real and symmetric with the eigenvector corresponding to eigenvalue @xmath42 being a uniform superposition of all basis states ( we remind that perturbative @xmath143 has always one eigenvalue equal to @xmath42 with the corresponding eigenvector though being in general more complicated ) . the gap ( i.e. , @xmath149 ) for small coupling @xmath279 of magnetization driving therefore depends on fluctuations of @xmath287 at the boundary two sites ; this can be contrasted with the dephasing case ( [ eq : boundaryr ] ) where fluctuations of @xmath18 matter . . ( a ) the largest non - zero eigenvalue @xmath149 of @xmath143 ( eq . ( [ eq : mur ] ) ) determining the gap for small @xmath279 . full line suggests asymptotic scaling @xmath1 . ( b ) dependence of exact gap on @xmath279 and perturbative @xmath149 from frame ( a ) , all for @xmath288 ( for @xmath129 behavior is similar ) . deviations from @xmath149 begin at an @xmath4-independent @xmath279.,title="fig:",width=148 ] . ( a ) the largest non - zero eigenvalue @xmath149 of @xmath143 ( eq . ( [ eq : mur ] ) ) determining the gap for small @xmath279 . full line suggests asymptotic scaling @xmath1 . ( b ) dependence of exact gap on @xmath279 and perturbative @xmath149 from frame ( a ) , all for @xmath288 ( for @xmath129 behavior is similar ) . deviations from @xmath149 begin at an @xmath4-independent @xmath279.,title="fig:",width=158 ] in fig . [ fig : weakxxzmu ] we show results for @xmath149 and verification of its validity . we can see that for small @xmath279 the gap also scales as @xmath1 irrespective of @xmath126 and that approximation @xmath289 holds upto an @xmath4-independent @xmath279 . what is interesting is that , compared to the xxz model with boundary dephasing we changed only boundary driving , that is , we modified only action of @xmath11 on @xmath39 out of @xmath4 sites , and nevertheless the scaling of gap for @xmath194 changes from exponential to algebraic . on the level of perturbative matrix @xmath143 this can be understood as being due to the breaking of the underlying symmetry of two domain - wall states . because the location of a symmetry - breaking perturbation is important ( [ eq : mursimpl ] ) , having @xmath290 dissipation at sites other than the boundary ones could still result in an exponentially small gap , see also recent ref . @xcite for a study of stability of edge modes to markovian dissipation . we note that the scaling of the average magnetization current in the nonequilibrium steady state looks diffusive @xcite for this model in the gapped regime while higher current cumulants show anomalous non - diffusive scaling @xcite . non - diffusive scaling @xmath245 of the gap observed here ( irrespective of @xmath126 ) is perhaps an additional indication speaking in favor of anomalous transport properties of the gapped heisenberg model . so - far in all models studied the gap was @xmath1 or smaller . on the other hand the lieb - robinson argument puts a larger upper bound on the gap of @xmath0 . one can wonder whether this upper bound can be saturated ? we are going to demonstrate that there are models with @xmath212 . to achieve that we are going to take chaotic models with boundary driving . we are first going to consider the xxz chain in a staggered magnetic field , @xmath291 with staggered field having a period of 3 sites , @xmath292 , for which the hamiltonian is quantum chaotic @xcite and shows diffusive magnetization transport @xcite . dissipation will be dephasing on the 1st and the last site . similarly as in other models , @xmath34 and @xmath35 are conserved and one can look at the gap in each @xmath35-particle sector with @xmath41 . in the 1-particle sector the staggered field has no influence on the asymptotic gap and eq.([eq : boundxxz1mag ] ) , with scaling @xmath1 describes asymptotic @xmath2 well , see fig . [ fig : stagg1mag ] . for the staggered xxz chain with boundary dephasing ( @xmath118 ) in the 1-particle sector . dashed black lines are eq.([eq : boundxxz1mag ] ) , having asymptotic decay @xmath1.,width=307 ] in the half - filling sector we are first going to evaluate perturbation theory for small @xmath21 . the procedure is exactly the same as for the xxz model without the field one has to calculate the largest non - zero eigenvalue @xmath149 of the matrix @xmath143 written in eq . ( [ eq : boundaryr ] ) using eigenstates of the staggered xxz chain with the gap then being given by @xmath293 . results are in fig . [ fig : staggxxzpert ] . for @xmath162 eigenvalue @xmath149 scales as @xmath0 , while for @xmath151 ( data not shown ) it is exponentially small in @xmath4 . and weak boundary dephasing . ( a ) the largest non - zero eigenvalue of @xmath143 , eq . ( [ eq : boundaryr ] ) , determining the gap for small @xmath21 . asymptotic behavior ( full line ) is @xmath294 . ( b ) with increasing size the convergence radius @xmath104 upto which perturbation theory holds decreases.,title="fig:",width=148 ] and weak boundary dephasing . ( a ) the largest non - zero eigenvalue of @xmath143 , eq . ( [ eq : boundaryr ] ) , determining the gap for small @xmath21 . asymptotic behavior ( full line ) is @xmath294 . ( b ) with increasing size the convergence radius @xmath104 upto which perturbation theory holds decreases.,title="fig:",width=158 ] from fig . [ fig : staggxxzpert]b we also see that the validity of small-@xmath21 approximation shrinks with growing @xmath4 . for non - small @xmath21 we numerically calculated gaps in the half - filling sector , see fig . [ fig : stagghalf ] . besides exact diagonalization we also used tdmrg to obtain the gap , observing relaxation of @xmath295 , however , the required matrix dimension increases with @xmath4 rapidly and one can not go to large system sizes . we see that for @xmath296 the gap scales as @xmath0 , which is different than without the staggered field , when it is @xmath1 ( fig . [ fig : boundxxzhalf]b ) . for @xmath151 though the gap is exponentially small , the same as without the staggered field . explanation is again in terms of localized modes . for the staggered xxz chain with boundary dephasing ( @xmath118 ) in the half - filled sector . the full black line in the main plot is @xmath236 ; in the inset it is @xmath297 . full symbols are obtained by exact diagonalization , empty with tdmrg.,width=307 ] in the gapless phase the smallest gap is from the 1-particle sector , in the gapped phase it is from the half - filling sector . the global gap therefore scales as @xmath1 for @xmath296 and as @xmath247 in the gapped phase . the fact that the gap remains @xmath1 in the 1-particle sector despite chaoticity is not surprising . what is interesting and puzzling is that in the half - filling sector and for @xmath296 the gap looks @xmath0 ( at least for the sizes available , fig . [ fig : stagghalf ] ) , despite chaoticity and diffusive transport . we do not understand at present how such fast relaxation is compatible with diffusion , we note though that relaxation towards the steady state and transport properties of the steady state are in principle two separate properties . it could be that the above fast relaxation is due to conservation of @xmath35 . we therefore take the same xxz chain with staggered field as above , eq . ( [ eq : xxzstagg ] ) , but this time with boundary magnetization driving of eq . ( [ eq : mu ] ) instead of with dephasing . the value of driving is chosen to be @xmath280 . now the liouvillian conserves only @xmath34 , and the reported gaps are for the @xmath41 sector ( eigenvalues in other sectors have larger gaps ; sectors @xmath52 also do not contain any steady state ) . the steady state is nontrivial and represents a nonequilibrium state with nonzero magnetization current . , eq . ( [ eq : mur ] ) , determining the gap for small @xmath279 , scales as @xmath0 . ( b ) agreement between perturbative @xmath149 from ( a ) with the exact gap @xmath298 , all for @xmath129 . convergence radius @xmath299 decreases with @xmath4.,title="fig:",width=148 ] , eq . ( [ eq : mur ] ) , determining the gap for small @xmath279 , scales as @xmath0 . ( b ) agreement between perturbative @xmath149 from ( a ) with the exact gap @xmath298 , all for @xmath129 . convergence radius @xmath299 decreases with @xmath4.,title="fig:",width=158 ] for the chaotic xxz model with staggered field ( [ eq : xxzstagg ] ) and boundary magnetization driving ( [ eq : mu ] ) , @xmath300 . full symbols are exact diagonalization while empty are obtained from tdmrg . straight lines suggest @xmath301 and @xmath302 asymptotic behavior.,width=307 ] for small dissipation one can again use perturbation analysis in dissipation strength @xmath279 . everything is similar to the case of the xxz model without staggered field . the perturbation matrix @xmath143 is given in eq . ( [ eq : mur ] ) . for the xxz model with a staggered field eigenvalues of @xmath143 do dependent on @xmath277 , however dependence of @xmath149 is very weak , with the correction at @xmath280 being less than @xmath303 . therefore one could again use a simplified expression for @xmath143 given in eq . ( [ eq : mursimpl ] ) . data in fig . [ fig : staggmupert ] show that @xmath149 , and therefore also @xmath2 for small @xmath279 , scales as @xmath0 , with the range of validity ( convergence radius @xmath299 ) decreasing with system size @xmath4 . observe also that the effect of staggered field is much more pronounced than in the staggered xxz with boundary dephasing ( fig . [ fig : staggxxzpert ] ) . going to non - small values of dissipation @xmath279 , [ fig : drivstaggxxz ] , the asymptotic scaling of the gap seem to be @xmath0 for @xmath148 , while for @xmath194 it looks like @xmath108 , in both cases though convergence is less clear than in other models . for @xmath148 the asymptotic gap certainly seems to be larger than @xmath7 , showing that the liouvillian gap for a boundary - only dissipation does not necessarily reflect diffusivity of the hamiltonian . for the tilted ising model ( [ eq : ti ] ) with boundary dephasing of strength @xmath118 . points are numerical calculations : full squares for exact diagonalization , and empty squares for tdmrg.,width=307 ] the last model that we are going to study is again quantum chaotic one , but one that no longer conserves @xmath34 . we take the ising chain in a tilted magnetic field , @xmath304 which is quantum chaotic @xcite for generic field direction ( we use @xmath305 and @xmath306 ) and with diffusive energy transport @xcite . for small dephasing strength @xmath21 the gap will again be equal to @xmath293 , where @xmath149 is the largest non - zero eigenvalue of matrix @xmath143 ( [ eq : boundaryr ] ) . numerically computed @xmath149 for sizes @xmath307 show scaling @xmath308 with @xmath309 ( data not shown ) . we have also numerically calculated @xmath2 ( fig . [ fig : ti ] ) for a system with boundary dephasing ( [ eq : dephasing ] ) of strength @xmath118 . for smaller @xmath4 we used exact full diagonalization or the arpack package , while for @xmath310 we used tdmrg @xcite , inferring @xmath2 from the observed relaxation rate of total energy , the initial state being a pure n ' eel state @xmath311 . for this model we could get @xmath2 with tdmrg for significantly larger systems than for the other two chaotic models . we can see from fig . [ fig : ti ] that the decay seems to be algebraic @xmath312 with @xmath271 being between @xmath313 . while smaller @xmath4 seem to be described by @xmath314 , for larger we get nicer fit with @xmath315 . we in principle can not exclude the scaling becoming @xmath7 for still larger @xmath4 , however , for the available data the scaling seems to be distinctively different than @xmath7 . let us briefly summarize our results and point to some interesting findings and open issues . we have studied open quantum spin chains with two types of sites at which dissipation acts . the first case was chains with bulk dissipation in which dissipation acts on all ( or most ) sites , while the second case was chains with boundary dissipation . scaling of the gap for models with bulk dissipation is summarized in table [ tab : bulk ] . rcrcrrcr & & + & & xx+deph . & & & & xx+hopp . + ( r)3 - 3 ( r)5 - 6 ( l)8 - 8 & & & & @xmath148 & @xmath194 & & + + @xmath316 & & @xmath317 & & @xmath317 & @xmath317 & & @xmath61 + @xmath318 & & @xmath61 & & @xmath61 & @xmath61 & & + @xmath104 & & @xmath236 & & @xmath236 & @xmath236 + + @xmath316 & & @xmath317 & & @xmath317 & @xmath317 & & + @xmath318 & & @xmath61 & & @xmath61 & @xmath319 & & + @xmath104 & & @xmath236 & & @xmath320 & @xmath321 & & rc r c rr c rr c rr c rr c r & & + & & xx+deph . & & & & & & & & & & tilted ising+deph . + ( r)3 - 3 ( r)5 - 6 ( r)8 - 9 ( r)11 - 12 ( r)14 - 15 ( l)17 - 17 & & & & @xmath148 & @xmath194 & & @xmath148 & @xmath194 & & @xmath148 & @xmath194 & & @xmath148 & @xmath194 & & + + @xmath316 & & @xmath322 & & @xmath322 & @xmath322 & & & & @xmath322 & @xmath322 & & & & + @xmath318 & & @xmath322 & & @xmath322 & @xmath322 & & & & & @xmath322 & @xmath322 & & + + @xmath316 & & @xmath322 & & @xmath322 & @xmath323 & & @xmath322 & @xmath322 & & @xmath236 & @xmath323 & & @xmath324 & @xmath325 & & @xmath326 + @xmath318 & & @xmath322 & & @xmath322 & @xmath323 & & @xmath322 & @xmath322 & & @xmath236 & @xmath323 & & @xmath236 & @xmath324 & & @xmath327 + @xmath104 or @xmath299 & & @xmath328 & & & & & & @xmath329 & @xmath330 & & @xmath331 & @xmath331 not surprisingly , because there are no fundamental limitations on the gap , one also finds all different scalings . quite universally , the gap scales differently for small dissipation strength than for non - small dissipation ( dissipation strength is in our cases mostly dephasing @xmath21 ) . systems with bulk dissipation therefore typically undergo a ( nonequilibrium ) phase transition from a phase dominated by @xmath16 to a phase where dissipation is dominating . such transition in the decay mode has been analyzed in detail for the xx chain with dephasing . in our models the critical dissipation strength @xmath104 at which this transition happens always goes to zero in the thermodynamic limit . for boundary driven models the gap scaling is summarized in table [ tab : bound ] . they all comply with a general bound prohibiting faster than @xmath212 relaxation . compared to bulk - dissipated cases , here the scaling seems the same for small and for non - small dissipation ( except perhaps for the tilted ising case , and the staggered xxz model with magnetization driving ; in both cases though finite - size effects could still be at play ) . in the 1-particle sector the gap is always @xmath1 due to essentially the solvability of the smallest nontrivial subspace ( eventhough a model might be chaotic in larger invariant subspaces ) and the longest eigenmodes having wavelength @xmath207 . a number of transitions in the scaling of @xmath2 can be identified . at each such transition there is a possible ( nonequilibrium ) phase transition that would be interesting to explore in more detail . changing a bulk parameter , like the anisotropy @xmath126 or the staggered field , can change the scaling ( e.g. , from algebraic to exponential , or from @xmath1 to @xmath0 ) . more interestingly , the scaling can also change by changing a boundary dissipation only , that is , changing only terms that have a relative weight @xmath332 in the liouvillian @xmath11 . an example is the gapped xxz chain for which the gap is exponentially small if dephasing is at the boundary , while it is algebraic for magnetization driving at the boundary . such a transition is due to a symmetry breaking of otherwise protected subspace . one finding that needs further exploration is a fast @xmath212 relaxation in chaotic models . in the present work we studied only the spectral gap , without detailed discussion of the associated eigenvector properties . of particular interest is locality of decay modes and with it connected relaxation of local observables . namely , it has been observed in lindblad equations as well as in classical systems @xcite that ( certain ) local observables can relax in a time that is smaller and scales differently than the global gap . for instance , in the xxx chain with lindblad magnetization driving the gap scales as @xmath245 whereas local magnetization and current relax as @xmath333 @xcite . in all systems studied that conserve magnetization as well as the number of particles in the bra and ket , that is @xmath34 and @xmath35 ( [ eq : const ] ) , the liouvillian eigenproblem has a block structure . in the @xmath41 sector one has a steady state in each @xmath334particle sector , and that steady state is an equal mixture of projectors to all basis states ( an ergodic diagonal state ) , @xmath335 with @xmath336 having @xmath35 spins in state @xmath75 and @xmath337 in state @xmath74 . the two most interesting subspaces are the 1-particle and the @xmath338-particle ( half - filling ) . the simplest case is @xmath339 , for which the steady state is @xmath340 taking a local operator basis composed of @xmath341 one can `` vectorize '' the operator @xmath342 , writing it as a vector in a hilbert space of operators . doing that on @xmath343 , it can be written as @xmath344 where we use the notation @xmath345 to denote vectors in the space of operators , written in the operator basis @xmath346 . the steady state in the @xmath198particle sector on @xmath4 spins , @xmath347 , is simply @xmath348 which is the so - called @xmath349-state . one can immediately see that for a bipartite cut after @xmath57 spins the schmidt decomposition is of rank @xmath39 with the two eigenvalues ( squares of schmidt coefficients ) being @xmath350 and @xmath351 , irrespective of @xmath57 . in the operator space such a steady state is therefore of finite rank @xmath39 ( it is weakly entangled ) regardless of the system size @xmath4 . the half - filling sector is composed of states with half of the spins pointing up and half pointing down . total magnetization is therefore zero . starting again with a simple example for @xmath352 , we have the steady state @xmath353 or , written as a vector , @xmath354 where @xmath355 is a uniform superposition of all @xmath356 basis states with zero magnetization ( for simplicity , we assume even @xmath4 ; for odd @xmath4 and the largest sector one has to replace @xmath338 with @xmath357 ) . regarding ( operator ) schmidt decomposition , for a cut after @xmath358 sites we see that the schmidt rank is @xmath39 with both eigenvalues being @xmath359 . for @xmath360 , i.e. , a cut of maximal size for @xmath361 , we have decomposition @xmath362 and therefore the state is of rank @xmath363 with eigenvalues being @xmath364 . we can see that the eigenvalue prefactors ( @xmath365 ) are actually of a combinatorial nature , resulting from the number of combinations of distributing @xmath366 ones on @xmath57 sites , e.g. , @xmath367 , @xmath368 , and @xmath369 . generalizing to a bipartite cut of @xmath355 after @xmath57 sites , one has @xmath370 nonzero schmidt coefficients with the eigenvalues being @xmath371 . we can see that the largest schmidt rank is for a half - cut , and is equal to @xmath372 . operator schmidt rank of the steady state in the half - filling sector grows linearly with the system size @xmath4 . we remind that a constant @xcite or a linear @xcite schmidt rank is a sign of an exact solvability of a steady state . however , our results show that the exact solvability of the lindblad steady state in general does not tell us anything about solvability of ( closest ) decay modes or the behavior of the gap . ref . @xcite uses periodic boundary conditions , facilitating a solution in fourier space . we prefer to keep open boundary conditions with a view of later - studied boundary - driven systems , where the first and the last sites are typically not coupled and the translational invariance is anyway lost due to localized dissipation . the mapping from @xmath11 to the ladder is such @xcite that ladder states from the 1-particle subspace have e.g. @xmath56 particles in @xmath75 and one in @xmath74 in the upper leg , and @xmath56 in @xmath74 and one in @xmath75 in the lower leg . for system sizes upto around @xmath373 we used full exact diagonalization of @xmath11 in the respective half - filling subspace , for @xmath4 upto @xmath374 the lanczos method ( arnoldi package ) , and for still larger @xmath4 the tdmrg method , see e.g. @xcite for further details and references about the tdmrg method .

understanding the chemical history of heavy elements ( metals " ) in the interstellar medium ( ism ) of galaxies is an important , but difficult undertaking . because refractory elements are found to be depleted from the gas phase ( locked into grains ) , they provide a substantial reservoir of coolants in the solid - state phase . interstellar grain surfaces are believed to furnish formation sites for molecular hydrogen ( h@xmath1 ) , which is both a coolant and the starting point for a rich cloud chemistry . grains also provide a critical transfer mechanism for reprocessing uv / o starlight into far - infrared and sub - mm emission from interstellar molecular clouds and high - redshift galaxies . in the high - redshift universe , the first dust grains may influence the thermodynamics of the primordial gas and thereby control rates of star formation . recent observations of dust grains at high redshift have been puzzling in the context of their formation . several observations of damped ly@xmath2 systems ( pettini 1994 ; ledoux , bergeron & petitjean 2002 ) have shown evidence of dust grains in them . also , thermal emission ( at @xmath3 mm ) from dust in high redshift ( @xmath4 ) qsos in the sloan digital sky survey ( sdss ) have been reported ( bertoldi 2003 ) . the implied dust mass of @xmath5 m@xmath6 within a gyr after the big bang appears difficult to explain with traditional models of grain formation in evolved low mass stars ( dunne 2003 ) , from which grains are thought to be transported to the interstellar medium ( ism ) through stellar winds ( whittet 1992 ) . recent models have , therefore , focused on other sites of dust grain formation , such as ejecta of core - collapse supernova explosions that can occur on shorter time scales than that of the evolution of low mass stars ( dwek & scalo 1980 ) . core - collapse supernovae are thought to be capable of synthesizing a significant amount of dust in the sn ejecta . applying a theory of nucleation and grain growth developed by kozasa & hasegawa ( 1987 ) , kozasa ( 1989 ; 1991 ) and todini & ferrara ( 2001 ) have studied grain formation in expanding sn ejecta , the latter estimating the dust mass formed @xmath7 m@xmath8 . nozawa ( 2003 ) have extended it to the case of zero - metallicity sne for population iii stars ( see also schneider , ferrara & salvaterra 2004 ) . although the sn `` dust factory model '' would explain the large observed dust masses in high - redshift galaxies , observations of dust in nearby sne appear inconsistent with this model , falling factors of 10100 short of the required amounts ( @xmath9 ; see above ) . as an example , in a review of type ii sn 2003gd ( ngc 628 ) , sugerman et al . ( 2006 ) state that radiative transfer models show that up to 0.02 solar masses of dust has formed within the ejecta " . however , a recent study ( meikle et al . 2007 ) of the same remnant concludes that the mid - ir flux is consistent with emission from @xmath10 of newly condensed dust in the ejecta " . in a young , oxygen - rich snr in the smc , stanimirovic et al . ( 2005 ) found only @xmath11 of hot dust ( @xmath12 k ) . the claimed sub - mm detection ( dunne et al . 2003 ) of @xmath13 of cold dust toward the cas a remnant was later shown to arise from interstellar dust in an adjacent molecular cloud ( krause et al . 2004 ) . observational tests of predictions made by models of sne are difficult , involving far - ir and sub - mm observations of young snrs . one must disentangle far - ir backgrounds , separate newly synthesized dust from circumstellar and interstellar dust , and understand the role of snr reverse shocks , which can destroy newly synthesized dust in high - speed ejecta . a case in point is sn 1987a . dust formation was inferred in ejecta of sn 1987a , but only @xmath11 was detected ( dwek 2006 ; dwek & arendt 2007 ) . additional dust is almost certainly present , because of observed changes in the optical and bolometric fluxes and emission - line asymmetries in the red and blue wings ( mccray 1993 , 2007 ) . moreover , the reverse shock has not yet reached the interior of the supernova debris in sn 1987a . these inner ejecta are most prone to dust formation , since they are cold , dense , and metal - enriched . a challenging theoretical issue is whether high - velocity dust ( @xmath14 km s@xmath15 ) can survive the shocks , thermal sputtering , and grain - grain collisions , as the ejecta are slowed down by the surrounding ism . a goal of this paper is to determine how much dust survives the sn event and is incorporated into the surrounding ism . the relative speed of reverse shocks with respect to the expanding ejecta can be very large , and can raise the temperature of the ejecta to the extent of destroying the nascent dust grains in it . it is therefore important to study the effect of reverse shocks in detail . recently , nozawa ( 2007 ) have considered this problem and came to the conclusion , with the help of hydrodynamical simulation , that reverse shocks can destroy a fraction of dust mass of order @xmath16% depending on the ambient gas density and explosion energy . bianchi & schneider ( 2007 ) also used numerical simulations and semi - analytical methods to conclude that only a small amount of dust mass survives ; e.g. , @xmath17% dust mass survives in the case of a sn with a progenitor of mass @xmath18 m@xmath8 and ambient particle density of @xmath19 @xmath20 . nozawa ( 2007 ) have included the motion of dust relative to gas within the sn remnant and calculated the evolution of dust grains over a period spanning the radiative phase ( @xmath21 yr ) , whereas bianchi & schneider ( 2007 ) neglected the effect of motion of grains on dust destruction and calculated dust evolution up to the non - radiative phase ( @xmath22 yr ) . we study in this paper a simple and analytically tractable problem , namely the effect of reverse shocks for a power - law mass distribution of the ejecta , in the regime of self - similar evolution of both forward and reverse shocks . although the simple formalism does not allow us to study several important processes including motion of dust grains and its effect on evolution of dust grains our approach allows one to segregate the processes of destruction within the reverse shock itself from those occuring between forward and reverse shocks ( which can be dealt with in numerical simulations involving grain motion ) . we first summarize the results of the theoretical predictions of dust grain formation in snrs . todini & ferrara ( 2001 ) calculated the formation of different types of dust grains amorphous carbon ( ac ) , silicates ( enstatites : @xmath23 , forsterites : @xmath24 ) and iron - bearing magnetites ( @xmath25 ) and others , in ejecta with different metallicities . they found that ac grains are generally of larger size , in the range of @xmath26 m , whereas silicates are produced with sizes @xmath27 m. nozawa ( 2003 ) reached similar conclusions for pop iii sn events , with sizes of newly synthesized grains spanning a range of three orders of magnitude , with maximum grain size being less than @xmath28 m . the dust grain parameters in the ism are inferred to be somewhat different from these estimates . the extinction observations are satisfactorily reproduced by assuming mostly of two components graphite and silicate with a common power - law size distribution , @xmath29 , truncated at a minimum size @xmath30 m and a maximum size @xmath31 m ( draine 2003 , and references therein ) . there also appears to be a population of very small grains with pah composition . nozawa ( 2003 ) found for grains synthesized in pop iii sn ejecta a size distribution with index @xmath32 for smaller grains and with @xmath33 for larger grains ( with the grain size at crossover point depending on supernovae models ) . we use these two power laws to bracket the possible size distributions of grains synthesized in sn ejecta , and assume that grains are formed in the range @xmath34 cm ( see figure 11 of nozawa ( 2003 ) . we also consider the case of an upper limit of @xmath35 cm and compare our results for both upper limits . it is instructive to estimate the destruction time scales before we calculate the effect of reverse shock in detail . the sputtering time scale for dust ( graphite , silicate and iron ) grains of size @xmath36 is approximately @xmath37 yr , for ambient temperatures larger than @xmath38 k ( draine & salpeter 1979 ) , where @xmath39 is electron density . for grains with size @xmath7 @xmath40 ( which incidentally contain most of the dust mass for a @xmath41 distribution ) , and for @xmath42 @xmath43 , the sputtering time scale is @xmath44 yr . we now briefly discuss the self - similar evolution of forward and reverse shocks in sn ( see truelove & mckee 1999 , hereafter referred to as tm99 , for further details ) . the parameters used to describe a supernova remnant which affect its evolution are the energy of the explosion ( @xmath45 ) , the mass contained in the expanding ejecta ( @xmath46 ) , the maximum velocity of the material within the ejecta ( @xmath47 ) , and the density of the ism ( @xmath48 ) . denoting the time of the explosion as @xmath49 , we construct our initial conditions at a later time , still early in the history of the remnant , assuming the ejecta to have already expanded to a radius @xmath50 without any deceleration due to the ambient medium . if we consider sn to be a point explosion , then the radial velocity profile will be linear as ejecta with velocity @xmath51 will travel out to a distance @xmath52 : @xmath53 this profile becomes flatter as time progresses and the faster - moving ejecta are flung out to greater distances . the corresponding radial density distribution of the ejecta matter is given by : @xmath54 where @xmath55 is the time - independent structure function of the density profile and the @xmath56 term arises from the free expansion of the ejecta . we define a dimensionless parameter to label shells of the ejecta starting from the centre outwards as w = , 0 w 1 following tm99 we consider the situation when _ f _ is given by a power law in _ w _ : f(w ) = \ { ll f_0 , & + f_nw^-n , & . where we have separated the ejecta into a uniform density core region and a power - law envelope region . the continuity of _ f _ at @xmath57 and the normalization of @xmath58 translate to expressions for parameters @xmath59 and @xmath60 in terms of the free parameters @xmath57 and _ n _ : f_0 & = & f_nw_c^-n + f_n & = & . as the freely expanding ejecta come into contact with the ambient medium , with the contact discontinuity moving at @xmath47 , a speed much larger than the sound speed in the ambient medium , a forward shock is set up in the surrounding gas which propagates into the medium . the ejecta at the contact surface decelerates suddenly as it hits the ambient gas , and consequently , a reverse shock is set up that propagates inwards through the ejecta . while the mass of this matter swept by the forward shock is less than the total mass of the ejecta driving the shock , the remnant is said to be in the _ ejecta dominated _ or ed stage . it is in this stage , before the reverse shock has attained a significant velocity , that the ejecta can be approximated as a spherical piston freely expanding into the ambient gas . when the mass swept by the forward shock becomes approximately equal to the ejecta mass , the corresponding late - time limit is known as the _ sedov - taylor _ or st stage of the evolution of the remnant . we follow tm99 to follow the evolution of the reverse shock analytically . let @xmath61 and @xmath62 denote the radii of the forward and reverse shocks , and @xmath63 and @xmath64 be the velocities of the forward and reverse shock waves respectively , all in the rest frame of the unshocked ambient medium . we have the velocity of the unshocked ejecta as seen by the reverse shock as it propagates into the ejecta : _ r v(r_r , t ) - v_r(t ) = - v_r(t ) . @xmath65 _ be the ratio of the pressures behind the reverse and forward shock waves and _ l _ be the ratio of the radii of the two shocks at a given time : ( t ) , l(t ) , w_f(t ) . it is useful to express various parameters in dimensionless forms in the units of characteristic values of these variables . so that , @xmath66 , @xmath67 , and @xmath68 . for example , the characteristic scales of length , time and velocity are given by , r_ch&= & m_ej^1/3 _ 0 ^ -1/3 5.25 n_0 ^ -1/3 pc , + t_ch&= & e^-1/2 m_ej^5/6 _ 0 ^ -1/3 1.66 10 ^ 3 n_0 ^ -1/3 yr , + v_ch&= & r_ch / t_ch 3.16 10 ^ 3 km s^-1 . [ eq : char ] ( we have assumed @xmath69 for he / h@xmath70 by number . ) we have estimated the numerical values for the case @xmath71 erg , @xmath72 g and @xmath73 @xmath20 . we will therefore write @xmath74 , @xmath75 , and @xmath76 , and the radius of the reverse shock , in particular , as @xmath77 . the values of @xmath65 and @xmath78 approach constant values as @xmath79 if the solutions remain self - similar in the same limit . we write @xmath80 and @xmath81 , and also @xmath82 . the solution to reverse shock evolution exists in the form of two branches , which we label the _ envelope _ ( @xmath83 ) and _ core _ ( @xmath84 ) branches . denoting @xmath85 , the evolution of the reverse shock is given by , & & t^(r_r^ ) = ( ) ^1/2r_r^ + & & \ { ll ^-2/(3-n ) , + ^-2/3 , . [ eq : ednl3_1 ] where the first part corresponds to the case @xmath86 and the second part , to @xmath87 . the corresponding velocity of reverse shock in these two cases are , & & v^(r_r^ ) = ( ) ^1/2 + & & \ { ll , & + , & . the relative velocity of the reverse shock in the frame of reference of the expanding ejecta , for the case when the reverse shock is in the envelope region , is given by , _ r^(r_r^ ) = lr_r^3/2 ( ) . [ eq : ednl3_2 ] note that , since @xmath88 describes the reverse shock , we will define it as ` @xmath89 ' and use it to label shells in the ejecta . this solution is valid ( in the ed stage ) only as long as the reverse shock remains in the envelope @xmath90 . in the case of @xmath91 ejecta , there is no need for a core and we can take the limit @xmath92 . for @xmath93 however , we must have a `` flat profile '' core to keep the integrated mass finite . for the steep index scenario , therefore , the functional form of the solution will change when the reverse shock reaches the core and we will return to this point later . we now describe the sedov - taylor stage , which is the late - time evolution limit of the non - radiative phase of a supernova remnant . following tm99 , we assume that the shocks trace a constant acceleration path after transition to the st stage at @xmath94 , which ( approximately ) begins when the mass of the ambient medium shocked by the forward shock equals the mass ejected by the supernova . thus the velocity of the oncoming ejecta in the frame of the reverse shock is _ r^*(t^ * ) = _ r , st^ * + _ r , st^*(t^ * - t_st^ * ) , [ eq : stnl3_2 ] where @xmath95 is the constant acceleration and @xmath96 is the velocity of the ejecta w.r.t the reverse shock at @xmath94 . note that @xmath97 . using this and integrating the resulting expression from @xmath94 to @xmath98 gives r_r ^(t^ ) = t^. [ eq : stnl3_1 ] for the @xmath91 case , we therefore have an implicit solution for the reverse shock position in the ed stage @xmath99 and an explicit solution in the form of a uniform acceleration description in the st stage . for the @xmath100 case , we again have an implicit solution in the ed stage at least while @xmath101 . in fact , an explicit solution for @xmath102 can be determined from the implicit solution in the ed stage for @xmath100 remnants . we have ( see equation ( 74 ) of tm99 ) , t^(r_r^ ) = ( ) ^1/2r_r^ ^-2/(3-n ) . in this expression , we can take the limit @xmath92 by allowing @xmath103 , the ejecta energy @xmath104 remaining finite . in this limit @xmath105 and when the second term in the brackets becomes much larger than unity , we have r_r^(t^ ) = ^1/n ( t^ ) ^ . [ eq : tlc_n5_1 ] this gives us the velocity of the reverse shock as , _ r ^(t^ ) = ^1/n ( t^ ) ^ -3/n , and correspondingly , _ r^(t^)&= & - v_r^ + & = & ^1/n t^-3/n . [ eq : tlc_n5_2 ] this is the result arrived at by chevalier ( 1982 ) and nadyozhin ( 1985 ) - henceforth referred to as ` the cn solution ' . the transition of a supernova remnant with @xmath100 to this stage has been shown to occur extremely fast ( tm99 ) . thus for all practical purposes , we may assume that the remnant begins with the cn stage . the reverse shock then moves into the core at @xmath106 . tm99 assumed that the reverse shock motion after this time is also described by a sedov - taylor trajectory - which leads to better match between the two regimes in terms of continuity of the shock radius and velocity . in the case of @xmath100 , therefore , if we replace @xmath94 by @xmath107 in the st stage equations above , the same analysis follows through . the numerical values of the parameters @xmath108 and @xmath109 are , of course , different from the parameters for the @xmath91 case : viz @xmath94 and @xmath95 . we have , r_r^(t^ ) & = & t^ + _ r^(t^ ) & = & _ r , core^+ _ r , core ^(t^- t_core^ ) . [ eq : tmc_n5_1 ] the values of the parameters used in our calculations have been taken from tm99 , which have been checked with results from simulations . we thus have the final solutions for the reverse shock as follows . in the case of @xmath91 , equations ( [ eq : ednl3_1 ] ; first part ) and ( [ eq : ednl3_2 ] ) describe the solutions in the ed stage , and , thereafter , equations ( [ eq : stnl3_1 ] ) and ( [ eq : stnl3_2 ] ) are appropriate for solutions in the st stage . for the case of @xmath110 , solutions are given by equations ( [ eq : tlc_n5_1 ] ) and ( [ eq : tlc_n5_2 ] ) in the @xmath101 stage . afterwards , in the @xmath111 stage , equations ( [ eq : tmc_n5_1 ] ) describe the solutions . the analytical solutions remain valid until @xmath112 ( tm99 ) , and we calculate the evolution of shocks until this point ( except for @xmath113 ; see below ) . for a given set of values of @xmath104 and @xmath46 , we can therefore solve for reverse shock position and relative velocity , assuming a density distribution described by index @xmath114 . the parameter @xmath2 in different cases of @xmath114 , is given by , =( 3 -n 5-n ) ( w_core^-(5-n)-n/5 w_core^-(3-n)-n/3 ) w_core^2 , for @xmath115 , we calculate this in the limit of @xmath116 , and for @xmath110 ( including the case of @xmath113 ) we assume ( following tm99 ) a value of @xmath117 . values of other parameters ( such as @xmath118 ) are taken from appropriate tables of tm99 for the relevant values of @xmath114 . we also consider the special case of @xmath113 , for which the mass integral diverges without a core . we follow the prescriptions of tm99 ( their section 8) for this case ( and , again , assume @xmath117 ) , and track the reverse shock with only ed solution without any transition to the st case . according to tm99 , based on their comparison of analytical and numerical results , this prescription gives tolerably good results ( accurate to within @xmath119 ) until @xmath120 . we have introduced a label to a given shell in the supernova ejecta by @xmath121 , the ratio of the velocity of the ejecta in the shell to the maximum ejecta velocity . the ejecta in a shell specified by a given @xmath89 first undergoes free expansion ( in the ed stage ) and self - similar expansion ( in the st stage ) until hit by the incoming reverse shock wave . this occurs at a time , which we call @xmath122 , which is a function of @xmath89 and which is found by equating @xmath89 to the @xmath123 found from the position of the reverse shock . that is , w = = r_r = w r_ej = w v_ej t_hit . or , in terms of dimensionless variables , r_r^= w = ( w ) t_hit ^. this expression for @xmath124 in terms of @xmath125 is equated to the explicit solution for the reverse shock , when available , as a function of time , and the resultant linear equation in @xmath126 and @xmath89 is solved using the bisection method for each given @xmath89 . this is applicable to the steep ( @xmath100 ) ejecta case ( both the cn and st stage ) as well as to the st stage for shallow ( @xmath91 ) ejecta . for the ed stage of shallow ejecta , we use @xmath126 as a function of @xmath127 in equation ( [ eq : ednl3_1 ] ) and then employ the bisection method exactly as before to solve for @xmath128 . either way , we can determine @xmath128 given any shell labeled by a @xmath89 in the sn ejecta . we assume that gas in a given shell of the ejecta evolves adiabatically before @xmath122 . the temperature is then raised to the postshock temperature and the density acquires a jump , corresponding to the value of relative shock velocity @xmath129 . after @xmath122 , the gas again evolves adiabatically , that is , its density decreases as @xmath56 ( equation [ eq : den ] ) . its temperature decreases as @xmath130 , for @xmath131 appropriate for monoatomic gas . for a shell with a certain value of @xmath89 , therefore , we have the following evolution of density , ( t , w ) = \ { ll f(w)t^-3 , & + 4 f(w ) t^-3 , & . [ eq : den_ev ] and temperature , t(t , w ) = \ { ll t_i ( ) ^-2 , & + t_rshock ( ) ^-2 , , & . where @xmath132 , the post - shock temperature due to the reverse shock with relative velocity @xmath133 . for a completely ionized gas the mean molecular weight @xmath134 . we assume a value of @xmath135 k at @xmath136 days , motivated by photometric observations of sn1987a ( catchpole 1987 ) where it was estimated that the photospheric temperature at @xmath137 d after the explosion was @xmath138 k. we have also checked that our results do not depend strongly on these assumptions . given the above mentioned evolution of density and temperature in shells of ejecta labeled by different values of @xmath89 , we can compute the steady decrease in grain radius , for different types of dust grains . we use the polynomial fit given by tielens et al ( 1994 ) for graphites and silicates as a function of gas temperature . they expressed the rate , @xmath139 in powers of @xmath140 , with coefficients for different grain composition tabulated in their table 4 ( see their equation 4.21 ) . these rates are consistent with recent results of nozawa ( 2006 ) . we follow a shell of ejecta material labeled by a value @xmath141 from @xmath142 10 years after the explosion ( by which time dust grains are believed to have formed in the cooling ejecta , since the formation time scale is of order a few years ( see , e.g. , todini & ferrara 2001 ) ) , to when the reverse shock hits the shell at @xmath122 and beyond , until a time @xmath143 , which is the limit of the validity of the analytical solutions . we have calculated the effect of reverse shocks on sputtering of graphite and silicate grains for typical value of explosion energy , stellar ejecta mass and ambient density . below we show the results for @xmath71 erg , @xmath144 g , @xmath145 g @xmath20 . first , we show the evolution of grain radius ( for graphites and silicates ) for two dust grains embedded in two particular shells , for example , in ejecta characterized by @xmath146 and @xmath71 erg . we show in figure 1 the cases for the shells marked @xmath147 , and @xmath148 . the reverse shock passes through these two shells , first hitting the @xmath149 shell at @xmath150 yr , and then the @xmath147 shell at @xmath151 yr . the bottom panel of figure 1 shows the rise in temperature for these two shells followed by adiabatic cooling . initially the temperature drops to very low temperatures due to strong adiabatic cooling , although , in reality , gas can not be cooled below the temperature of the cosmic microwave background , but we have not used such a lower bound in our calculations . at any rate , the post - shock temperature does not depend on the low temperature of the gas upstream , being determined only by the strength of the reverse shock , and so the results of our calculation remains realistic in spite of the strong cooling before the reverse shock hits the gas in a given shell . figure 2 plots the ratio of final to initial grain sizes as a function of the initial grain sizes for silicates and graphites forming ( and being sputtered ) in shells @xmath152 for the particular case of @xmath146 . _ since the sputtering yield of silicate grains is larger than that of graphites for a given set of parameters _ , silicates are eroded more rapidly than graphite grains for sputtering in a given ejecta shell . also , the relatively outer shell of @xmath149 experiences less sputtering than the inner shells of @xmath153 and @xmath154 . this is expected since the reverse shock ( relative ) speed picks up as it plows through the ejecta , increasing with decreasing values of @xmath89 . to illustrate this point , we plot the trend of the relative velocity of the reverse shock as a function of @xmath89 for the shells hit by the shock in figure 3 , for @xmath146 ( solid ) , @xmath155 ( dotted ) , and @xmath156 ( dashed ) . the relative velocity is a rising function with decreasing @xmath89 , for @xmath146 and for @xmath110 , and it is a peaked function for @xmath155 ( and also for @xmath113 , which is not shown here , but as shown by tm99 ) . however , we notice in figure 2 that grains ( both graphites and silicates ) in shell @xmath153 ( solid curve ) are sputtered to a somewhat lesser extent than grains in the @xmath157 shell , although the relative speed of the reverse shock should be ( according to figure 3 , solid curve for @xmath146 ) higher when it hits the @xmath153 shell than at the @xmath157 shell . although the post - shock temperature at @xmath153 is higher than at @xmath157 , the shells in the inner region encounter the reverse shock later than shells in the outer region of ejecta , and by the time they encounter it , the density in the inner shells would have decreased rapidly ( see equation [ eq : den_ev ] ) . the rarefaction in the inner shells at the time of encounter with the reverse shocks lessens the effect of dust sputtering in these shells . we illustrate this point by considering grains which are sputtered below a certain size , say , @xmath158 cm ( @xmath159 m ) , as a function of the shell in which they reside ( and get sputtered ) . we determine the maximum size of grains which are sputtered below a size @xmath160 cm , as a function of @xmath89 , for different shells , and present the results in figure 4 . curves for different values of @xmath114 show the maximum size of grains sputtered below @xmath161 m as a function of the parameter @xmath89 ( left panels ) , as well as against the fraction of ejecta mass that is contained inside of the concerned shell , @xmath162 in the right panels . results for graphites and silicate grains are shown in the upper and lower panels respectively . figure 4 clearly shows that sputtering becomes intense in intermediate shells , and decreases ( as a result of rarefaction ) with decreasing value of @xmath89 for small values of @xmath89 . the sizes of grains which are sputtered below @xmath163 m lie in the range of @xmath164 m @xmath165 m , with silicates more strongly sputtered than graphites . in all cases , the curves of the maximum size of sputtered grains increases with decreasing values of @xmath89 and @xmath162 , finally plunging to very low values for @xmath166 , indicating negligible effect of the reverse shock deep inside the ejecta . for example , the ( dotted ) curve for @xmath155 shows a peak at @xmath167 below which the maximum size of sputtered grains decreases . also , although it is not explicitly shown in figure 4 , the corresponding results for @xmath168 show that reverse shocks do not penetrate most of the ejecta mass in the time period of the validity range of analytical solutions @xmath112 . analytical means allow us to probe deep into the ejecta mass distribution only for @xmath91 , and full hydrodynamical models are needed to address the cases of steeper profiles . for steep density distributions , the limiting time of @xmath169 is reached by the time the reverse shock only skims the surface of the ejecta , as far as mass fraction is concerned . the characteristic time in the case considered here ( @xmath71 erg , @xmath72 g , ambient density @xmath145 g @xmath20 ) is @xmath170 yr . although , by the limiting time of @xmath171 yr for the validity of self - similar regime , reverse shocks can hit shells with small @xmath89 in these steep cases , the mass contained inside of these shells remains very large . in other words , in these cases we can not probe what happens to most of the ejecta mass , or the dust contained in them with analytical means , within the limits of the self - similar regime of shock evolution . we also note that a substantial amount of grain sputtering takes place in the hot gas between the forward and reverse shock that analytical methods can not capture . detailed hydrodynamical calculations are needed to address these issues , and our result can only provide a lower limit to the total mass fraction of dust that is destroyed . finally , we calculate the total dust mass sputtered away as a function of initial grain sizes in the following manner . we first calculate for a given shell @xmath89 , the final grain size @xmath172 corresponding to the initial grain size @xmath173 . then the fraction @xmath174 is the total dust mass sputtered for grains formed in this shell . we then calculate the total dust mass sputtered for all shells ( for a given initial grain size @xmath173 ) , weighted by the shell mass in which they are situated , as given by , f_d ( a_i)= d m(<w ) m_ej , where @xmath175 is the ejecta mass contained within the shell @xmath89 . we plot these weighted fractions @xmath176 for graphite and silicate grains as functions of @xmath173 for @xmath177 in figure 5 . thick curves denote the results for silicates and thin ones are for graphite grains . as expected , the sputtered fraction decreases with increasing @xmath173 , and the fraction for silicates is in general larger than that for graphites . the total dust mass fraction that is destroyed can then be evaluated from these results by convolving @xmath176 with an initial grain size distribution . as mentioned earlier , nozawa ( 2003 ) found that the size distribution of synthesized grains is bounded by two power laws , with indices @xmath33 for larger grains and @xmath32 for smaller grains . we therefore estimate the total dust mass fraction destroyed , assuming a size distribution @xmath178 , with @xmath179 and @xmath180 . the destroyed fraction of dust mass is given by , f = n(a ) a^3 f_d(a ) da n(a ) a^3 da + , where the grain volume scales as @xmath181 . the results for graphites and silicates are presented in table 1 , for @xmath177 ( @xmath71 erg , @xmath144 g ) , leaving the cases of @xmath100 for which the reverse shocks do not reach the interiors of the ejecta within the validity range of self - similar solutions . we assume the lower and upper limit for the integral to be @xmath158 cm and @xmath182 cm respectively , as explained earlier in 2 . we also calculate the destroyed fractions for a truncated grain size distribution , with an upper limit of @xmath35 cm , and the results are shown inside brackets in table 1 for comparison . we have further studied the variation of the destroyed dust mass fraction with explosion energy and ejected mass . tables 2 and 3 show the fraction of dust mass supttered for two sets of parameters : table 2 for @xmath144 g and @xmath183 erg , and table 3 for @xmath184 g and @xmath71 erg . figure 6 plots the varition of ( logarithm ) of the destroyed dust mass fraction , @xmath185 , for silicates , for different values of explosion energy and ejected mass , and for different mass profiles ( @xmath177 ) . in the upper panel , we plot the variation of @xmath185 with @xmath46 ( in logarithm ) , keeping the explosion energy constant at @xmath71 erg , and in the lower palnel , we plot ( logarithms of ) @xmath185 with explosion energy @xmath104 , keeping the ejected mass fixed at @xmath186 g. the destroyed fraction generally increases with ejected mass , but its variation with explosion energy is not monotonic . the reason is that the process of dust destruction depends mainly on two factors : the speed of the reverse shock ( which depends on the characteristic speed @xmath187 ) and the characteristic time , @xmath188 . now , @xmath189 , whereas @xmath190 . it would appear that a change in ejected mass manifests itself in increasing @xmath188 , and thereby increasing the destroyed fraction , whereas , a change in @xmath104 affects @xmath188 and @xmath187 in a complicated manner , which results in the variation shown in the bottom panel of figure 6 . we therefore find that a mass fraction @xmath191 % of silicates and graphite grains created in cooling ejecta can be sputtered away by the reverse shock if the density distribution of the ejecta has a shalow profile ( @xmath192 ) . nozawa et al ( 2007 ) recently studied sputtering of dust in reverse shock and found a destroyed dust mass fraction of @xmath193 ; their result for an ambient hydrogen number density of @xmath194 @xmath20 and a progenitor mass of @xmath195 m@xmath8 is @xmath196 ( their table 1 ) . in fact , their result for the ` unmixed grains ' model ( which has to do with spatial mixing of the core heavy - element distribution ) in which the maximum size is @xmath197 m is @xmath198 . it should be noted , however , that a considerable amount of sputtering in their calculation is caused by the hot plasma between the forward and the reverse shocks ( see 3.3 in nozawa et al ( 2007 ) ) , which we have not considered . in our case , for analytical simplicity , we have assumed the post - reverse shocked gas to be cooling adiabatically . our results are , therefore , expected to give lower bounds on the actual dust mass fraction destroyed in sn remnants . there is another reason why the destroyed fraction of dust mass calculated here is lower than that of nozawa et al ( 2007 ) and bianchi & schneider ( 2007 ) . strictly speaking , the destruction rate of tielens et al ( 1994 ) used here is valid for a gas with solar abundance . the gas in the dust formation region in the supernova ejecta is expected to be metal rich , and the destruction of dust grains in this gas can be more efficient in this case , since sputtering yields by metal ions are much higher in general than by hydrogen and helium ions . it should be noted that our results for the destroyed fraction is independent of the ambient density , since the gas density enters into our calculation only to change @xmath188 ( notice in equations [ eq : char ] that @xmath187 is independent of ambient density , and so the relative speed of reverse shock , and in turn the post - shock temperature , are also independent of ambient density ) . in reality , however , a lot of grain sputtering would take place between the forward and reverse shock , where the gas density will crucially depend on the ambient density . this aspect of grain sputtering is neglected in our analytical calculation , but is captured in the simulation results of bianchi & schneider ( 2007 ) and nozawa ( 2007 ) . another difference between our results and those of bianchi & schneider ( 2007 ) arises from the assumption of initial grain size distribution . bianchi & schneider ( 2007 ) used grains of small size in their calculation , with silicate grains smaller than @xmath199 m and graphites smaller than @xmath200 m . also , their size distribution is more confined to a narrow range than the power law distribution assumed here . both these factors would enhance the destroyed fraction of dust mass in reverse shocks . for example , if the sizes of @xmath201 grains have a delta - function distribution around @xmath202 ( see figure 1 of bianchi & schneider 2007 ) , then the curves in figure 2 of the present paper ( say , for @xmath153 ) yield a mass destruction fraction of @xmath203 % ( with @xmath204 ) . for @xmath205 grains of size @xmath206 , we similarly estimate a destroyed mass fraction of @xmath207 % . from figure 2 , we estimate that one requires @xmath208 in order to get @xmath209% destruction of silicate grains . a note on the observed density distribution of supernovae ejecta is in order here . chevalier & fransson ( 1994 ) studied a model of the ejecta from a type ii supernova with a shallow ( @xmath115 ) core surrounded by a steep ( @xmath110 ) envelope . in the case of type ia supernovae , a steep density distribution in the ejecta mass has also been discussed in the literature ( colgate & mckee ( 1969 ) ; dwarakadas & chevalier ( 1997 ) ) . there has been success , however , in modelling data of type ia supernova with uniformly distributed ( @xmath210 ) ejecta ( hamilton & sarazin ( 1984 ) ; hamilton , sarazin & szymkowiak ( 1986 ) ) . it is worth considering what observations might improve our understanding of the net dust production by sne . far - infrared and sub - millimeter observations of reverse - shocked dust in young sne by _ spitzer _ , _ herschel _ , and other facilities will help , but our theoretical studies suggest that unshocked ( cold ) dust may harbor a considerable mass of undetected dust . to constrain this dust and distinguish it from surrounding interstellar clouds will probably require sub - mm observations with small beam sizes . this brings in complications of the circumstellar environment of young sne . furthermore , the long times ( thousands of years ) required for reverse shocks to reach the core ejecta suggest that late - time observations of supernova remnants would be useful . we have studied the effect of reverse shocks analytically , in the regime of self - similar evolution , on the sputtering of dust grains that are believed to form in cooling ejecta . for representative cases we found that fractions of dust mass that is destroyed are of order of order @xmath211 % of silicates and graphites , for sn of explosion energy @xmath212 erg , ejecta mass @xmath213 g , and an ambient density @xmath214 g @xmath20 . our analytical formalism provides only a lower bound on the dust mass fraction that is destroyed by the reverse shock , since it ignores further sputtering of grains in hot plasma between the forward and reverse shocks . our results are , therefore , consistent with the recent estimates from numerical simulations which include these additional effects . furthermore , our study provides a formalism to generalize these results to cases with different values of sn parameters . we thank drs . r. chevalier , e. dwek , t. nozawa , a. ray and y. shchekinov for helpful discussions . this work was supported at the university of colorado by nasa theory grant nnx07-ag77 g and nsf theory grant ast07 - 07474 . bertoldi , f. , carilli , c. l. , cox , p. , fan , x. , strauss , m. a. , beelen , a. , omont , a. & zylka , r. 2003 , a&a , 406 , l55 bianchi , s. & schneider , r. 2007 , mnras , 378 , 973 catchpole , r. m. et al . , mnras , 229 , 15 chevalier , r. a. 1982 , apj , 258 , 790 colgate , s. a. & mckee , c. 1969 , 157 , 623 deneault , e. a. , clayton , d. d. & heger , a. 2003 , apj , 594 , 312 dunne , l. , eales , s. , ivison , r. , morgan , h. , edmunds , m. 2003 , nature , 424 , 285 dwek , e. 2006 , science , 313 , 178 dwek , e. , & arendt , r. g. 2007 , in aip conf . proc . , supernova 1987a : 20 years after , ed . s. immler , k. weiler , r. mccray , vol . 937 , 58 dwek , e. , & werner , m. w. 1981 , apj , 248 , 138 gerardy , c. l. , fessen , r. a. , hfflich , p. & wheeler , j. c. 2000 , aj , 119 , 2968 gerardy , c. l. 2002 , apj , 575 , 1007 hamilton , a. j. s. & sarazin , c. l. 1984 , apj , 281 , 682 hamilton , a. j. s. , sarazin , c. l. & szymkowiak , a. e. 1986 , apj , 300 , 698 kozasa , t. , hasegawa , h. & nomoto , k. 1989 , apj , 344 , 325 krause , o. , et al . 2004 , nature , 432 , 596 ledoux , c. , bergeron , j. , petitjean , p. 2002 , a&a , 385 , 802 mccray , r. 1993 , ara&a , 31 , 175 mccray , r. 2007 , in aip conf . proc . , supernova 1987a : 20 years after , ed . s. immler , k. weiler , r. mccray , vol . 937 , 3 meikle , w.p.s . , mattila , s. , et al . 2007 , apj , 665 , 608 morgan , h. l. & edmunds , m. g. 2003 , mnras , 343 , 427 moseley , s. h. , dwek , e. , giaccum , w. , graham , j. r. , loewenstein , r. f. , silverberg , r. f. 1989 , nature , 340 , 697 nadyozhin , d. k. 1985 , ap&ss , 112 , 225 nozawa , t. , kozasa , t. , umeda , h. , maeda , k. & nomoto , k. 2003 , apj , 598 , 785 nozawa , t. , kozasa , t. , habe , a. 2006 , apj , 648 , 435 nozawa , t. et al . 2007 , astro - ph/0706.0383 pettini , m. , smith , l. j. , hunstead , r. w. & king , d. l. 1994 , apj , 426 , 79 schneider , r. , ferrara , a. & salvaterra , r. 2004 , mnras , 351 , 1379 seab , c. g. & shull , j. m. 1983 , apj , 275 , 652 shull , j. m. 1978 , apj , 226 , 858 stanimirovic , s. , et al . 2005 , apj , 632 , l103 sugerman , b. e. k. et al . 2006 , science , 313 , 196 tielens , a. g. g. m. , mckee , c. f. , seab , c. g. & hollenbach , d. j. , 1994 , apj , 431 , 321 todini , p. & ferrara , a. 2001 , mnras , 325 , 726 truelove , j. k. & mckee , c. f. 1999 , apj , 120 , 299 ( tm99 ) whittet , d. c. b. 1992 , _ dust in the galactic environment _ , iop publishing , great britain .

intermolecular vibrational energy transfer occurs in a quite large number of chemical processes ranging from simple chemical reactions to very complicated molecular systems like molecular motors . it has been the subject of numerous publications @xcite . among the large variety of experimental techniques used to investigate intramolecular or intermolecular energy transfer , two important spectroscopic methods have been quite successful these last decades . the first one is based on the infrared pump - probe spectroscopy technique . it enables an indirect measurement of the intermolecular energy transfer by bleaching the population at the pumping laser frequency and testing the subsequent ground state recovery by a time delayed probe pulse . a typical situation is the one reported by bakker _ @xcite on the study of vibrational energy transfer , after excitation of the c - h stretching mode in chloroform and bromoform dissolved in polar and non - polar solvent , where an important increase of vibrational energy transfer resulting from the molecule - solvent dipole - dipole interaction has been observed . similar conclusions have been reached for resonant intermolecular vibrational energy transfer in liquid water from rotational anisotropy measurements @xcite . the second archetype of experiments takes advantage of the proportionality between the time - dependence of the raman active transition and the population time dependence to apply infrared pump coupled to anti - stokes - raman - probe technique to study vibrational energy redistribution in polyatomic liquids such as water , deuterated water and methanol @xcite . today , ultrafast two - dimensional infrared ( 2d - ir ) spectroscopy which has been used for more than a decade to study dynamics and structures of large molecular complexes @xcite , charge transfer @xcite , chemical reactions @xcite as well as vibrational coupling and vibrational energy relaxation @xcite , offers an interesting method to directly probe the intermolecular vibrational energy transfer @xcite . in addition , it does not require the probe vibrational mode to be involved in the intermolecular vibrational energy transfer and can display interactions and dynamics that are not accessible by traditional linear infrared vibrational absorption measurements . in a recent experimental work , et al _ @xcite have used 2d - ir spectroscopy to directly probe the intermolecular vibrational energy transfer between benzonitrile and acetonitrile-@xmath0 molecules . using an experimental setup based on three transform - limited ir femtosecond laser pulses and heterodyne detection to phase - resolve and to amplify the vibrational echo signal , they directly probe the vibrational energy transfer between the two molecules . in these 2d - ir spectroscopy experiments , three different types of variables are involved . the first one is the emitted vibrational echo frequency . the second one is the delay time between two laser pulses , most frequently the first and the second ones . it will be numerically fourier transformed for each vibrational echo frequency to give a second frequency mapping . finally , the third one corresponds to the time of delay between the two last laser pulses . it is just due to this last parameter that the cross peak , which is the signature of the intermolecular energy transfer , appear in 2d - ir spectra . when the molecular vibrational states participating in the transfer of energy are dark states , most probably made of combination states of low frequency modes with very weak dipole moments or not at all optically active , coherence transfer between these combination states and an optically active vibrational state induces a cross - peak which becomes the signature of the energy transfer process . it is the intention of this work to describe the appearance of this cross - peak and to analyze in detail the influence of the dynamical constants on its magnitude . in sec.2 , we introduce the vibrational structure of the individual molecules participating in the energy transfer . then , the energy level structure of the global system undergoing the dynamics induced by the femtosecond laser pulses is built in terms of the physical constants of the individual molecules . in sec.3 , we describe the general dynamics induced by the three femtosecond laser pulses and resulting from the relaxation , dephasing , and coherence transfer processes . then , the 2d - ir spectra will be evaluated and discussed according to the roles of the physical constants . finally , in the last section , numerical simulations exhibit the main features which characterize the identification of the energy transfer process through the appearance of the cross - peak observed experimentally . the physical process underlying the 2d - ir vibrational spectrum of a molecular system is based on the interaction of three delayed infrared pulses with the molecular system under investigation . as usual , the interaction hamiltonian of the laser - system interaction is given by @xmath1 \label{2.1}\end{aligned}\ ] ] where the notation c.c . stands for the complex conjugate part . the symbol @xmath2 stands for the normalized envelops of the three laser pulses defined as @xmath3 with @xmath4 . this is a convenient form for later analytical evaluation and its time - integrated form is physically acceptable and has been used extensively . all the laser pulses are centered at @xmath5 with a time duration of @xmath6 and with their amplitudes arbitrarily fixed . as usual , @xmath7 is the dipole moment operator and the quantities @xmath8 and @xmath9 are standard notations for the frequency and wave vector of the field @xmath10 , respectively . it is worthy of note that in time - domain nonlinear spectroscopy , the pulse delay time @xmath11 are retained in the pulse envelops only , because they act just like arbitrary factors @xmath12 and can be compensated by the phase of the local field oscillator for heterodyne detected signals or just cancel in homodyne detection . to analyze 2d - spectra , it is necessary to perform a fourier transform over pulse delay time . the above mentioned factors have to be retained . moreover , for 2d - ir experiments , it is convenient to use the photon - echo geometry . therefore , it is also adopted in our theoretical calculation . it is well established that photon - echo and many other related nonlinear optical processes @xcite are described by the third - order perturbation term of the density matrix with respect to the laser - molecule interaction @xmath13 defined previously by eq.([2.1 ] ) . then , the contribution to the third - order term of the density matrix , @xmath14 , that is relevant to the three - pulse process discussed here , takes the form @xmath15 where the interaction liouvillian is defined by @xmath16 $ ] . the various @xmath17 are free evolution liouvillians of the vibrational system , corresponding to @xmath18 if @xmath19 $ ] with @xmath20 the free vibrational model hamiltonian . the evolution liouvillian @xmath21 account for the evolution of the populations from @xmath22 . the other two evolution liouvillians , namely @xmath23 with @xmath24 and @xmath25 with @xmath26 and @xmath27 account for the evolution of the coherences . notice that the last liouvillian describes coherence transfer between the quasi - resonant transitions @xmath28 and @xmath29 . the emitted signal along the rephasing phase - matched direction @xmath30 and the nonrephasing direction @xmath31 is deduced from the polarization @xmath32 \label{2.3}\end{aligned}\ ] ] where @xmath33 stands for @xmath34 or @xmath35 . heterodyne detection of nonlinear signals has been discussed extensively by joffre _ et al _ @xcite . and will be introduced to amplify the emitted signal . to this end , an appropriate local field oscillator is introduced @xmath36 \label{2.4}\end{aligned}\ ] ] where the various constants are similar to the ones introduced for the exciting laser fields and @xmath37 is an additional phase of this local field . also , we use the relation between polarization @xmath38 and signal field @xmath39 obtained from maxwell equations @xcite , say @xmath40 to simulate the experimental procedure where the fourier transform and next the squared magnitude are performed successively , the detected heterodyne signal intensity can be expressed as @xcite @xmath41 using heterodyne detection . if , as usual , the intensity of the signal field is negligible and the one of the local field is subtracted from the total intensity , then @xmath42 three conditions were imposed according to the assumptions made by lepetit _ et al _ @xcite . first , the laser pulses present clean fast leading edges . second , time origin can be chosen for convenience , and here it is fixed at the center of the last pulse . third , the signal field obeys the causality principle . these conditions imply that @xmath39 is negligible for @xmath43 so that the fourier transform with respect to time can be performed over positive time only . notice that , all high frequency terms are neglected . for our purpose , the evaluation of the 2d - ir spectrum requires the calculation of the double fourier transform over the delay time @xmath44 between the pulses and the experimental time @xmath45 . in the first fourier transform over @xmath46 , the delay time @xmath46 changes sign . therefore , the experimental waiting time @xmath47 has to be defined as @xmath48 , according to the prescription of jonas @xcite . since the center of the third laser pulse is fixed at the time origin , the 2d - ir spectra can be obtained as @xmath49 and this is the basic quantity for evaluating any 2d - ir spectra . the formal expression of the third - order polarization in the phase - matched directions @xmath50 or @xmath35 , required for simulating any 2d - ir experiment , is given in the basis set @xmath51 of the total vibrational molecular system . from eq.([2.2 ] ) , the density matrix elements can be written as @xmath52 each @xmath53 specifies a particular pathway in the liouvillian space of the total molecular system contributing to the density matrix elements @xmath54 . the symbol @xmath55 stands for the summation over the various combinations of fields . these combinations must satisfy the phase - matching condition in the framework of the rotating wave approximation . that is , only the combinations of field components satisfying the secular approximation are retained . in addition , the symbol @xmath56 means that only the values of @xmath53 associated with the particular density matrix element @xmath54 must be included . at least for markovian systems , the general mathematical structure of @xmath57 is of the type @xmath58 which enables a great simplification for further time integrations . the amplitudes @xmath59 and the exponential arguments @xmath60 , @xmath61 , @xmath62 , and @xmath63 will all be obtained by identification of eq.([2.9 ] ) with all the compatible combinations of the product @xmath64 of matrix elements deduced from eq.([2.2 ] ) . each combination with specified indices @xmath65 is associated with a specific pathway @xmath53 . this implies that each particular value of @xmath53 stands for one particular combination of @xmath65 . all of the pathways will be evaluated in the next section according to our specific model . at this stage , the different matrix elements of liouvillian evolution operator participating in free evolution have to be evaluated . notice that the matrix elements of the free evolution liouvillian of the type @xmath23 are diagonal in the coherence liouvillian subspace and have the general form @xmath66 where , as usual , @xmath67 . meanwhile , other matrix elements of the type @xmath68 and @xmath69 with @xmath70 and @xmath71 need to be evaluated based on a specific model . this evaluation will be done in the next section based on a specific model of vibrational energy exchange taking place between two different molecules . the molecular model undergoing intermolecular vibrational energy transfer between two molecules through a process of combination band absorption @xcite and occurring among low frequency vibrational modes , can be described by the hamiltonian @xmath72 where each individual molecular hamiltonian @xmath73 is made of two parts , namely , an anharmonic vibrational mode @xmath74 and a low frequency vibrational modes @xmath75 . the particular system studied in the following is the complex of acetonitrile-@xmath0 ( a ) and benzonitrile ( b ) , and the combination bands are most probably identified as the c - d bending mode in the deuterated methyl group of a and the benzene ring breathing mode in b @xcite . the relevant parameters of the individual molecules including energy levels and nonradiative transition rate constants , as well as the transition dipole moments @xmath76 of the individual molecules with corresponding transition energies pertaining to the 2d - ir spectral range investigated , are shown in fig.[fig1 ] . then , according to the sum rules ( a ) and benzonitrile ( b ) . the various transition constants @xmath77 associated with the individual molecules are indicated by the single - headed arrows in the figure . the allowed dipole moments are indicated by double - headed arrows . the sets of states @xmath78 and @xmath79 stand for the corresponding combination bands of their respective low frequency modes . ] @xmath80 and the definitions of the dephasing constants @xmath81 for the individual molecules , we can evaluate the various dynamical parameters of the total system . as usual , @xmath82 , @xmath77 , @xmath83 and @xmath84 stand for the total decay rates , the transition constants , the dephasing and the pure dephasing constants of the individual molecules . for the total system , the corresponding quantities are denoted @xmath85 , @xmath86 , @xmath87 and @xmath88 , respectively . notice that for small energy gaps , temperature effects can be efficient so that direct and reverse transition rate constants , related by detailed balance , are introduced . for the @xmath89 transitions , we have the temperature dependence @xmath90 and similar relation exists for the @xmath91 and @xmath92 transitions . and dipole moments associated to the total system are indicated using the same symbols as in fig.[fig1 ] . the set of states of the total vibrational system are defined at the bottom of the figure . ] we still have to introduce the coherence transfer constants @xmath93 . they introduce a coupling between the coherences associated to the transitions @xmath94 and @xmath95 . the phenomenon of coherence transfer is now well established experimentally and has been observed in a wide range of systems including photodissociation of triatomic molecules @xcite , photobiological process in light harvesting compounds @xcite or temperature - dependent spontaneous raman spectra of high - frequency c - h vibrations in microwave rotational spectra @xcite . moreover , it has also been considered theoretically @xcite . in the present system , the more important consequence of coherence transfer , especially between closely - lying vibrational energy levels , is the ability of observing dark transitions appearing as extra - peaks in the 2d - ir spectra . all of the vibrational states of the total molecular system are described at the bottom of fig.[fig2 ] where their corresponding dipole moments and transition constants are presented . from the laser - molecule interactions among the set of states ranging from @xmath96 to @xmath97 and their corresponding relaxation constants , the various pathways in the total liouvillian space participating in the dynamical evolution can be determined . as usual , only pathways satisfying the rotating wave approximation ( rwa ) will be retained . the individual pathways are given in appendix a and their corresponding contributions to the various density matrix elements can be obtained from eq.([2.2 ] ) , from the knowledge of the coherence evolution previously established , and also from the population and coherence transfer evolutions that need to be evaluated . this has to be done according to the specific vibrational model of energy transfer . the evolution of the populations is straightforwardly deduced from the zeroth - order liouville equation @xmath98 where , in the present model , the relaxation liouvillian takes the form @xmath99 } \;&\;\ssc{0}\;&\;\ssc{{\gamma}_{2233}}\;&\;\ssc{{\gamma}_{1'1'2'2'}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{22ll}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{11ll}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{1'1'l'l'}}\;\ssc{0}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[3,3,3,3]}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{1122}}\;&\;\ssc{{\gamma}_{2'2'3'3'}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{2'2'l'l'}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{1'1'l'l'}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{11ll}}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[4,4,4,4]}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{22ll}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[5,5,5,5]}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{22ll}}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{2'2'l'l'}}\;&\;\ssc{{\gamma}_{22ll}}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[6,6,6,6]}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{22ll}}\;&\;\ssc{{\gamma}_{2'2'l'l'}}\;&\;\ssc{0}\;&\;\ssc{0}\\ \ssc{0}\;&\;\ssc{{\gamma}_{ll22}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[7,7,7,7]}\;&\;\ssc{{\gamma}_{lll'l'}}\;&\;\ssc{{\gamma}_{1122}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{1'1'2'2'}}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{l'l'2'2'}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{l'l'll}}\;&\;\ssc{\gamma[8,8,8,8]}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{1'1'2'2'}}\;&\;\ssc{{\gamma}_{1122}}\;&\;\ssc{0}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{ll22}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[9,9,9,9]}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{lll'l'}}\;&\;\ssc{0}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{l'l'2'2'}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{\gamma[10,10,10,10]}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{l'l'll}}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{l'l'2'2'}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{l'l'll}}\;&\;\ssc{0}\;&\;\ssc{\gamma[11,11,11,11]}\;&\;\ssc{0}\\ \ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{ll22}}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{0}\;&\;\ssc{{\gamma}_{lll'l'}}\;&\;\ssc{0}\;&\;\ssc{\gamma[12,12,12,12 ] } \end{pmatrix } \nonumber\\ & & \label{3.4}\end{aligned}\ ] ] besides , from our model , the total decay rates are related to the transition rates constants by the expressions given in table [ 1 ] xxx @xmath100=0}$ ] & @xmath100=-\sum_{u=1,7}\gamma[u , u,2,2]}$ ] & @xmath101=-\sum_{u=1,8}\gamma[u , u,3,3]}$ ] + @xmath102=-\sum_{u=2,9}\gamma[u , u,4,4]}$ ] & @xmath103=-\sum_{u=2,3,11,12}\gamma[u , u,5,5]}$ ] & @xmath104=-\sum_{u=3,10}\gamma[u , u,6,6]}$ ] + @xmath105=-\sum_{u=1,2,8}\gamma[u , u,7,7]}$ ] & @xmath106=-\sum_{u=1,3,7}\gamma[u , u,8,8]}$ ] & @xmath107=-\sum_{u=2,4,7,11}\gamma[u , u,9,9]}$ ] + @xmath108=-\sum_{u=6,8,3,12}\gamma[u , u,10,10]}$ ] & @xmath109=-\sum_{u=2,5,8,9}\gamma[u , u,11,11]}$ ] & @xmath110=-\sum_{u=3,5,7,10}\gamma[u , u,12,12]}$ ] + then , by identifying the integral representation of the density matrix with the definition of the population evolution liouvillian , it is found that @xmath111^{-1}_{mmnn}\g{\rho}_{nn}(t_0 ) = \sum_{m}\g{g}_{mmnn}(t - t_0)\g{\rho}_{nn}(t_0 ) \label{3.5}\end{aligned}\ ] ] and the evolution liouvillians in the population subspace are deduced from the diagonalization of @xmath112^{-1}$ ] . they correspond to @xmath113 for the ground state and all the other matrix elements can be expressed in terms of the eigenvalues @xmath114 and the components of their corresponding eigenvectors @xmath115 of @xmath112^{-1}$ ] , so that @xmath116 and @xmath117 . the same processes occurs between transitions @xmath118 and @xmath119 as illustrated in this figure . ] it remains to calculate the evolution liouvillian associated with the coherence transfer processes depicted in fig.[fig3 ] . to this end , we have to diagonalize the evolution liouvillian in the coherence subspace . only , the coherence subspace @xmath120 coupled by the coherence transfer constants need to be considered . as shown in fig.[fig3 ] , they are the @xmath121 and @xmath122 transitions in the first case , as well as the @xmath123 and @xmath124 transitions in the second case . therefore , in the liouvillian subspace where coherence transfer takes place , we have for @xmath70 and @xmath71 @xmath125 now , we have all the required quantities to perform the analytical evaluation and subsequent numerical simulations of the 2d - ir spectra for a system undergoing energy transfer . eq.(3.7 ) indicates that the coherence transfer processes are described by the coherence transfer constants @xmath126 and @xmath127 involving the dark states @xmath128 and @xmath129 of the benzonitrile and acetonitrile-@xmath130 molecules , respectively . therefore , by coupling a dark state to an optically active state , it is possible to visualize on the two - dimensional spectrum the resonance associated with the energy transfer . from the previous sections , the expressions of the liouville evolution operators in the population , coherence and coherence transfer subspaces can be deduced straightforwardly and evaluated for all the matrix elements appearing in the various pathways given in appendix a. once their expressions , taken for a given path , are introduced in eq.([2.2 ] ) and identified with the density matrix element ( [ 2.9 ] ) using the formal expression ( [ 2.10 ] ) for @xmath131 , the various quantities @xmath60 , @xmath61 , @xmath62 , @xmath63 and @xmath59 can be obtained for each individual pathway . the final step to obtain the analytical expression of the 2 d - ir spectrum is the triple time integration plus the double fourier transform over @xmath46 and @xmath45 . they need to be performed for each field ordering , as well as for the two definitions of @xmath46 according to the sign of @xmath132 . their expressions are a little bit cumbersome . for this reason , they will not be presented here . as an example , we give the expression corresponding to the field ordering @xmath133 . other ones , can be evaluated as well . at this stage , it is necessary to introduce all of the numerical values required to perform the numerical simulations from the analytical results describing the vibrational energy exchange between the acetonitrile-@xmath0 and benzonitrile molecules , chosen as the model system . @xmath134&@xmath135&@xmath136&@xmath137 + + @xmath138&@xmath139&@xmath140&@xmath141 + for simplicity , only one combination band state per molecule will be retained in the following simulations . the energies of the acetonitrile-@xmath130 and benzonitrile molecules are given in table [ 2 ] . .numerical values of the total decay rates , transition rate constants and pure dephasing constants of the individual molecules . all of the numerical values are given in units of ps . [ cols="<,<,<,<",options="header " , ] first , we introduce various constants useful for simplify the final expressions @xmath142 then , for the laser field interaction ordering @xmath133 and a positive delay time @xmath143 , there are @xmath144=\;\;\;\frac{c_0^+\e^{-c[n,\alpha , a]t}}{c[n,\alpha , a]+i\omega_d - i\omega_a}-\frac{\e^{-c[n,\alpha , a]t}}{(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_a+\gamma_a)}+\frac{\e^{-c[n,\alpha , a]t}}{(c[n,\alpha , a]+\gamma_a)(i\omega_d - i\omega_a-\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr c_{2,cba}^+[n,\alpha]=-\frac{\e^{-b[n,\alpha , a , b]t}-c[n,\alpha , a]t}{(b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_a+\gamma_a)}-\frac{c_{1,cba}^+[n,\alpha]\e^{-b[n,\alpha , a , b]t}}{b[n,\alpha , a , b]-\gamma_b } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm + \frac{c_0^+\e^{-b[n,\alpha , a , b]t - c[n,\alpha , a]t}}{(b[n,\alpha , a , b]+\gamma_b+c[n,\alpha , a]+i\omega_d - i\omega_a)(c[n,\alpha , a]+i\omega_d - i\omega_a)}+\frac{\e^{-b[n,\alpha , a , b]t - c[n,\alpha , a]t}}{(b[n,\alpha , a , b]+\gamma_b+c[n,\alpha , a]+\gamma_a)(c[n,\alpha , a]+\gamma_a)(i\omega_d - i\omega_a-\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr c_{3,cba,1}^+[n,\alpha]=-\frac{\e^{-\gamma_at-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_a+\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm + \frac{\e^{-\gamma_at-\gamma_bt}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_a+\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm -\frac{c_{1,cba}^{+}[n,\alpha]\e^{-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b ) } + \frac{c_{1,cba}^{+}[n,\alpha]\e^{-\gamma_bt}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm -\frac{c_{2,cba}^+[n,\alpha]}{a[n,\alpha , a , b , c]-\gamma_c}+\frac{c_{2,cba}^+[n,\alpha]}{a[n,\alpha , a , b , c]+\gamma_c } $ } \nonumber\\ & & \mbox{$\scr c_{3,cba,2}^+[n,\alpha]=-\frac{\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t - c[n,\alpha , a]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_a+\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm + \frac{c_0^+\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t - c[n,\alpha , a]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]+\gamma_b+c[n,\alpha , a]+i\omega_d - i\omega_a)(b[n,\alpha , a , b]+\gamma_b+c[n,\alpha , a]+i\omega_d - i\omega_a)(c[n,\alpha , a]+i\omega_d - i\omega_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm + \frac{\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t - c[n,\alpha , a]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]+\gamma_b+c[n,\alpha , a]+\gamma_a)(b[n,\alpha , a , b]+\gamma_b+c[n,\alpha , a]+\gamma_a)(c[n,\alpha , a]+\gamma_a)(i\omega_d - i\omega_a-\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm -\frac{c_{1,cba}^+[n,\alpha]\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b)}-\frac{c_{2,cba}^+[n,\alpha]\e^{-a[n,\alpha , a , b , c]t-\gamma_ct}}{a[n,\alpha , a , b , c]+\gamma_c } $ } \label{b.2}\end{aligned}\ ] ] and the first contribution for positive delay time @xmath132 and field ordering @xmath133 is given by @xmath145=-\sum_{\alpha}tf_{\omega_{lo}}(\omega_t)q[n,\alpha , a , b , c]k_t $ } \nonumber\\ & & \mbox{$\scr\hskip 1.4truecm\times\biggl\lbrack\;\;\ ; \frac{\e^{-\gamma_at-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a)(c[n,\alpha , a]-\gamma_a)}$ } \nonumber\\ & & \mbox{$\scr\hskip 2.4truecm\times \frac{1}{(i\omega_d - i\omega_a+\gamma_a)(k[n,\alpha , a , b , c]+i\omega_t+a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b+c[n,\alpha , a]-\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm + \frac{c_{1,cba}^+[n,\alpha]\e^{-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b)(k[n,\alpha , a , b , c]+i\omega_t+a[n,\alpha , a , b , c ] -\gamma_c+b[n,\alpha , a , b]-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm + \frac{c_{2,cba}^+[n,\alpha]}{(a[n,\alpha , a , b , c]-\gamma_c)(k[n,\alpha , a , b , c]+i\omega_t+a[n,\alpha , a , b , c]-\gamma_c ) } + \frac{c_{3,cba,1}^+[n,\alpha]}{k[n,\alpha , a , b , c]+i\omega_t } + \frac{c_{3,cba,2}^+[n,\alpha]}{k[n,\alpha , a , b , c]+i\omega_t}$ } \biggr\rbrack . \label{b.3}\end{aligned}\ ] ] with the same ordering of the laser pulses , say @xmath133 and a negative delay time @xmath146 , it is found that @xmath147=\;\;\;\frac{\e^{-c[n,\alpha , a]t}}{c[n,\alpha , a]-\gamma_a}-\frac{\e^{-c[n,\alpha , a]t}}{(c[n,\alpha , a]+\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr c_{2,cba}^-[n,\alpha]=\;\;\;\frac{\e^{-b[n,\alpha , a , b]t - c[n,\alpha , a]t}}{(b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_b-\gamma_b)}-\frac{c_{1,cba}^-[n,\alpha]\e^{-b[n,\alpha , a , b]t}}{(b[n,\alpha , a , b]-\gamma_b)(i\omega_d - i\omega_b-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.6truecm + \frac{c_0 ^ -\e^{-b[n,\alpha , a , b]t - c[n,\alpha , a]t}}{(-b[n,\alpha , a , b]-c[n,\alpha , a]-\gamma_a + i\omega_d - i\omega_b)(c[n,\alpha , a ] + \gamma_a)}-\frac{\e^{-b[n,\alpha , a , b]t - c[n,\alpha , a]t}}{(b[n,\alpha , a , b]+c[n,\alpha , a]+\gamma_a+\gamma_b)(c[n,\alpha , a]+\gamma_a)(i\omega_d - i\omega_b+\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr c_{3,cba,1}^-[n,\alpha]=\;\;\;\frac{\e^{-\gamma_at-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_b-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.9truecm -\frac{\e^{-\gamma_at-\gamma_bt}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_b-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.9truecm -\frac{c_{2,cba}^-[n,\alpha]}{a[n,\alpha , a , b , c]-\gamma_c } + \frac{c_{2,cba}^-[n,\alpha]}{a[n,\alpha , a , b , c]+\gamma_c } -\frac{c_{1,cba}^-[n,\alpha]\e^{-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b)(i\omega_d - i\omega_b+\gamma_b)}$ } \nonumber\\ & & \mbox{$\scr\hskip 1.9truecm + \frac{c_{1,cba}^-[n,\alpha]\e^{-\gamma_bt}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b)(i\omega_d - i\omega_b-\gamma_b)}$ } \nonumber\\ & & \mbox{$\scr c_{3,cba,2}^-[n,\alpha]=\;\;\;\frac{\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t - c[n,\alpha , a]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(c[n,\alpha , a]-\gamma_a)(i\omega_d - i\omega_b-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.9truecm -\frac{c_0 ^ -\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t - c[n,\alpha , a]t-\gamma_ct}}{(-a[n,\alpha , a , b , c]-\gamma_c - b[n,\alpha , a , b]-c[n,\alpha , a]-\gamma_a+i\omega_d - i\omega_b)(-b[n,\alpha , a , b]-c[n,\alpha , a]-\gamma_a+i\omega_d - i\omega_b)(c[n,\alpha , a]t+\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.9truecm -\frac{\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t - c[n,\alpha , a]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]+c[n,\alpha , a]+\gamma_a+\gamma_b)(b[n,\alpha , a , b]+c[n,\alpha , a]+\gamma_a+\gamma_b)(c[n,\alpha , a]+\gamma_a)(i\omega_d - i\omega_b+\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.9truecm -\frac{c_{1,cba}^-[n,\alpha]\e^{-a[n,\alpha , a , b , c]t - b[n,\alpha , a , b]t-\gamma_ct}}{(a[n,\alpha , a , b , c]+\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b)(i\omega_d - i\omega_b-\gamma_b)}-\frac{c_{2,cba}^-[n,\alpha]\e^{-a[n,\alpha , a , b , c]t-\gamma_ct}}{a[n,\alpha , a , b , c]+\gamma_c } $ } \label{b.4}\end{aligned}\ ] ] and the second contribution for negative delay time @xmath146 is @xmath148=-\sum_{\alpha}tf_{\omega_{lo}}(\omega_t)q[n,\alpha , a , b , c]k_t $ } \nonumber\\ & & \mbox{$\scr\hskip 0.8truecm\times\biggl\lbrack\;\;\ ; \frac{\e^{-\gamma_at-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(b[n,\alpha , a , b]-\gamma_a+c[n,\alpha , a]-\gamma_b)(c[n,\alpha , a]-\gamma_a ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.8truecm\times \frac{1}{(i\omega_d - i\omega_b-\gamma_b)(k[n,\alpha , a , b , c]+i\omega_t+a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]+c[n,\alpha , a]-\gamma_a-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.3truecm + \frac{c_{1,cba}^-[n,\alpha]\e^{-\gamma_bt}}{(a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b)(b[n,\alpha , a , b]-\gamma_b)(i\omega_d - i\omega_b-\gamma_b ) ( k[n,\alpha , a , b , c]+i\omega_t+a[n,\alpha , a , b , c]-\gamma_c+b[n,\alpha , a , b]-\gamma_b ) } $ } \nonumber\\ & & \mbox{$\scr\hskip 1.3truecm + \frac{c_{2,cba}^-[n,\alpha]}{(a[n,\alpha , a , b , c]-\gamma_c)(k[n,\alpha , a , b , c]+i\omega_t+a[n,\alpha , a , b , c]-\gamma_c ) } + \frac{c_{3,cba,1}^-[n,\alpha]}{k[n,\alpha , a , b , c]+i\omega_t } + \frac{c_{3,cba,2}^-[n,\alpha]}{k[n,\alpha , a , b , c]+i\omega_t } $ } \biggr\rbrack \label{b.5}\end{aligned}\ ] ] and the contribution to the 2d - ir spectrum with field ordering @xmath133 is given by @xmath149=h(t_b - t_a)i_{cba}^+[n,\omega_d,\omega_t]+h(t_a - t_b)i_{cba}^-[n,\omega_d,\omega_t ] \label{b.6}\end{aligned}\ ] ] where @xmath150 stands for the heaviside function . the contributions for any other field ordering can be evaluated as well . 99 a. a. stuchebrukhov and r. a. marcus , j. chem . phys . * 98 * , 8443 ( 1993 ) . z. h. wang and a. pakoulev and d. d. dlott , science * 296 * , 2201 ( 2002 ) . g. c. schatz and r. p. van duyne , in _ handbook of vibrational spectroscopy _ , edited by j. m. chalmers and p. r. griffiths ( wiley , chichester , 2002 ) . r. rey and k. b. moller and j. t. hynes , chem . ( washington d. c. ) * 104 * , 1915 ( 2004 ) . c. g. elles and f. f. crim , annu . chem . * 57 * , 273 ( 2006 ) . g. marcotte and p. ayotte , j. chem . phys . * 134 * , 114522 ( 2011 ) . h. j. bakker and p. c. m. planken and a. lagendijk , nature * 347 * , 745 ( 1990 ) . s. woutersen and h. j. bakker , nature * 402 * , 507 ( 1999 ) . d. d. dlott , chem . phys . * 266 * , 149 ( 2001 ) . m. f. decamp and l. deflores and j. m. mccraken and a. tokmakoff and k. kwak and m. cho , j. phys . chem . * b 109 * , 11016 ( 2005 ) . i. j. finkelstein and j. r. zheng and h. ishikawa and s. kim and k. kwak and m. d. fayer , phys . * 9 * , 1533 ( 2007 ) . m. kozinski and s. garrett - roe and p. hamm , j. phys . b 112 * , 7645 ( 2008 ) . l. w. barbour and m. hegadorn and j. b. asbury , j. phys . b 110 * , 24281 ( 2006 ) . l. w. barbour and m. hegadorn and j. b. asbury , j. am . chem . soc . * 129 * , 15884 ( 2007 ) . j. zheng and k. kwak and j. xie and m. d. fayer , science * 313 * , 1951 ( 2006 ) . c. kolano and j. helbing and m. kozinski and w. sander and p. hamm , nature * 444 * , 469 ( 2006 ) . j. f. cahoon and k. r. sawyer and j. p. schlegel and c. b. harris , science * 319 * , 1820 ( 2008 ) . m. khalil and n. demirdven and a. tokmakoff , j. chem . * 121 * , 362 ( 2004 ) . m. j. nee and c. r. baiz and j. m. anna and r. mccanne and k. j. kubarich , j. chem . phys . * 129 * , 084503 ( 2008 ) . s. r. g. naraharisetty and v. m. kasyanenko and i. v. rubtsov , j. chem . phys . * 128 * , 104502 ( 2008 ) . h. bian and j. li and x. wen and j. zheng , j. chem . phys . * 132 * , 184505 ( 2010 ) . h. bian and w. zhao and j. zheng , j. chem . * 131 * , 124501 ( 2009 ) . h. bian and x. wen and j. li and j. zheng , j. chem . phys . * 133 * , 034505 ( 2010 ) . h. bian and j. li and x. wen and z. sun and j. song and w. zhuang and j. zheng , j. phys . chem . * a 115 * , 3357 ( 2011 ) . h. bian and h. chen and j. li and x. wen and j. zheng , j. phys . chem . * a 115 * , 11657 ( 2011 ) . m. cho and n. f. scherer and g. r. fleming and s. mukamel , j. chem . phys . * 96 * , 5618 ( 1992 ) . y. tanimura and s. mukamel , phys . rev . * e 47 * , 118 ( 1993 ) . v. v. lozovoy and i. pastirk and m. g. comstock and m. dantus , chem . phys . * 266 * , 205 ( 2001 ) . k. f. everitt and e. geva and j. l. skinner , j. chem . phys . * 114 * , 1326 ( 2001 ) . b. i. grimberg and v. v. lozovoy and m. dantus and s. mukamel , j. phys . chem . * a 106 * , 697 ( 2002 ) . v. v. lozovoy and b. i grimberg and i. pastirk and m. dantus , chem . phys . * 267 * , 99 ( 2001 ) . m. khalil and a. tokmakoff , chem . phys . * 266 * , 213 ( 2001 ) . m. khalil and n. demirdven and a. tokmakoff , phys . lett . * 90 * , 047401 ( 2003 ) . l. lepetit and m. joffre , opt . * 21 * , 564 ( 1996 ) . j. p. likforman and m. joffre and v. thierry - mieg , opt . * 22 * , 1104 ( 1997 ) . n. belabas and m. joffre , opt * 27 * , 2043 ( 2002 ) . m. j. burns and w. k. liu and a. h. zewail , in _ spectroscopy and excitation dynamics of condensed molecular systems _ , edited by v. m. agranovich and r. m. hochstrasser ( north - holland publishing company , amsterdam , 1983),chapter 7 . m. cho , in _ two - dimensional optical spectroscopy _ ( crc press , new york , 2009 ) , p. 61 . p. hamm and m. zanni , in _ concepts and methods of 2d infrared spectroscopy _ ( cambridge university press , cambridge , 2011 ) , chapter 4 . l. lepetit and g. cheriaux and m. joffre , j. opt . . am . * b 12 * , 2467 ( 1995 ) . s. m. g. faeder and d. m. jonas , j. phys . chem . * a 103 * , 10489 ( 1999 ) . m. khalil and n. demirdven and a. tokmakoff , j. phys . chem . * a 107 * , 5258 ( 2003 ) . e. gershgoren and e. gordon and s. ruhman , j. chem . phys . * 106 * , 4806 ( 1997 ) . j. m. jean and r. a. friesner and g. r. fleming , j. chem . phys . * 96 * , 5827 ( 1992 ) . s. a. henck and k. k. lehmann , chem . * 144 * , 281 ( 1988 ) . d. w. oxtoby , j. phys . chem . * 87 * , 3028 ( 1983 ) . k. f. freed and a. a. villaeys , chem . * 39 * , 51 ( 1979 ) . p. de bree and d. a. wiersma , j. chem . phys . * 70 * , 790 ( 1979 ) . y. ohtsuki and y. fujimura , j. chem . phys . * 104 * , 8321 ( 1996 ) .

crystalline samples of cold atoms , now known as optical lattices , were initially investigated in the dissipative regime , as a tool to provide velocity damping , and hence a reduction of the kinetic energy of the atomic samples ( for reviews see @xcite ) . in fact the sub - doppler cooling regime relies on the action of laser beams on the atomic motion in a standing wave configuration . the study of that regime also implied the possibility of trapping atoms in the sub - wavelength sized potential wells created by the laser beams . as soon it was clear that the standing wave patterns created by several intersecting laser beams provided by low - power diode lasers could be used to trap atoms in periodic structures , the field exploded . for instance a large experimental effort was made to probe directly the bound states of the atoms within the optical potential . after the initial experiments using one - dimensional ( 1d ) lattices , several schemes were developed allowing an extension to two- and three - dimensions ( 2d and 3d ) . very soon optical lattices were used as a flexible tool to modify the spatial periodicity of the cold atomic samples , and in some cases fancy spatial structures could be produced which the solid state physics community could only dream of . moreover , important applications for atomic nanolithography have been realized . later , the research effort moved into the nondissipative , or conservative , regime , with the aim of reducing the scattering rate in the optical potential wells which ruled out coherent control over the wave - packet atomic motion . in fact , the interest in optical lattices shifted to using them as a test - bed for quantum mechanics . such a shift in interest was enhanced when ultracold atomic samples represented by quantum degenerate gases were available for loading into the optical periodic potential . bose - einstein condensates ( bec s ) represent flexible sources whose spatial dimensions and velocity spread can be controlled with large freedom , so that a condensate may be loaded with great accuracy into the periodic potential created by intersecting laser beams . thus the study of conservative optical lattices used to modify the spatial macroscopic wavefunction of bec s has greatly expanded in the last few years . in this work we will report on the most important aspects of the interaction between ultracold atoms , above and below the bec temperature , within optical lattices . we will concentrate on those features more directly connected with the long - range interactions within the optical lattice . we will not , therefore , discuss some exciting investigations on the dynamical tunneling and chaotic behavior for atoms located within an optical periodic potential whose amplitude is periodically or randomly modulated @xcite . furthermore , we will not discuss the use of optical lattices in quantum computation schemes @xcite . moreover the subject of self generated periodic spatial structures will not be covered because separately treated in this book @xcite . section [ basicaa ] will introduce the basic notions on the creation of optical lattices and the atomic response within the lattice , discussing first the near - resonant optical lattices that evolved from laser cooling schemes and then the non - dissipative , far - detuned optical lattices . section [ ultracold ] briefly reviews some of the experiments on ultracold atoms in optical lattices carried out since 1992 , whilst section [ condensed ] deals with the more recent experiments in which bose - einstein condensates within optical lattices have been explored . the experimental results obtained by the pisa group on bloch oscillations , landau - zener tunneling and optical potential renormalization are there reported . section [ conclude ] concludes the presentation . between the lattice beams can be used to create a moving or uniformly accelerated lattice . ] the simplest possibility to create a periodic potential for neutral atoms is to exploit the light - shift experienced by the atoms in a spatially modulated light field . in one dimension , this can be achieved by superposing two linearly polarized , counter - propagating laser beams with parallel or perpendicular polarizations ( see fig . [ latticegeometries](a ) ) . if the polarizations of the two laser beams are perpendicular to each other ( see fig . [ latticeprinciple ] ) , an atom with two magnetic sub - levels in the ground state will see two interleaved standing waves of @xmath0 and @xmath1 circularly polarized light . this so - called @xmath2 configuration was typical of the early experiments on optical lattices , as it provided both localization of the atoms at the troughs of the potential wells and a sub - doppler cooling mechanism ( `` sisyphus - cooling '' , in which atoms are preferentially pumped from a sub - level with locally high potential energy to the other sub - level with a potential minimum at that point , thus reducing the kinetic energy of the atoms as shown in fig . [ sisyphus ] . ) . for this combination of effects to work , the laser beams creating the optical lattice were detuned by a few natural linewidths from the atomic resonance ( near - resonant optical lattices ) . in most configurations , the detuning was to the red side of the resonance , resulting in the atoms being trapped at the antinodes of the standing wave creating the lattice . configuration . ] the 1d periodic structure created by two - counter - propagating laser beams can be generalized to two and three dimensions in different ways . theodor hnsch and his group in munich used orthogonal pairs of beams whose relative phases were stabilized to create a constant lattice geometry @xcite . a simpler approach was pioneered by gilbert grynberg s team at the ens in paris @xcite . figure [ latticegeometries](b ) shows their beam geometry used to create a 3d lattice . in this setup , no active phase - control is necessary as a variation in phase of the beams will only lead to a spatial translation of the lattice without changing its intrinsic geometry . in 2d and 3d lattices , the arrangement of lattice wells can either be anti - ferromagnetic , with adjacent wells having orthogonal circular polarizations ( as in the 1d example shown above ) , or ferromagnetic , in which case adjacent wells have the same circular polarization . by changing the splitting angle @xmath3 between the two pairs of beams , the distance between neighbouring lattice wells in the ens setup could be varied . this possibility has later also been used for experiments on becs . in a 1d optical lattice with angle @xmath3 , the distance between neighboring wells ( lattice constant ) @xmath4 can be varied through the angle @xmath3 between between the two laser beams creating a lattice with @xmath5 where @xmath6 is the laser wavenumber . after the exploration of the properties of near - resonant lattices , the research effort was concentrated on far - detuned lattices with detunings ranging from hundreds to thousands of linewidths , using both ultra - cold atoms and becs . in these optical lattices the dissipative cooling mechanisms are not active and the optical lattice is described by a conservative potential . for a 1d lattice configuration as in fig.[latticegeometries ] with the two counterpropagating laser beams having the same polarization , the ac - stark shift created by an off - resonant interaction between the electric field of the laser and the atomic dipole results a potential of the form @xmath7 where @xmath8 is the depth of the potential and @xmath4 the lattice constant . in a lattice configuration in which the two laser beams with wavevector @xmath9 are counter - propagating , the usual choice of units are the recoil momentum @xmath10 and the recoil energy @xmath11 . in the case of an angle - geometry , it is more intuitive to base the natural units on the lattice spacing @xmath4 and the projection @xmath12 of the laser wavevector @xmath9 onto the lattice direction . however , the spatially periodic external potential leads naturally to a solid state physics approach . one can then define a bloch momentum @xmath13 corresponding to the full extent of the first brillouin zone @xcite or , alternatively , to the net momentum exchange in the lattice direction between the atoms and the two laser beams . in that frame of reference a possible choice for the energy unit is the bloch energy defined as @xmath14 . these units can also be used for the case of the angle - geometry , making use of eq . ( [ angle ] ) for the connection between @xmath4 and @xmath3 . by introducing a frequency difference @xmath15 between the two beams , the lattice potential of eq . ( [ potential ] ) can be moved at a constant velocity @xmath16 given by @xmath17 or accelerated with an acceleration @xmath18 given by @xmath19 the first experiments on optical lattices were aimed at local properties of the atoms trapped in the wells . making a harmonic approximation to the potential at the centre of each well , it is straightforward to calculate the harmonic trapping frequency for a one - dimensional lattice as @xmath20 in a typical optical lattice experiment , atoms were first trapped and laser - cooled in a magneto - optical trap ( mot ) . after further cooling using optical molasses ( essentially by switching off the magnetic fields of the mot and increasing the detuning of the trap beams ) , the lattice laser beams were switched on . in order to demonstrate that the atoms were truly localized in the potential wells of the lattice , a series of experiments was carried out using pump - probe techniques @xcite . by propagating a probe beam through the optical lattice that had a variable detuning from the lattice beams , raman resonances could be identified ( see fig . [ raman ] ) . these corresponded to an atom absorbing a photon from a lattice beam and emitting another into the probe beam ( or vice versa ) whilst changing its vibrational quantum number by unity . the measured positions of the raman frequencies agreed with the calculated vibrational frequencies . these experiments were helped by the fact that due to lamb - dicke suppression of inelastic scattering , the raman resonances were narrow enough to be individually resolvable . axis . ] other experiments investigating the motion of the atoms in the lattice wells included the observation of breathing oscillations of the atomic wavepackets when then the well - depth was suddenly changed @xcite as well as the collapse and revival of coherent wavepacket oscillations induced by rapidly shifting the lattice in space @xcite . in both experiments , the wavepacket motion was inferred from the fluorescence light as the re - distribution of photons between the lattice beams due to the motion of the atoms led to intensity fluctuations that could be experimentally detected . stimulated revivals were observed in 2000 by the lattice group in hannover @xcite . information about coherence times of the atomic motion in optical lattices can also be obtained by creating coherent transients @xcite . in this method , coherences between vibrational states are first created by a short probe pulse . subsequently , during the lifetime of the coherence lattice photons will be preferentially scattered in the direction of the probe beam ( now switched off ) . from the spectrum of these photons , vibrational frequencies and coherence times can be extracted @xcite . the nick - name `` crystals bound by light '' that was invented for optical lattices soon after their first experimental realization highlighted a property of these physical systems that was not visible in the early experiments , namely their periodic spatial structure . if optical lattices were , as predicted , `` egg cartons '' for atoms , then some direct evidence to that effect was desirable . an obvious experiment to carry out was bragg - scattering of a probe beam off the planes of the `` crystal '' , and in 1995 the groups in munich and gaithersburg @xcite managed to do just that ( see fig . [ bragg ] ) . they were able to show that when the lattice beams were switched on , long - range spatial order was built up as the atoms were further cooled and localized in the periodically arranged potential wells of the lattice . when the lattice was switched off , the spatial order was lost on a time - scale consistent with the thermal motion of the atoms . by measuring the structure factor associated with the bragg reflection , it was also possible to obtain information about the spatial localization of the atoms in the lattice wells @xcite . another global property of near - resonant optical lattices that was accessible to experimental verification was their response to external magnetic fields . using an optical lattice with anti - ferromagnetic order of the atomic spins , the ens group showed that by applying a magnetic field that differentially shifted the potential energies of the @xmath0 and @xmath1 wells , a redistribution of population between the two sets of wells resulted in a macroscopic magnetization of the lattice @xcite . the entire lattice , therefore , exhibited paramagnetic behaviour . in near - resonant optical lattices , only a few percent of all the lattice sites actually contain an atom . despite the consequent lack of atom - atom interactions , it was shown in 1996 that propagating excitations similar to sound waves can be created in optical lattices @xcite . through cycles of half - oscillations in the potential wells and subsequent optical pumping into the other magnetic sub - level of the ground state , propagating density modulations were created that could be detected via pump - probe spectroscopy . moreover the first direct observation of brillouin - like propagation modes in a dissipative periodic optical lattice was recently reported @xcite . whilst so far we have stressed the regular , periodic structure of optical lattices , a number of interesting experiments have also been carried out in a quasi - periodic geometry . by adding an additional laser beam to the four - beam geometry described above , a quasi - periodic lattice structure can be created . the diffusion properties in such a lattice were studied by the group of philippe verkerk @xcite . the fact that in a red - detuned , near resonant lattice the atoms are trapped at local maxima of the light intensity means that the coherent motion of the atomic wavepackets is frequently interrupted by spontaneous emission events . early on in the development of optical lattices , schemes were developed that would provide trapping in locally `` gray '' or `` dark '' states in which the atoms absorb very few photons or no photons at all . these schemes were , appropriately , called `` gray '' or `` dark '' optical lattices @xcite . in order to get rid completely of dissipative effects , it was , however , necessary to increase the detuning of the lattice beams from the atomic resonance and , at the same time , the beam intensity in order to keep a fixed lattice depth . as the spontaneous scattering rate decreases more rapidly with detuning than the light - shift or dipole force providing localization , in this manner optical lattices could be realized in which for the duration of the experiment ( usually on the order of tens to hundreds of milliseconds ) virtually no photons were scattered . the lack of photon scattering , on the other hand , also ruled out the sisyphus cooling mechanism present in near - resonant lattices . in order further to cool atoms transferred into a far - resonant lattice , resolved - sideband raman cooling was successfully employed by several groups @xcite . as in trapped ion experiments , transitions towards lower vibrational levels were induced using a combination of laser beams and magnetic fields . in this way , a large fraction of the atoms could be cooled to the ground vibrational state of the lattice . another possibility to achieve the loading of atoms into the ground vibrational state ( or , for shallow lattices , the ground state bloch band ) is to start from a bose - condensed cloud of atoms , as will be discussed in the second part of this chapter . apart from the intrinsic interest in the characteristics of optical lattices , a number of experiments have used optical lattices as a tool for creating a specific quantum system whose properties could then be studied . especially the precise experimental control of the lattice parameters ( well depth , geometry ) means that optical lattices can be used as a construction kit for quantum systems . in a beautiful experiment @xcite , the group of christophe salomon at the ens in paris demonstrated bloch oscillations of atomic matter waves delocalized over many lattice sites of a shallow optical lattice by accelerating the lattice and measuring the resulting velocity of the atoms . related experiments by mark raizen s team in austin investigated landau - zener tunneling and wannier - stark ladders @xcite . in an extension to this work , they were able to experimentally verify the non - exponential nature of the initial decay of a quantum system @xcite and the associated zeno and anti - zeno effects exhibited by such systems when they are subject to frequent observations @xcite . in a near - resonant lattice , gilbert grynberg s group at the ens was able to create an asymmetric optical lattice with a ratchet - like potential that converted the random thermal motion of the atoms into directed motion @xcite . while in most of the original optical lattice experiments the atomic clouds had temperatures in the the micro - kelvin range , corresponding to a few recoil energies of the atoms , atomic samples with sub - recoil energies are now routinely produced in bose - einstein condensation experiments . since the first experimental realizations in 1995 , many aspects of bose - einstein condensed atomic clouds ( becs ) have been studied @xcite , ranging from collective excitations to superfluid properties and quantized vortices . the properties of becs in periodic potentials constitute a vast new field of research initially explored in @xcite . several experiments have made use of the periodic optical potential produced by a pulsed standing wave to manipulate the condensate or to explore its properties @xcite . in the following , we will concentrate on studies of the condensate within the periodic optical lattice . the first step in that direction was taken by the investigation of the tunneling of becs out of the one - dimensional potential wells of a shallow optical lattice in the presence of gravity @xcite . more recently the phase properties of the condensate wavefunction occupying the whole optical lattice have involved such intriguing concepts as number squeezing @xcite and the mott insulator transition @xcite . the tunneling of the condensate between neighboring wells , controlled by varying the optical lattice potential depth , determine the overall properties of the macroscopic wavefunction . thus the condensate response within a 1d optical lattice can be described as an array of tunneling junctions , as pointed out in @xcite and later explored in @xcite in connection with the superfluid properties of the condensate wavefunction . coherent acceleration of becs adiabatically loaded into optical lattices was demonstrated in @xcite , with bloch oscillations observed for small values of the lattice depth , while a landau - zener breakdown occurred when the lattice depth was further reduced and/or the acceleration increased . the expansion of the condensate array was explored initially in @xcite and later in @xcite . the high level coherent control over this artificial solid state system was demonstrated in @xcite , where the bec was carefully loaded into the lattice ground state by adiabatically turning on the optical lattice . the different dependence of the condensate population on the temperature for the 1d optical lattice was pointed out by ref . @xcite . most experiments were based on the production of the condensate through the standard technique , followed by an adiabatic load into the optical lattice . often the magnetic trap was switched off when the optical lattice was switched on . however , a larger condensate density is realized when the interaction between the condensate and the lattice takes place inside the magnetic trap , and both of them are subsequently switched off to allow time - of - flight imaging . the florence group @xcite has developed a different approach by producing the condensate directly inside the optical lattice which is adiabatically loaded during the evaporating cooling stage . the main advantage of this approach is that the phase coherence of the condensate over the whole optical lattice structure is built up during the condensate formation process . the main disadvantage is that the optical lattice is on during the whole evaporation process , and the spontaneous losses produced by the optical lattice should be heavily reduced by further increasing the detuning of the optical lattice lasers . while a large majority of the experiments have loaded the condensate into 1d optical lattices , experiments on 2d lattices were performed by greiner _ @xcite , and the mott - hubbard transition of ref . @xcite was realized in a 3d optical lattice . if the momentum spread of the atoms loaded into an optical lattice structure is small compared to the characteristic lattice momentum @xmath21 , then their thermal de broglie wavelength will be large compared to the lattice spacing @xmath4 and will , therefore , extend over many lattice sites . a description in terms of a coherent delocalized wavepacket within a periodic structure is then appropriate and leads us directly to the bloch formalism first developed in condensed matter physics . in the tight - binding limit ( @xmath22 ) , the condensate in the lattice can be approximated by wavepackets localized at the individual lattice sites ( wannier states ) . this description is more intuitive than the bloch picture in the case of experiments in which the condensate is released from a ( deep ) optical lattice into which it has previously been loaded adiabatically . in bose - einstein condensates , interactions between the constituent atoms are responsible for the non - linear behavior of the bec and can lead to interesting phenomena such as solitons @xcite and four - wave mixing with matter waves @xcite . as the atoms are extremely cold , collisions between them can be treated by considering only @xmath23-wave scattering , which is described by the scattering length @xmath24 . modeling the interatomic interaction as hard - core collisions , one can simplify the treatment using a mean - field description which leads to the famous gross - pitaevskii equation @xcite , the validity of which has been demonstrated in numerous experiments . for a bec in an optical lattice , one expects an effect due to the mean - field interaction similar to the one responsible for determining the shape of a condensate in the thomas - fermi limit : the interplay between the confining potential and the density - dependent mean - field energy leads to a modified ground state that reflects the strength of the mean - field interaction . applied to a bec in a periodic potential , one expects the density modulation imposed on the condensate by the potential ( higher density in potential troughs , lower density where the potential energy is high ) to be modified in the presence of mean - field interactions . in particular , the tendency of the periodic potential to create a locally higher density where the potential energy of the lattice is low will be counteracted by the ( repulsive ) interaction energy that rises as the local density increases . the description of a condensate in a 1d array of coupled potentials wells is based on the total hamiltonian @xmath25 with the interaction parameter @xmath26 given by @xmath27 and @xmath28 the condensate wavefunction at position @xmath29 . as the interaction term is expected to distort the band structure of the condensate in the lattice @xcite , it should affect all measurable quantities ( rabi frequency , amplitude of bloch oscillations , and tunneling probability ) . in ref . @xcite , the authors derived an analytical expression in the perturbative limit ( assuming @xmath30 ) for the effect of the mean - field interaction on the ground state of the condensate in the lattice . starting from the gross - pitaevskii equation for the condensate wavefunction in a one - dimensional optical lattice ( _ i.e. _ a one dimensional hamiltonian equivalent to that of eq . ( [ hamiltonianaa ] ) ) , they found that the effect of the mean - field interaction could be approximately accounted for by substituting the potential depth @xmath8 with an effective potential @xmath31 with the dimensionless parameter @xmath32 given by @xcite @xmath33 corresponding to the ratio of the nonlinear interaction term and the bloch energy . the @xmath32 parameter contains the peak condensate density @xmath34 , the scattering length @xmath35 , and the atomic mass @xmath36 . from the dependence of @xmath32 on the lattice angle @xmath3 it follows that a small angle @xmath3 ( meaning a large lattice constant @xmath4 ) should result in a large interaction term @xmath32 . the reduction of the effective potential given by eq . ( [ potentialred ] ) agrees with the intuitive picture of the back - action on the periodic potential of the density modulation of the condensate imposed on it by the lattice potential . for repulsive interactions , this results in the effective potential being lowered with respect to the actual optical potential created by the lattice beams . the properties of bose - einstein condensates within optical lattices have been examined in a large number of theoretical papers , predicting a variety of phenomena , often making use of the strict analogies with cases previously studied within the context of solid state physics and nonlinear dynamics . we will list here the principal research lines that have characterized this research so far . the collective excitations of the condensate within optical lattices , and their probe , have been determined @xcite , and those analyses have stimulated the experimental investigations in refs . a strong deformation of the bloch energy bands of the condensate produced by the nonlinear atomic interactions has been predicted in refs . different mechanisms of breaking down the bloch oscillation , all of them connected to the interatomic interactions , have been discussed in refs . @xcite . the thermal and quantum decoherence for an array of multiple condensates within an optical lattice , introduced in refs . @xcite , are an important issue requiring more detailed studies , both theoretically and experimentally . the bose - hubbard - hamiltonian for atoms in an optical lattice , introduced by jaksch _ et al . _ @xcite , was analyzed in ref . @xcite through a mean - field approximation generalization of the bogoliubov approach ; later it was applied to determine the conditions for the number squeezing in that transition @xcite . a number of papers have pointed out the existence of additional solutions for the evolution of the condensates within optical lattices : instabilities , solitons ( shape preserving excitations ) , breathers ( excitations characterized by internal oscillations ) , and self - trapping states or intrinsic localized modes ( wavepacket localized around few lattice sites ) @xcite . those predictions have often made use of theoretical analogies with other nonlinear classical and quantum problems , involving the sine - gordon equation , the discrete nonlinear schrdinger equation and other nonlinear physics problems . spatial instabilities of the condensate within the optical lattice , with a spontaneous breaking in the spatial periodicity , have been predicted by wu and niu @xcite . @xmath37 , the condensate being accelerated for @xmath38 and @xmath39 , respectively . the separations between the different spots vary because the detection occurred after different time delays . ] for 1d optical lattice , a linear increase of the detuning @xmath15 between the two laser beams forming the lattice provides to the optical lattice the constant acceleration given by eq . ( [ acceleration ] ) . as the lattice can only transfer momentum to the condensate in units of @xmath40 , the acceleration of the condensate leads to higher momentum classes as the acceleration time increases . with an initial momentum spread of the condensate much less than the bloch momentum @xmath41 ) and since the adiabatic switching transfers the momentum spread into lattice quasimomentum , the different momentum classes @xmath42 ( where @xmath43 ) occupied by the condensate wavefunction can be resolved directly after the time - of - flight . in the experiments of refs . @xcite with the optical lattice in the horizontal direction and the atomic wave diffraction monitored in the time - of - flight detection , the accelerated momentum classes showed up as diffraction peaks in the time - of flight absorption images as in fig.[leshouches2 ] . up to @xmath44 momentum could be transferred to the condensate preserving the phase - space density of the condensate during the acceleration , a result indicating that no heating or reduction of the condensate fraction occurred . measuring the average velocity of the condensate from the occupations of the different momentum states , the bloch oscillations of the condensate velocity corresponding to a bloch - period @xmath45 could be detected . note in fig.[leshouches2 ] that while the bloch oscillation takes place , the condensate wavefunction coherently occupies two neighboring velocity classes . in a configuration with the optical lattice oriented along the vertical direction @xcite , the different momentum classes emitted from the condensate travel in space separately because of the acceleration due to gravity . thus the condensate absorption images after the time of flight show one or two atom laser pulses corresponding to the single or double momentum class occupied by the condensate at the time of the release . at large acceleration of the lattice or at a decreased lattice depth , not all of the condensate could be coherently accelerated up to the final velocity of the lattice . such condensate loss can be interpreted in terms of landau - zener tunneling of the condensate out of the lowest band when the edge of the brillouin zone is reached . each time the condensate is accelerated across this edge , the fraction undergoing tunneling into the first excited band is given by the landau - zener probability @xmath46 @xcite : @xmath47 with the critical acceleration @xmath48 given by @xmath49 such tunneling produces the following mean velocity @xmath50 of the condensate at the end of the acceleration process for a final velocity @xmath51 of the lattice : @xmath52 in ref . @xcite it was verified that this formula correctly described the tunneling of the condensate at low values of the condensate density by varying both the potential depth and the lattice acceleration . in the configuration with the optical lattice oriented along the vertical direction @xcite , the different momentum classes emitted from the condensate travel in space separately because of the acceleration due to gravity . thus the condensate absorption images after the time of flight ( fig.[atomlaserpulses ] ) showed a part of the population confined in the optical lattice , and several laser pulses corresponding to the different momentum class occupied by the condensate . atomic momentum classes are separated by the gravity , pointing downwards . the large condensate amplitude near @xmath53 corresponds to the population confined in the lattice . the atomic momentum classes generate the atom laser pulses detected after a @xmath54 time of flight , with @xmath55 and condensate acceleration for @xmath56 . ] in order to measure accurately the variation of the effective potential @xmath57 with the interaction parameter @xmath32 as given by eq . ( [ potentialred ] ) , the landau - zener tunneling out of the lowest bloch band for small lattice depth was studied in a regime where the parameter @xmath32 modified the optical potential experienced by the condensate @xcite . therefore , the variation of the final mean velocity @xmath58 was studied as a function of the condensate density . the density was varied by changing the mean harmonic frequency of the magnetic trap . from the mean velocity the effective potential was then calculated using the landau - zener probability given above , with the critical acceleration determined by the effective potential @xmath59 fig . [ meanfield ] shows the ratio @xmath60 as a function of the parameter @xmath32 for the experimental investigation in two different geometries of the optical lattice , counter - propagating and in an angle - geometry with @xmath61 deg . as predicted by eq . ( [ cparameter ] ) , the reduction of the effective potential is much larger in the angle geometry . the theoretical predictions of eq . ( [ potentialred ] ) given by the theory of @xcite are also shown in the figure . an effective potential may also be derived within the framework of a tight - binding approximation using wannier states , i.e. describing the condensate within each potential minimum potential through the solution of the gross - pitaevskii equation while neglecting the overlap with neighboring potential minima @xcite , a very weak tunneling of the condensate preserving the overall phase relation of the condensate wavefunction . the tight - binding predictions presented in fig . [ meanfield ] provide a better agreement with experimental results than those of eq.([potentialred ] ) . parameter for the effective potential @xmath57 , normalized to the applied optical potential @xmath62 , for the two lattice geometries of ref . the experimental results for the counter - propagating geomotry ( open circles ) and angle geometry with @xmath61 deg ( filled squares ) are plotted together with the theoretical prediction of ref . @xcite(dashed line ) and of ref . @xcite(continuous line ) . parameters in these experiments were @xmath63 and @xmath64 for the counter - propagating lattice and @xmath65 and @xmath66 for the angle geometry . ] because the condensate is a superfluid , described by a wavefunction exhibiting long - range phase coherence , if the lattice potential is turned on smoothly , the condensate remains in the superfluid phase . in this regime a delocalized condensate wavefunction minimizes the total energy of the many - body system , the phase of the atomic wavefunction being well determined , with a fluctuating atom number at each lattice site . this applies as long as the atom - atom interactions are small compared to the tunnel coupling . instead , when the repulsive atom - atom interactions are large compared to the tunnel coupling , the total energy is minimized when each site of the optical lattice is filled with the same number of atoms . approaching this mott - isulator quantum phase transition , with the lattice site number commensurate to the atom number , the wavefunction phase coherence is lost . meanwhile , the fluctuations in the number of atoms per site are reduced and finally go to zero . in addition , in the superfluid regime the excitation spectrum is gapless , whereas the mott insulator phase exhibits a gap in the excitation spectrum . the first important step in modifying a condensate from a coherent state to a fock state was taken by orzel _ et al . _ @xcite who realized squeezed states in a 1d optical lattice . by properly choosing the depth of the optical potential and the amplitude of the mean field interaction term in eq . ( [ hamiltonianaa ] ) , the authors were able to reach a regime where deviation from the coherent state wavefunction for the condensate became significant . an increase in the phase fluctuations of the condensate wavefunction was measured in the interference of atom waves released from the optical lattice . from this , a squeezing in the number of atoms occupying the lattice sites was derived by assuming a minimum uncertainty state . the experimental realization of the mott isulator in a 3d optical lattice in ref . @xcite required a cubic optical lattice where a volume with 65 sites in each single direction was occupied , with potential depth around @xmath67 . adiabatic loading of a condensate of @xmath68 atoms into such a lattice produced a state whose coherence properties were tested as usual by the interference pattern following a time - of - flight . at a potential depth larger than @xmath69 the interference pattern presented an incoherent background , gaining more strength increasing the potential depth until that background was the only visible feature in the interference . that loss of interference was a first sign of the mott - isulator regime , with a fock state without macroscopic phase coherence characterizing the occupation of the single wells . evidence of the phase transition was gained by monitoring the quick restoration of the phase coherence , quick as compared to the restoration produced by the inhomogeneous dephasing of the superfluid condensate wavefunction under the application of a magnetic energy gradient . additional strong evidence was provided by the direct measurements of the gap in the excitation spectrum of the mott insulator , in contrast to the superfluid phase where the excitation spectrum is gapless . we have provided a brief overview of the research work performed on cold atomic samples ( laser cooled or evaporatively cooled to the quantum degenerate regime ) located within the periodic potential created by an optical interaction , concentrating our attention on the phenomena where the long range order of the external potential is imprinted onto the atomic wavefunction . as far as quantum degenerate gases within optical lattices are concerned , recently the field has greatly flourished with new experimental groups joining the crowded space and theory groups proposing new exciting schemes and pointing out the occurrence of interesting phenomena . it seems that bose - einstein condensates in optical lattices could be the test - bed for a large variety of theoretical models . it is obvious that this flourishing of ideas and of measurements will continue still for some time , because additional configurations could be explored , the control on the atoms and the optical lattices will progress , and finally because other phenomena may occur when fermi degenerate gases will be loaded in optical lattices . the detailed comparison between theory and experiments dealing with bose - einstein condensates within optical lattices requires heavy numerical solutions of the 3d gross - pitaevskii equation . however , the recent realization of nearly 1d bose - einstein condensates could simplify the theoretical analysis . but even with non - condensed , ultra - cold gases the versatility of optical lattices in realizing and studying quantum systems will certainly provide interesting challenges to physicists in the future . studies of cold collisions and diffusion properties in optical lattices are but two of the areas in which a lot of work still needs to be done . moreover , experiments will test the schemes in which the coherent motion of the atoms in far - detuned optical lattices could be exploited to realize quantum logic gates . the collaboration of d. ciampini , m. cristiani and j.h . mller to the experimental results reported in the present work is gratefully acknowledged . this work was supported by the murst through the prin2000 initiative , by the infm through the progetto di ricerca avanzata ` photon matter ' , and by the the eu through the cold quantum gases network , contract no . hprn - ct-2000 - 00125 . o.m . gratefully acknowledges a marie curie fellowship from the eu within the ihp programme . jessen and i.h . deutsch , adv . at . . phys . * 37 * , 95 ( 1996 ) . g. grynberg and c. trich , in _ coherent and collective interactions of particles and radiation beams _ edited by a. aspect , w. barletta , and r. bonifacio ( ios press , amsterdam ) ( 1996 ) . meacher , cont . phys . * 39 * , 329 ( 1998 ) . g. grynberg and c. robilliard , phys . rep . * 355 * , 335 ( 2001 ) . steck , w.h . oskay and m.g . raizen , science * 293 * , 274 ( 2001 ) and references therein . hensinger , h. hffner , a. browaeys , n.r . heckenberg , k. helmerson , c. mckenzie , g.j . milburn , w.d . phillips , s.l . rolston , h. rubinstein - dunlop , and b. upcroft , nature * 412 * , 52 ( 2001 ) and references therein . see , e.g. , i.h . deutsch and p.s . jessen , phys . a * 57 * , 1972 ( 1998 ) , and d. jaksch , h .- j . briegel , j.i . cirac , c.w . gardiner , and p. zoller , phys . lett . * 82 * , 1975 ( 1998 ) . g. kurizki , s. giovanazzi , d. odell , a. i. artemiev__new regimes in cold gases via laser - induced long - range interactions _ _ , in `` dynamics and thermodynamics of systems with long range interactions '' , t. dauxois , s. ruffo , e. arimondo , m. wilkens eds . , lecture notes in physics vol 602 , springer ( 2002 ) , ( in this volume ) a. hemmerich and t.w . hnsch , phys . lett . * 70 * , 410 ( 1993 ) . g. grynberg , b. lounis , p. verkerk , j.y . courtois , and c. salomon , phys . * 70 * , 2249 ( 1993 ) . see , e.g. , c. kittel , _ introduction to solid state physics _ ( wiley , new york , 1996 ) . p. verkerk , b. lounis , c. salomon , c. cohen - tannoudji , j.y . courtois , and g. grynberg , phys . lett . * 68 * , 3861 ( 1992 ) . p. rudy , r. ejnisman , and n.p . bigelow , phys . * 78 * , 4906 ( 1997 ) . g. raithel , w.d . phillips , and s.l . rolston , phys . lett . * 81 * , 3615 ( 1998 ) . buchkremer , r. bumke , h. levsen , g. birkl , and w. ertmer , phys . lett . * 85 * , 3121 ( 2000 ) . c. trich , l. guidoni , p. verkerk , and g. grynberg , in _ osa tops on ultra - cold atoms and bec _ , edited by k. burnett , vol . 7 , p. 82,(osa , washington d.c . , 1996 ) . o. morsch , p.h . jones , and d.r . meacher , phys . a * 61 * , 023410 ( 2000 ) . m. weidemller , a. hemmerich , a. grlitz , t. esslinger , and t.w . hnsch , phys . lett . * 75 * , 4583 ( 1995 ) . g. birkl , m. gatzke , i.h . deutsch , s.l . rolston , and w.d . phillips , phys . lett . * 75 * , 2823 ( 1995 ) . a. grlitz , m. weidemller , t.w . hnsch , and a. hemmerich , phys . 78 * , 2096 ( 1997 ) . meacher , s. guibal , c. mennerat , j.y . courtois , k.i . petsas , and g. grynberg , phys . * 74 * , 1958 ( 1995 ) . courtois , s. guibal , d.r . meacher , p. verkerk , and g. grynberg , phys . * 77 * , 40 ( 1996 ) . l. sanchez - palencia , f .- carminati , m. schiavoni , f. renzoni , and g. grynberg , phys . lett . * 88 * , 133903 ( 2002 ) . l. guidoni , b. dpret , a. di stefano , and p. verkerk , phys . a * 60 * , r4233 ( 1999 ) . a. hemmerich , m. weidemller , t. esslinger , c. zimmermann , and t.w . hnsch , phys . * 75 * , 37 ( 1995 ) . hamann , d.l . haycock , g. klose , p.h . pax , i.h . deutsch , and p.s . jessen , phys . lett . * 80 * , 4149 ( 1998 ) . h. perrin , a. kuhn , i. bouchoule , and c. salomon , europhys . lett . * 42 * , 395 ( 1998 ) . v. vuletic , c. chin , a.j . kerman , and s. chu , phys . * 81 * , 5768 ( 1998 ) . madison , m.c . fischer , and m.g . raizen , phys . a * 60 * , r1767 ( 1999 ) . wilkinson , c.f . bharucha , m.c . fischer , k.w . madison , p.r . morrow , qian niu , b. sundaram , and m.g . raizen , nature * 387 * , 575 ( 1997 ) . fischer , b. guttierez - medina , and m.g . raizen , phys . lett . * 87 * , 040402 ( 2001 ) . c. mennerat - robilliard , d. lucas , s. guibal , j. tabosa , c. jurczak , j.y . courtois , and g. grynberg , phys . * 82 * , 851 ( 1999 ) . m. ben - dahan , e. peik , j. reichel , y. castin , and c. salomon , phys . lett * 76 * , 4508 ( 1996 ) ; e. peik , m. ben dahan , i. bouchoule , y. castin , and c. salomon , phys . rev . a * 55 * , 2989 ( 1997 ) . wilkinson , c.f . bharucha , k.w . madison , qian niu , and m.g . raizen , phys . * 76 * , 4512 ( 1996 ) . qian niu , xian - geng zhao , g.a . georgakis , and m.g . raizen , phys . lett . * 76 * , 4504 ( 1996 ) . m. raizen , c . salomon , and qian niu , phys . today * 50 * ( 7 ) , 30 ( 1997 ) . see reviews by w. ketterle , d.s . durfee and d.m . stamper - kurn , and by e. cornell _ et al . _ in _ bose - einstein condensation in atomic gases _ , edited by m. inguscio , s. stringari and c. wieman ( ios press , amsterdam ) ( 1999 ) . k. berg - srensen and k. mlmer , phys . a * 58 * , 1480 ( 1998 ) . d. jaksch , c. bruder , j.i . cirac , c.w . gardiner , and p. zoller , phys . lett . * 81 * , 3108 ( 1998 ) . m. kozuma , l. deng , e.w . hagley , j. wen , r. lutwak , k. helmerson , s.l . rolston , and w.d . phillips , phys . lett . * 82 * , 871 ( 1999 ) . ovchinnikov , j.h . mller , m.r . doery , e.j.d . vredenbregt , k. helmerson , s.l . rolston , and w.d . phillips , phys . lett . * 83 * , 284 ( 1999 ) . hagley , l. deng , m. kozuma , m. trippenbach , y. b. band , m. edwards , m. doery , p. s. julienne , k. helmerson , s. l. rolston , and w. d. phillips , phys . lett . * 83 * , 3112 ( 1999 ) . j. stenger , s. inouye , a. p. chikkatur , d. m. stamper - kurn , d. e. pritchard , and w. ketterle , phys . rev . lett . * 82 * , 4569 ( 1999 ) ; and erratum : phys . rev . lett . * 84 * , 2283 ( 2000 ) . simsarian , j. denschlag , m. edwards , c.w . clark , l. deng , e.w . hagley , k. helmerson , s.l . rolston , and w.d . phillips , phys . * 85 * , 2040 ( 2000 ) . j. m. vogels , k. xu , c. raman , j. r. abo - shaeer , and w. ketterle , phys . 88 , 060402 ( 2002 ) j. steinhauer , r. ozeri , n. katz , and n. davidson , phys . 88 , 120407 ( 2002 ) . r. ozeri , j. steinhauer , n. katz , and n. davidson , phys . 88 , 220401 ( 2002 ) . anderson and m. kasevich , science * 282 * , 1686 ( 1998 ) . c. orzel , a.k . tuchman , m.l . fenselau , m. yasuda , and m.a . kasevich , science * 231 * , 2386 ( 2001 ) . m. greiner , o. mandel , t. esslinger , t.w . hnsch , and i. bloch , nature * 415 * , 6867 ( 2002 ) . s. burger , f. cataliotti , c. fort , f. minardi , m. inguscio , m.l . chiofalo , and m.p . tosi , phys . lett . * 86 * , 4447 ( 2001 ) . cataliotti , s. burger , c. fort , p. maddaloni , f. minardi , a. trombettoni , a. smerzi , and m. inguscio , science * 293 * , 843 ( 2001 ) . o. morsch , j.h . mller , m. cristiani , d. ciampini , and e. arimondo , phys . * 87 * , 140402 ( 2001 ) . m. cristiani , o. morsch , j. h. mtller , d. ciampini , and e. arimondo , phys . a * 65 * , 063612 ( 2002 ) . p. pedri , l. pitaevskii , s. stringari , c. fort , s. burger , f.s . cataliotti , p. maddaloni , f. minardi , and m. inguscio , phys . 87 * , 220401 ( 2001 ) . o. morsch , m. cristiani , j.h . mller , d. ciampini , and e. arimondo , cond - mat 0204528 . j. hecker - denschlag , j. e. simsarian , h. hffner , c. mckenzie , a. browaeys , d. cho , k. helmerson , s. l. rolston , and w. d. phillips , j. phys . b : at . mol . phys in press ( 2002 ) . s. burger , f. s. cataliotti , c. fort , p. maddaloni , f. minardi and m. inguscio , europhys * 57 * 1 ( 2002 ) . m. greiner , i. bloch , o. mandel , t.w . hnsch , and t. esslinger , phys . lett . * 87 * , 160405 ( 2001 ) . s. burger , k. bongs , s. dettmer , w. ertmer , k. sengstock , a. sanpera , g.v . shlyapnikov , and m. lewenstein , phys . lett . * 83 * , 5198 ( 1999 ) . l. deng , e.w . hagley , j. wen , m. trippenbach , y. band , p.s . julienne , j.e . simsarian , k. helmerson , s.l . rolston , and w.d . phillips , nature * 398 * , 6724 ( 1999 ) . f. dalfovo , s. giorgini , l.p . pitaevskii , and s. stringari , rev . 71 * , 463 ( 1999 ) . choi and qian niu , phys . lett . * 82 * , 2022 ( 1999 ) . a. brunello , f. dalfovo , l. pitaevskii , and s. stringari , phys . lett . * 85 * , 4422 ( 2000 ) . chiofalo , s. succi , and m.p . tosi , phys . a * 63 * , 063613 ( 2001 ) . m.krmer , l. pitaevskii , and s. stringari , phys . lett . * 88 * , 180404 ( 2002 ) . bronski , l.d . carr , b. deconink , and j.n . kutz , phys . lett . * 86 * , 1402 ( 2001 ) . b. wu , r.b . diener , and qian niu , phys . a * 65 * , 025601 ( 2002 ) . d.diakonov , l.m . jensen , c.j . pethick , and h. smith , e - print : cond - mat/0111303 . b. wu and qian niu , phys . a * 61 * , 023401 ( 2000 ) . o.zobay and b.m . garraway , phys . rev . * 61 * 033603 ( 2000 ) . a. trombettoni and a. smerzi , phys . lett . * 86 * , 2353 ( 2001 ) . b. wu and qian niu , phys . a * 64 * , 061603 ( 2001 ) . l. pitaevskii , and s. stringari , phys . * 88 * , 180402 ( 2001 ) . a. cuccoli , a. fubini , v. tognetti , and r. vaia , phys . rev . a * 64 * , 061601 ( 2001 ) . d. van oosten , p. van der straten , and h.t.c . stoof , phys . a * 63 * 053601 ( 2001 ) . k.burnett , m. edwards , c.w . clark , and m. shotter , j. phys . b : at . mol . . phys . * 35 * , 1671 ( 2002 ) . abdullaev , b. b. baizakov , s. a. darmanyan , v. v. konotop , and m. salerno , phys . a * 64 * , 043606 ( 2001 ) . konotop and m.salerno , phys . a * 65 * , 021602(r ) ( 2002 ) . n. tuskada , phys . a * 65 * , 063608 ( 2002 ) . m. kalinski , g. metakis , and w.p . schleich , private communication .

from the sizes of protoplanetary disks , we expect planets to form out to tens of au . the many giant planets found on small orbits by radial velocity and transit techniques are believed to have formed beyond the water ice lines of their stars and migrated inward . inward migration of massive planets can have a profound ( and usually destructive ) effect on the smaller objects in a planetary system , such as earth - sized planets within the habitable zones . it is therefore important to determine how many systems retain their giant planets on wide orbits . direct detection of these planets is limited to the most massive examples . indirect detection through the use of debris disks addresses this limitation . a debris disk consists of the circumstellar solid material that remains after the protoplanetary disk gas has dispersed and giant planets might have formed . although most of the mass in a debris disk is harbored by the parent bodies ( planetesimals ) , dust generated in their collisions accounts for the majority of the disk s surface area , and observations of debris disks via their thermal emission or reflected stellar radiation trace this dust . while tens of debris disks have been spatially resolved , the majority are detected only as an infrared excess in the spectral energy distribution ( sed ) of their star . typical debris disk temperatures are tens to a few hundred kelvins , emitting as modified blackbodies that peak in the mid to far infrared . these wavelengths are well - suited for study with the _ spitzer space telescope _ @xcite infrared spectrograph ( irs ; @xcite ) and multiband imaging photometer for _ spitzer _ ( mips ; @xcite ) . for a recent review of debris disks , see @xcite . debris disks often appear constrained to one or two discrete rings . this is evident from images of resolved disks such as fomalhaut @xcite , and from seds of unresolved debris disks that are fit well by one or two blackbody functions , corresponding to dust at one or two distinct radial locations . of the 28 disks without strong emission features presented by @xcite , all but one were fit better with a blackbody function than with a continuous disk model . the exception was hr8799 , which was later determined to be best modelled by two blackbodies @xcite . @xcite originally fit some excess seds with a power law , representing a continuous radial distribution of dust ; however , @xcite argued that these power law fits require the optical depth of the disk to increase with orbital radius ( which is theoretically implausible ) , and they found that these targets can be fit well by two blackbodies instead . these warm " and cold " debris disk components may be analogous to the asteroid belt and kuiper belt in the solar system . why do rings form , and what sets their location ? ice lines are one possibility . during the protoplanetary disk phase , a radial pressure gradient in the gas partially counteracts the gravitational force on the gas , allowing it to rotate at a sub - keplerian velocity . solid particles orbit at keplerian rates , and thus experience a head wind that slows their rotation and makes them spiral inwards . when these solids reach the ice line , the volatile component sublimates , producing a local pressure increase that counteracts the overall pressure gradient . this creates a zone where the particles can settle without a headwind . thus , there is a tendency to have a planetesimal belt near the ice line , and this belt can then produce grains that assume a specific temperature . @xcite found similar warm disk temperatures ( 190 k ) around stars of different stellar types . this can potentially be explained by the presence of the water ice line at 150 - 170 k. @xcite argue that ice lines of other species ( e.g. co , @xmath0 , @xmath1 ) at lower temperatures may play an important role in planet formation in the outer solar system . if ice lines are also responsible for setting the location of the cold components , we would expect these components to have similar temperatures over a range of stellar types . planets may also cause the discrete ring structures . planetesimals will be scattered away once a planet is massive enough to dominate the gravity in its vicinity . observations support the expected relation between planets and disk structure : the four giant planets imaged around hr 8799 @xcite are located in the gap between the warm and cold debris disk components @xcite ; the imaged planet orbiting @xmath2 pic appears to sculpt the inner edge of its debris disk @xcite ; the well - resolved cold debris ring around fomalhaut is likely confined by planets @xcite ; and in the solar system , neptune sculpts the inner edge of the kuiper belt @xcite . if the locations of debris disks are set by planets , we can use cold debris disks to investigate wide - orbit planets , which are not easily studied by other means . of the over 850 confirmed exoplanets , 35 have orbits larger than 5 au , and only 17 have orbits larger than 10 au ( nasa exoplanet archive ) . this is likely due to the observational biases of the radial velocity and transit detection techniques . direct imaging can detect wide - orbit planets , but the current technology is only sensitive to very massive planets around young , nearby stars . uranus and neptune , for instance , could not be directly detected from outside the solar system . the planet formation processes beyond 5 - 10 au are poorly understood . for example , did the planets around hr 8799 form at their current locations or did they ( or at least their cores ) form on smaller orbits , then migrate outwards via scattering interactions with other planets or planetesimals ? cold debris disks provide an observational test of these planet formation and migration theories . planet formation by core accretion becomes less efficient farther from the star due to a radial increase in the dynamical timescale and decline in the surface density of solids in the protoplanetary disk @xcite . therefore , the inner edge of a cold debris disk may represent the outer limit of efficient core accretion @xcite . a planet migrating outwards into a debris disk will push the inner edge of the disk outward as well . uranus and neptune may have formed on smaller orbits and migrated into the kuiper belt @xcite . in this alternate scenario , the inner edge of a cold debris disk may represent the limits of outward migration . in this paper , we focus on _ spitzer _ measurements of cold debris disks , showing how their temperatures vary with the temperature of their central star and what this implies about planet formation and migration on wide - orbits . first , we describe the selection of our target stars ( [ sec : targetselection ] ) . then , we outline our photometric ( [ sec : photometry ] ) and irs ( [ sec : irsdatareduction ] ) data acquisition / reduction . next , we detail our modeling of the stellar photosphere sed ( [ sec : photospheremodel ] ) , our derivation of the infrared excess , and our fitting of blackbodies to the excess ( [ sec : blackbodyfitting ] ) . finally , we analyze the results ( [ sec : results ] ) and discuss the implications for wide - orbit planet formation and migration ( [ sec : discussion ] ) , before offering a summary and concluding remarks ( [ sec : conclusions ] ) . we searched the _ spitzer _ observers log for main - sequence stars that were observed with the irs long low ( ll ) module ( both orders ) in staring mode and with mips at 24 @xmath3 and 70 @xmath3 , and we accumulated a sample of 546 targets . the stellar properties of our target list are summarized in table [ table : stellarproperties ] . after reducing and analyzing the data , we refined the sample as described in [ sec : blackbodyfitting ] . this yielded 225 stars with significant excess ( of which 174 had cold components ) , which are detailed in table [ table : excessdata ] . it is important to note that our sample comprised stars from a variety of _ spitzer _ observational programs , each having targets selected in a different manner ( e.g. nearby stars , stars with previously detected infrared excess , stars with rv - detected planets , etc . ) . hence , our sample was not selected to be statistically representative in any one sense ; rather , it was designed to include as many relevant targets as possible . we calculated the stellar temperature , @xmath4 , from the star s v-@xmath5 color using tabulated values for main - sequence stars ( or interpolations between these values ) from @xcite , and we estimated a 200 k one - sigma uncertainty on these values . @xmath4 is listed for targets with significant excess in table [ table : excessdata ] . over the range of spectral types and metallicities of our sample of stars , v-@xmath5 is an accurate indicator of stellar effective temperature , with a peak - to - peak scatter of less than @xmath6 1.5% and weak metallicity dependence @xcite . we use @xmath4 to parametrize stellar type , and this degree of accuracy is sufficient for our purpose . some of the uncertainty in @xmath4 reflects the effect of interstellar reddening on v-@xmath5 , although the great majority of stars in this sample are within the local bubble and therefore should not be strongly reddened ( of the 174 targets for which we detect cold components , only 29 are beyond 100 pc , 19 are beyond 120 pc , and 4 are beyond 150 pc ) . ages for these stars were estimated from a combination of chromospheric activity measurements , x - ray emission , placement on the hr diagram , surface gravity , membership in clusters and associations , and gyrochronology collected from the literature . @xcite describe the intercomparison of these methods and how they are applied , and the quoted ages are on the scale calibrated by @xcite . the ages are listed in table [ table : stellarproperties ] , along with references for the measurements used to derive the ages and a quality flag for the age accuracy ( ranging from 0 if no age could be determined to 3 if there were three or more age measurements with good agreement ) . cccccccccc hip110 & hd224873 & & g5 & 49.6 & 0.42 & 2 & 2,20 & 5346816 & not included + hip345 & hd225200 & hr9102 & a0v & 124.8 & 0.1 & 2 & 31 & 12720128 & warm + cold + hip394 & hd225239 & hr9107 & g2v & 39.2 & 0.25 & 1 & 2 & 4028672 & not included + hip490 & hd105 & & g0v & 39.4 & 0.17 & 3 & 2,3,4,5,6,9,12 & 5295616 & cold + hip522 & hd142 & hr6 & f7v & 25.7 & 4.8 & 2 & 3,17,27 & 22296064 & not included + hip544 & hd166 & hr8 & k0ve & 13.7 & 0.24 & 3 & 1,2,6,15,23 & 12717824 & warm + cold + hip560 & hd203 & hr9 & f3vn & 39.4 & 0.01 & 2 & 38 & 14983424 & cold + hip682 & hd377 & & g2v & 39.1 & 0.22 & 3 & 2,4,9,12,20,32 & 5268736 & warm + cold + hip910 & hd693 & hr33 & f8vfe-08h-05 & 18.7 & 3 & 3 & 3,4,23,24 & 14494720 & not included + & & & & & & & & 4008448 & + hip919 & hd691 & & k0v & 34.2 & 0.24 & 3 & 2,9,12,14,18,20,39 & 5345280 & not included + hip1031 & hd870 & & k0v & 20.2 & 2.3 & 2 & 3,5 & 16026112 & cold + hip1134 & hd984 & & f5 & 47.1 & 0.25 & 3 & 2,4,9,20 & 5271808 & not included + hip1292 & hd1237 & & g8.5vk : & 17.5 & 0.3 & 3 & 2,3,4,5,6 & 22290176 & not included + hip1368 & & & k7 & 15 & 0.9 & 3 & 1,2,39 & 25674240 & cold + hip1473 & hd1404 & hr68 & a2v & 41.3 & 0.45 & 1 & 7 & 14160384 & warm + although we built our target list around the _ spitzer _ irs spectra , we supported these spectra in our analysis with a suite of photometric data . the properties of the photometric systems we used are summarized in table [ table : photometrybands ] and the photometric data for the targets with significant excess ( see [ sec : blackbodyfitting ] ) are given in table [ table : excessdata ] . mips photometry at 24 @xmath3 provided an important calibration reference for the irs data ( [ sec : irsdatareduction ] ) , while 70 @xmath3 photometry provided a crucial constraint on the temperature of cold debris disks ( [ sec : blackbodyfitting ] ) . we used our in - house debris disk pipeline to reduce and extract photometry for the mips data as part of the effort to preserve the legacy of _ spitzer _ measurements on debris disk studies ( @xcite ; k. su et al . 2013 , in prep ) . basic reduction ( up to the post - bcd mosaics ) and calibrations of the mips data follow the descriptions by @xcite and @xcite . in addition , the extraction of the 24 @xmath3 photometry was briefly described in @xcite , where the source position at 24 @xmath3 is determined by a combination of psf fitting and 2d gaussian fitting . both psf fitting and aperture photometry aperture photometry used an aperture radius of 6.255@xmath7 , a sky annulus of 19.92 - 29.88@xmath7 , and an aperture correction factor of 1.6994 . ] were performed , and we preferentially used the aperture photometry at 24 @xmath3 because this was used to calibrate the photosphere model at 24 @xmath3 , as described in [ sec : photospheremodel ] and the appendix ( the psf and aperture photometry agree to within a few percent ) . we then used the 24 @xmath3 source position to perform psf fitting for data at 70 @xmath3 by minimizing the residual signal at the source position ( for details , see k. su et al . 2013 , in prep ) . for faint sources located in areas with structured background , the resultant 70 @xmath3 photometry can be negative , which reflects non - detection . the 1@xmath8 photometry uncertainty ( listed in table [ table : excessdata ] ) includes the pixel - to - pixel variation near the source of interest , and detector repeatability ( 1% and 5% of the source flux at 24 and 70 @xmath3 , respectively ) . mips photometry data for all of our targets can be found in @xcite , @xcite , and k. su et al . ( 2013 , in prep ) . v , j , h , and @xmath5 photometry were used to model the stellar photosphere sed ( [ sec : photospheremodel ] ) and to estimate @xmath4 ( [ sec : targetselection ] ) . we obtained _ hipparcos _ v band @xcite and 2mass j , h , and k band photometry @xcite from the vizier online database . many of the targets in the sample are nearby stars and are severely saturated in the 2mass data . to overcome this , we used heritage aperture photometry , which we transformed to match the 2mass system . when both 2mass and transformed heritage photometry of high quality were available , we averaged them . the references for the heritage photometry are in table [ table : excessdata ] . irac photometry ( 3.6 @xmath3 ) was also used to model the stellar photosphere sed ( [ sec : photospheremodel ] ) . these data were taken in _ spitzer _ cycle 7 ( pid 70076 , pi : su ) . all the irac data were taken in subarray mode with four dithered positions to avoid saturation . we used the basic calibrated data ( bcd ) products provided by the _ science center ( pipeline version s18.18 ) , and performed the necessary steps ( pixel solid angle correction and pixel phase correction ; k. su et al . 2013 , in prep ) to extract the photometry . aperture photometry was used for each individual data frame ( 64 frames per dithered position ) ; and the final quoted flux is the median value of all measurements per star . wise data @xcite were obtained from vizier , but are not included in table [ table : excessdata ] . because wise photometry is less accurate than _ spitzer _ photometry , we did not use it quantitatively . instead , we used it as a qualitative confirmation of our _ spitzer _ data . our irs reduction started with the level 1 bcd products , downloaded from the _ spitzer _ heritage archive . in addition to irs ll data , we also reduced short low ( sl ) module data , when available , and included them in our analysis . the astronomical observation requests ( aors ) used for each target are listed in table [ table : stellarproperties ] . the basic reduction was performed using the spectroscopic modeling analysis and reduction tool ( smart ) software package @xcite , scripted into a series of automated routines . for each spectral order of each irs aor , three files ( 2d spectra , uncertainty , and mask ) were combined into a single 3-plane file . bad and rogue pixels were removed by the routine irsclean with the clean parameter set to 4096 ( pixels with this value or higher were included in the clean ) . next , when available , multiple data collection events ( dces ) for the same nod position were combined , and then the background was removed from each 2d spectrum by subtraction of the opposite nod . the 2d spectra were then converted into 1d spectra using smart s optimal 2 nod extraction @xcite . the 1d spectra from both nods were combined . the result was a wavelength , flux , and uncertainty vector of each spectral order for each aor . the bonus " third order data were not used . the remainder of the data reduction and analysis was performed with the matlab software package . irs data at wavelengths near the order edges often were of poor quality , so data longward of 38 @xmath3 and shortward of 20.75 @xmath3 were discarded from the ll first order , data shortward of 14.3 @xmath3 were discarded from the ll second order , data longward of 14.7 @xmath3 and shortward of 7.55 @xmath3 were discarded from the sl first order , and data shortward of 5.25 @xmath3 were discarded from the sl second order . systematic offsets in flux existed between irs orders , which we fixed by applying a multiplicative correction factor to the ll first order spectrum and to both sl order spectra ( if they existed ) , in order to align them with the ll second order flux . initially , we determined the order correction factors using an automated routine , but due to the variety of shapes of the spectra and the presence of outlying data points , we found that fine - tuning these factors by eye was more reliable . the data from all available orders and modules were then combined into a single spectrum . next , we cut outlying data points from the spectrum in an iterative process . the spectrum was fit with a polynomial and the standard deviation of residual flux values around the fit was calculated . points lying more than three standard deviations from the fit were discarded . this process was iterated six times . because the scatter of the data generally increased towards longer wavelengths , we applied this process separately to the short and long ends of our spectrum ; if sl data were available , the two sections were divided at 14 @xmath3 and each section fit with a fourth degree polynomial , whereas if no sl data were available , the sections were divided at 25 @xmath3 and each section was fit with a second degree polynomial . this procedure , on average , cut 10 data points from each spectrum , with the first iteration typically cutting five points , and the sixth iteration only cutting zero , one , or two points ( nearly 90% of spectra had no points cut in the sixth iteration ) . before cutting , the spectra had 181 , 296 , or 373 points , depending on the available orders . after the outliers were cut , the spectrum was smoothed by binning to a wavelength resolution of 0.7 @xmath3 . the absolute calibration of irs data is only accurate to @xmath910% , whereas the mips calibration is accurate to @xmath92% @xcite . to reduce systematic errors in our irs data , we multiplicatively scaled our spectrum to be consistent with the measured mips 24 @xmath3 flux density for each target . the irs spectra were calibrated as point sources , so if a source was slightly extended , it would have an incorrect slit - loss correction . the stellar photospheres ( point - like ) are dominant in the irs spectra for the majority of sources in the sample ; therefore , this has no significant impact on the result . if the mips photometry for a source was contaminated by background emission , this contamination would be passed to the irs spectrum of the source through this scaling ( the irs data may or may not have picked up the contamination , depending on the slit orientation ) . nevertheless , an additional multiplicative factor was later applied to the irs data ( see [ sec : photospheremodel ] ) , which corrected any lingering errors from this process . some targets were observed with more than one irs aor . we combined the spectra from these aors , interleaving their data points , then re - smoothing to a spectral resolution of 0.7 @xmath3 . although the spectrum of a main - sequence star peaks in the visible or near - ir , the stellar photosphere can still be the dominant source of flux density in the mid - ir . thus , to extract the thermal emission of the debris disk ( the infrared excess ) , an accurate model spectrum of the stellar photosphere must be generated and subtracted from the data . it is common to use photosphere spectra from detailed numerical models of stellar structures , such as kurucz / atlas9 @xcite . however , these models are not well - tested in the mid- and far - infrared . @xcite compared various families of models ( kurucz / atlas9 , marcs , and nextgen / phoenix ) and found that the choice of model family can affect whether a star is determined to have an infrared excess or not . we opted for a simpler model , a rayleigh - jeans relation given by @xmath10 , where rj is an amplitude scale factor . this model is appropriate for wavelengths beyond 5 @xmath11 m because stars in this study , mostly a through k dwarfs , have virtually no spectral features in the mid - ir , behaving as blackbodies in the rayleigh - jeans regime . for example , in the range of most interest for this study , the departure of a reference a0 star spectrum @xcite from a rayleigh - jeans fit between 15 and 30 @xmath3 is no more than @xmath6 0.3% , and the atlas9 @xcite solar model follows rayleigh - jeans behavior to within @xmath6 0.24% from 20 to 40 @xmath3 . however , the theoretical spectra show minor systematic differences from the observations @xcite , so use of them to improve the knowledge of the seds would be risky . modeling the photosphere thus came down to finding the appropriate scale factor ( rj ) for each star in our sample . we used a set of color relations ( derived from a sample of stars known to have no ir excess ) to predict the mips 24 @xmath3 magnitude of the photosphere , [ 24 ] , from measured v , j , h , @xmath5 , and irac magnitudes bands where ir excesses due to very close - in dust are very rare . the details of these relations are described in the appendix . this procedure estimates the 24 @xmath11 m photospheric outputs to about 2% rms , based solely on data at wavelengths short of 4 @xmath11 m . we then converted [ 24 ] to flux density units and used it to find the scaling factor for the photosphere model , which is given in table [ table : excessdata ] for targets with significant excess ( see [ sec : blackbodyfitting ] ) . we finally applied a small multiplicative factor to the irs data if it clearly looked slightly offset from the photosphere model . then , we subtracted the photosphere model from the irs spectra and mips 70 @xmath3 point to obtain the infrared excess , @xmath12 . the photosphere model was assumed to have an uncertainty of 2% , thus the uncertainty in the excess flux density was @xmath13 ^ 2}.\ ] ] data are omitted from the right panels , as these points were not used to constrain the fits . ] to interpret the excesses in a general way , we needed to assign them fiducial temperatures . the emission of disk grains is a function of their optical constants ; however , as discussed in [ sec : introduction ] , debris disks radiate like blackbodies at one or two temperatures . the simplest possible description of the emission is in terms of these temperatures . in this section , we describe how we determined if each target had significant excess and whether the excess was best described by one or two blackbodies . we fit @xmath14 with a one component blackbody model @xmath15 where @xmath16 is the planck function . the best - fit parameters , @xmath17 and @xmath18 were found by minimizing the reduced chi - squared , @xmath19 ^ 2 } { \sigma_\text{excess}\left(\lambda_i\right)^2 } \nonumber\\ & + 28\frac { \left[f_{\nu\,\text{excess}}\left(70 \micron\right)-f_{\nu\,\text{model}}\left(70 \micron\right)\right]^2 } { \sigma_\text{excess}\left(70 \micron\right)^2 } \bigg\},\end{aligned}\ ] ] where @xmath20 , n is the number of data points in the irs excess data , and n is the number of free parameters in the fit ( n=2 ) . the mips 70 @xmath3 data point was weighted in the fit as 28 irs data points , equivalent to the number of irs data points ( 0.7 @xmath3 resolution ) that would fit inside the equivalent width of the mips 70 @xmath3 spectral response function ( 19.65 @xmath3 ) . the fit was performed using matlab s lsqcurvefit algorithm . @xmath18 was constrained to between 0 and 500 k. while @xmath17 represents the amplitude of the debris disk emission , a more useful measure of a debris disk s brightness is the fractional excess , @xmath21 . we calculated @xmath22 according to @xmath23 from equation 2 of @xcite . @xmath24 and @xmath25 are the peak flux density values of the disk emission and stellar photosphere emission , respectively , which occur at wavelengths @xmath26 and @xmath27 . while the peak flux densities for the disk components were easily calculated from our best fit model , the rayleigh - jeans stellar photosphere model had no maximum . to overcome this , we created a blackbody function using the temperature of the star from table [ table : stellarproperties ] , and scaled it to match the flux density of our rayleigh - jeans model at 24 @xmath3 . from this representation of the photosphere , we found the maximum flux density , and calculated @xmath22 . data are omitted from the right panels , as these points were not used to constrain the fits . ] with our best fits in hand , we refined our sample to only those targets with statistically significant excess . first , we inspected all fits by eye and discarded targets with poor quality data that resulted in clearly unrealistic fits . second , we discarded targets with excess that was too faint , @xmath28 . third , we inspected the fits and identified all targets whose excess relied solely on the mips 70 @xmath3 data point ( i.e. there was no excess in the irs data ) . for these cases , we required this mips point to represent a significant excess , so we discarded targets where @xmath29 or where @xmath30 . this process resulted in 321 of the original 546 stars having no significant excess . of the remaining 225 targets with significant excess , we next determined whether the excess consisted of one or two components . to do this , we fit the excess sed of each target with a model consisting of the sum of two blackbodies , @xmath31 finding the optimal set of the four parameters @xmath32 , @xmath33 , @xmath34 , and @xmath35 was again done by minimizing the reduced chi - squared ( equation [ eq : reducedchisq ] ) , now with n=4 . data are omitted from the right panels , as these points were not used to constrain the fits . ] our definition of cold " was the coldest well - detected component of the excess that was below 130 k. @xcite found a continuous distribution of warm components above @xmath9130 k , centered around 190 k , whose temperatures were set by the water ice line . thus , any component above 130 k was considered warm " for the purposes of this study . a component with a temperature of 110 k , for example , would be considered warm if another , colder temperature component was also detected , but would be considered cold if no colder component was detected . to implement this , we fit each target once using 100 k as the division between warm and cold , and again with the division at 130 k. we then selected the better of these two cases ( based on reduced chi - squared ) to represent the best two - component model . for many targets , these two cases yielded identical fits . in both cases , the minimum cold temperature allowed by the fit was 0 k , and the maximum allowed warm temperature was 500 k. deciding if each target warranted a two - component model was a two - step process . first , we required that the reduced chi - squared of the two - component fit be at least three times better than that of the one - component fit ( i.e. all targets where @xmath36/@xmath37 were deemed to be better fit by a single component ) . second , for the remaining targets , we identified those for which the presence of the cold component relied entirely on the mips 70 @xmath3 point , meaning the irs excess was entirely fit by the warm component . for these targets to have significant cold components , we required their mips 70 @xmath3 excess to fall more than 3@xmath8 above the excess predicted by the warm component . that is , if one of these targets had @xmath38 we concluded that it was best fit by one component . this step was analogous to the third step we performed when deciding if each target had excess or not . the targets that passed both of these criteria were deemed to have two components . this process identified 100 single - component cold disks , 51 single - component warm disks , and 74 two - component disks . we found only one target ( hip45585 ) that required two warm components ( both @xmath39130 k ) to fit properly , and we discarded this target from our results as it had no cold components ( it is counted in the 321 targets that we discarded ) . we found no targets requiring two components colder than 100 k. our methods were insensitive to possible additional cold components below @xmath950 k. we concluded that our fitting procedure allowed us to find all cold components above this limit . equivalently , our method reliably fit the inner edges of the cold disks . for the two - component disks , we calculated the fractional excesses for the warm and cold components , @xmath40 and @xmath41 , using equation [ eq : fwyatt2 ] . one - component excesses were deemed warm or cold depending if they had temperatures above or below 130 k. the verdicts for all targets are given in table [ table : stellarproperties ] and the parameters of our best fits for targets with significantly - detected components are listed in table [ table : excessdata ] . we estimated that the uncertainty in @xmath33 was 10 k. this estimate was conservative ; warmer cold components had more irs data points in excess of the photosphere , so their fits were more constrained , with a temperature uncertainty of 3 to 7 k. we were unable to accurately constrain the temperatures of very cold components , which were detected only at 70 @xmath11 m . with only one data point , blackbodies at a wide range of cold temperatures could be given the amplitude necessary to match the point , so a degeneracy existed between t and @xmath22 in blackbody fits to the excess sed . we made this distinction at 45 k ; above this temperature there was enough information in the irs spectra to constrain the temperature . cold temperatures above 45 k were considered true detections , while cold temperatures below 45 k were considered unconstrained and were replaced with upper limits at 45 k. by using the 70 @xmath3 data to determine if such targets had significant excesses , we ensured that these disks were truly very cold , rather than warm but very faint . of the 174 cold components in our sample , 25 had temperature upper limits . some irs spectra exhibit mineralogical emission features ( e.g. hip41081 , hip57971 ) . our model did not explicitly account for these features , but they peak sharply at @xmath910 @xmath3 and do not resemble blackbody functions . these features did not influence our fits , except to raise the value of the minimum reduced chi - squared . the data and best fits for a small sample of our targets are shown in figures [ fig : twocompgallery ] , [ fig : coldcompgallery ] , and [ fig : warmcompgallery ] ( for targets with two components , one cold component , and one warm component , respectively ) . the left panels in all of these figures show the measured data with the photosphere and blackbody models , while the right panels illustrate the excess data and models with the photosphere subtracted . histograms of the significant warm and cold debris disk temperatures are shown in figure [ fig : temphistogram ] . 45 k around cooler stars . the green line is the best fit trend to the data , determined by a bayesian linear regression , @xmath42 . a representative error bar is in the lower right . ] @xcite studied the behavior of the warm disk components in detail , but their sample ( restricted to stars with ages less than 1 gyr ) had too few 70 @xmath3 detections of solar - type stars ( 9 ) to determine any trends between the cold component temperature and stellar type . to look for such a trend in our larger sample , we plot the cold component temperature versus stellar temperature in figure [ fig : tdisktstarscatter ] . well - constrained disk temperatures are black circles ( open circles are young systems with age less than 25 myr ) , and upper limits are downward facing blue triangles . although there is substantial scatter in the cold component temperatures , a positive correlation between cold debris disk temperature and stellar temperature is evident . to quantify this trend , we fit a linear relation to the data by bayesian analysis using the routine linmix_err . the routine properly handles upper - limits , and it uses uncertainties in both x and y directions ( we assumed an uncertainty of 10 k in @xmath33 and 200 k in @xmath4 ) . in its model , the routine also includes the normally - distributed intrinsic scatter of the data around the trend . the posterior distributions of the slope , intercept , and intrinsic scatter from the bayesian regression are show in figure [ fig : trendposteriors ] . the slope of the linear regression was @xmath43 , thus the existence of a trend between @xmath33 and @xmath4 is significant at level greater than 4.5@xmath8 . the 1@xmath8 intrinsic scatter around the trend was @xmath44 k. the best - fit trend line from the bayesian analysis , @xmath42 , is plotted in green in figure [ fig : tdisktstarscatter ] . the choice of a linear fit is not physically motivated ; its purpose is to show that to first order there is a correlation between cold debris disk temperature and stellar type . a significant source of scatter around the trend results from the 10 targets with @xmath45 k and @xmath46 k. six of these stars ( hip560 , hip64184 , hip64995 , hip65875 , hip67497 , and hip78663 ) are less than 25 myr old ( most are part of the scorpius - centaurus association ) . although we show in [ sec : discussion ] that there is no significant trend in @xmath33 with age across the broad range of stellar ages in our sample , the excesses of these targets may be a product of their relatively young ages . they represent a period ( 5 - 50 myr ) when considerable collisional activity may still be occurring as a result of planet building @xcite . vs. @xmath4 trend slope , intercept , and intrinsic scatter , generated by the linmix_err routine . ] we now consider what determines the temperature of cold debris disks and how a trend with stellar type might arise . as discussed in 1 , @xcite found similar warm disk temperatures ( 190 k ) around stars of different stellar types , and they attributed the effect to a particle trap at the water ice line . if ice lines of other species set the location of cold debris disks , we would expect cold disks around different type stars to have a common temperature . however , the trend of cold component temperature with stellar type ( figure [ fig : tdisktstarscatter ] ) is inconsistent with any strictly temperature - dependent mechanism for setting the location of cold debris disks . @xcite show that during the protoplanetary disk phase , the water ice line location is time dependent , and predicting it requires accounting for viscous heating in the disk , as well as the evolution of the pre - main - sequence stellar luminosity . such considerations have not been applied to potential ice lines of species other than water , and we do not consider this level of detail here . delayed stirring , in which the radial location of dust in a debris disk moves outwards with time , is a postulated mechanism for debris disk evolution @xcite . this could occur if the parent bodies are distributed in a broad ring . particles in a debris disk collide and grind into dust faster on smaller orbits where the dynamical timescale is shorter . thus , the location of the emitting dust moves outward with time , and , therefore , becomes cooler with time . because late type stars have longer lifetimes than early type stars , the late type stars in our sample are generally older than the early type stars . if delayed stirring occurs , then this age bias in our sample could explain the observed trend in @xmath33 with @xmath4 . to test this hypothesis , we plot @xmath33 against the age of the system ( if available ) for targets in three @xmath4 bins ( 5000 to 6000 k , 6000 to 7000 k , and 7500 to 9500 k ) , shown in figure [ fig : agevstdisk ] . we see no trend in @xmath33 with age in any bin , suggesting that delayed stirring does not produce the trend of disk temperature with stellar type . is plotted against the age of the system for targets in three @xmath4 bins . no trend is seen with age , suggesting age does not play a confounding role in our discovered @xmath33 vs. @xmath4 trend . this also argues against the occurrence of delayed stirring . ] perhaps cold debris disks are all at roughly the same orbital distance , regardless of stellar type . assuming the disk is heated to its equilibrium temperature , the relation between the disk s temperature , location , and stellar type is given by @xmath47 the second and third lines of equation [ eq : equilibriumtemp ] are derived assuming @xmath48 and @xmath49 ( valid in the roughly solar mass range ) . so with @xmath50 constant , early type stars would host warmer disks . @xcite perform detailed simulations of the evolution of debris disks with inner and outer edges at 30 and 150 au , respectively , around stars with mass ranging from 1 to 3 @xmath51 . their models output the disk emission at 24 and 70 @xmath3 , which show that the ( color ) temperature of these cold debris disks does increase with stellar mass , in agreement with this expectation . but is the orbital location of cold debris disks truly constant with stellar type ? if the size of cold debris disks traces the size of the original protoplanetary disks , we can use observations of the latter to address this question . @xcite find that protoplanetary disk mass and radius are related by @xmath52 furthermore , observations @xcite show that @xmath53 combining these relations yields @xmath54 so protoplanetary disk size does increase with earlier stellar type , albeit slowly . how quickly must the disk size increase for it to maintain a constant temperature ? from equation [ eq : equilibriumtemp ] , @xmath55 although the size of disks does increase with earlier spectral type , it does so more slowly than required to maintain a constant temperature , thus disks are expected to be warmer around early type stars , consistent with our findings . substituting equation [ eq : ppradius ] into equation [ eq : equilibriumtemp ] reveals how the disk temperature would vary with spectral type in this case : @xmath56 new results from @xcite suggest that equation [ eq : ppandstarmass ] may hold only for @xmath57 , with disk mass constant or possibly even decreasing as @xmath58 for higher mass stars . if this were true and translated into a flat or decreasing @xmath50 with @xmath59 , then the disk temperature would increase even faster with spectral type than derived in equation [ eq : migresult ] . is it plausible that the inner edge of cold debris disks scales with the size of the original protoplanetary disk ? some mechanism must clear the material inside cold debris disks and set their inner edge , and planets are a common explanation . an outwardly migrating planet would set an inner edge , but migration from planet - planet scattering can be a chaotic and unpredictable phenomenon ; simulations show that the planet s final location depends sensitively on the initial conditions of the system @xcite . this suggests that a trend with stellar temperature would not arise . outward migration via scattering through a smooth disk of planetesimals would proceed in a more orderly manner , and the planet would halt its migration when the surface density of planetesimals decreased below a certain threshold , at a location that scales with the size of the original protoplanetary disk . planet formation by core accretion ( without migration ) would also clear debris inside the cold component and create its inner edge . planetesimals would be either incorporated into the planets as they formed , or scattered away once the planets became massive enough to dominate the gravitational field in their vicinity . as mentioned in [ sec : introduction ] , core accretion efficiency declines with increasing orbital radius , leaving an outer zone of planetesimals beyond the planets . the timescale for the formation of a planet with a given mass scales as @xmath60 where p is the orbital period and @xmath61 is the surface density of solids . substituting kepler s third law , @xmath62 ( @xmath63 is the orbital distance ) and the typical protoplanetary disk structure , @xmath64 @xcite into equation [ eq : formationtime ] gives @xmath65 after a time ( @xmath66 ) planets will have formed out to a given orbital location ( @xmath63 ) , which sets the inner edge of the cold debris disk . so setting @xmath66 constant and @xmath67 predicts that the size of the disk would scale with spectral type as @xmath68 by comparing this with equation [ eq : equilibriumradius ] , we see that in this case as well , the size of the disk grows more slowly with stellar type than required to maintain a constant equilibrium temperature , consistent with our observational results . substituting equation [ eq : formradius ] into equation [ eq : equilibriumtemp ] shows how the disk temperature would vary with stellar type if its inner edge were set by the limits of planet formation : @xmath69 we studied the circumstellar environment of 546 main - sequence stars via the mid infrared emission of their debris disks , as measured by the _ spitzer _ space telescope . after subtracting a model of the flux expected from the stellar photosphere , we obtained an sed of the infrared excess for each target . we found 225 targets with significant excess : 100 with a single cold component , 51 with a single warm component , and 74 with two components . examining the results revealed a trend between the temperature of the inner edge of the cold debris disk component and that of its host star . photosphere magnitude for a sample of 1037 stars . the blue curve shows the gaussian fit to the distribution with @xmath70 and @xmath71 . ] this trend is inconsistent with theories that predict the location of cold debris disks to be strictly temperature - dependent , i.e. we rule out the dominance of ice lines in sculpting the outer regions of planetary systems . we also rule out delayed stirring as the source of this trend . the trend can potentially be explained if the outward migration of planets traces the extent of the primordial protoplanetary disk , which tends to be limited to warmer equilibrium temperatures for hotter stars . the trend can also be explained if planets form in situ out to a distance where the core accretion efficiency drops below a certain threshold , leaving a cold debris disk that is warmer around earlier type stars . we thank scott kenyon and benjamin bromley for helpful science discussions , and the referee for many useful comments . we also thank vianney lebouteiller for help in scripting the irs reduction pipeline . this work was supported by contract 1255094 from caltech / jpl to the university of arizona . we used the simbad database and the vizier catalogue access tool , operated at cds , strasbourg , france . we also used the nasa / ipac infrared science archive , which is operated by the jet propulsion laboratory , california institute of technology , under contract with nasa . this work is based on observations made with the _ spitzer _ space telescope , which is operated by the jet propulsion laboratory , california institute of technology under a contract with nasa . and we used data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by nasa and the nsf . finally , we used data products from the wide - field infrared survey explorer , which is a joint project of the university of california , los angeles , and the jet propulsion laboratory / california institute of technology , funded by nasa . here we describe how we used photometry to derive a model photosphere magnitude at 24 @xmath3 . the relation among v , @xmath5 , and [ irac ] used to predict [ 24 ] is equation [ eq : mips24fromivk ] , but a number of other relations were used beforehand to improve the accuracy of the result . first , we preprocessed @xmath5 , h , and j photometry to remove small nonlinearities in the relations . these small corrections to the 2mass photometry were derived by comparison with our all - sky warm mission measurements in irac band 1 . the preprocessing replaced @xmath5 with @xmath72 according to @xmath73 h with h according to @xmath74 and j with j according to @xmath75 equation [ eq : mips24fromivk ] utilizes @xmath76 , a combination of measured and derived @xmath5 magnitudes . one of these , @xmath77 , was derived by solving the relation among v , @xmath77 , and j , given by @xmath78 where @xmath79 . this relation is a fit to the behavior of more than 1000 main - sequence stars , as are the other relations that are given below . @xmath80 was derived ( if h was measured ) by solving the relation among v , @xmath80 , and h , given by @xmath81 where @xmath82 . @xmath76 was then calculated by averaging @xmath72 , @xmath77 , and @xmath80 with relative weights of 1 , 0.75 , and 0.47 , respectively . the weights were determined from typical 2mass errors and by minimizing the residuals in comparing the projected photospheric levels with the 24 @xmath11 m measurements . equation [ eq : mips24fromivk ] also uses @xmath83_\text{super}$ ] , a combination of measured and derived irac magnitudes . the derived irac magnitude , @xmath83_1 $ ] , was calculated by solving @xmath84_1 = k_{s\,\text{super } } + 0.0012x^4 - 0.01448x^3 + 0.0419x^2 - 0.056x + 0.0295,\ ] ] where @xmath85 . the measured [ irac ] and derived @xmath83_1 $ ] were averaged together with relative weights of 1 and 0.5 , respectively , yielding @xmath83_\text{super}$ ] . finally , [ 24 ] was calculated with @xmath86 = [ \text{irac}]_\text{super } + 0.001x^5 - 0.0097x^4 + 0.0179x^3 + 0.0305x^2 - 0.0821x + 0.0312,\ ] ] where , again , @xmath85 . this process required v and at least one of j , h , or @xmath5 to proceed . targets that did not have this minimum photometry available were not included in our sample , although nearly all targets in our sample had a complete suite of v , j , h , and @xmath5 photometry available ( table [ table : excessdata ] ) . the accuracy and precision of this method is illustrated in figure [ fig:24accuracy ] , which shows the distribution of the difference between the observed and predicted [ 24 ] for over 1000 stars . a gaussian fit to this distribution ( the blue curve ) has a mean of 0.00158 and a standard deviation of 0.0224 . cccccccccccccccc hip345 & 6.39 & 6.28 & 6.25 & 6.26 & & 8843 & & 35.7 @xmath6 0.38 & 97.24 @xmath6 5.35 & 11.9 & 94.9 @xmath6 5.35 & 53 & 185 & 9.42 & 4.56 + hip490 & 7.53 & 6.46 & 6.19 & 6.12 & & 5923 & & 28.39 @xmath6 0.3 & 152.7 @xmath6 9.71 & 14.2 & 149.9 @xmath6 9.71 & 48 & & 42 & + hip544 & 6.13 & 4.73 & 4.63 & 4.31 & & 5405 & & 159.1 @xmath6 1.58 & 105.8 @xmath6 6.32 & 72.6 & 91.57 @xmath6 6.33 & 50 & 126 & 4.99 & 4.78 + hip560 & 6.19 & 5.45 & 5.33 & 5.24 & & 6635 & & 117.2 @xmath6 1.12 & 67.08 @xmath6 7.06 & 31.6 & 60.89 @xmath6 7.06 & 129 & & 18.4 & + hip682 & 7.59 & 6.42 & 6.15 & 6.12 & & 5809 & & 36.49 @xmath6 0.39 & 170.6 @xmath6 10.62 & 14.7 & 167.7 @xmath6 10.6 & @xmath8745 & 119 & 47.3 & 11.6 + hip1031 & 7.23 & 5.85 & 5.47 & 5.38 & & 5372 & 1.9 @xmath6 0.03 & 50.67 @xmath6 0.51 & 22.51 @xmath6 4.98 & 27.8 & 17.07 @xmath6 4.99 & 50 & & 3.29 & + hip1368 & 9 & 6.38 & 5.75 & 5.58 & & 3903 & 1.7 @xmath6 0.03 & 46.6 @xmath6 0.48 & 20.74 @xmath6 5.15 & 26.5 & 15.55 @xmath6 5.15 & @xmath8745 & & 9.99 & + hip1473 & 4.52 & 4.34 & 4.42 & 4.46 & & 9164 & & 154.4 @xmath6 1.56 & 43.82 @xmath6 6.5 & 65.7 & 30.94 @xmath6 6.51 & & 133 & & 1.38 + hip1481 & 7.46 & 6.46 & 6.25 & 6.15 & & 6258 & & 34.76 @xmath6 0.36 & -0.83 @xmath6 2.98 & 13.7 & -3.518 @xmath6 2.98 & & 217 & & 11.9 + hip1499 & 6.46 & 5.33 & 5.04 & 4.89 & & 5692 & 2.95 @xmath6 0.03 & 80.94 @xmath6 0.81 & 61.43 @xmath6 6.81 & 43.1 & 52.98 @xmath6 6.81 & 54 & & 5.54 & + hip2072 & 3.9 & 3.74 & 3.58 & 3.51 & & 8131 & & 311.2 @xmath6 3.1 & 72.97 @xmath6 4.43 & 143 & 44.94 @xmath6 4.47 & @xmath8745 & 169 & 0.577 & 1.47 + hip2472 & 4.77 & 4.67 & 4.77 & 4.7 & & 9096 & & 112.5 @xmath6 1.12 & 76.98 @xmath6 6.54 & 49.5 & 67.27 @xmath6 6.54 & 69 & 192 & 1.47 & 1.5 + hip2578 & 5.07 & 5.06 & 5.16 & 4.99 & & 9028 & & 232 @xmath6 2.31 & 56.48 @xmath6 4.13 & 35.5 & 49.52 @xmath6 4.13 & & 194 & & 15.6 + hip2710 & 6.91 & 6.04 & 5.85 & 5.75 & & 6473 & 1.33 @xmath6 0.02 & 40.62 @xmath6 0.44 & 104.7 @xmath6 6.5 & 19.8 & 100.8 @xmath6 6.5 & @xmath8745 & 110 & 16.5 & 2.94 + hip2843 & 6.71 & 5.84 & 5.66 & 5.59 & & 6510 & 1.57 @xmath6 0.03 & 42.13 @xmath6 0.46 & 21.54 @xmath6 3.76 & 23.3 & 16.98 @xmath6 3.76 & 64 & & 2.43 & + ccc v & 0.5423 & 3729 + 2mass j & 1.241 & 1623 + 2mass h & 1.6513 & 1075 + 2mass @xmath5 & 2.1657 & 676 + wise 1 & 3.35 & 309.54 + wise 2 & 4.60 & 171.79 + wise 3 & 11.56 & 31.67 + wise 4 & 22.09 & 8.36 + irac 1 & 3.6 & 269.53 + mips 24 & 23.675 & 7.17 + mips 70 & 71.42 & 0.778 +

the problem of time synchronization in large distributed systems consists of giving all the physically disjoint elements of the system a common time scale on which to operate . this common time scale is usually achieved by periodically synchronizing the clock at each element to a reference time source , so that the local time seen by each element of the system is approximately the same . time synchronization plays an important role in many systems in that it allows the entire system to cooperate and function as a cohesive group . time synchronization is an old problem @xcite , but the question of scalability is not . recent advances in sensor networks show a clear trend towards the development of large scale networks with high node density . for example , a hardware simulation - and - deployment platform for wireless sensor networks capable of simulating networks with on the order of 100,000 nodes was recently developed @xcite . as well , for many years the smart dust project sought to build cubic - millimeter motes for a wide range of applications @xcite . also , there is work in progress on the drastic miniaturization of power sources @xcite . these developments ( and many others ) indicate that large scale , high density networks are on the horizon . large scale , high density networks have applications in a variety of situations . consider , for example , the military application of sniper localization . large numbers of wireless nodes can be deployed to find the shooter location as well as the trajectory of the projectile @xcite . since the effective range of a long - range sniper rifle can be nearly @xmath0 km , in order to fully track the trajectory of the projectile it may be essential that our deployed network be tightly synchronized over distances of a few kilometers . another example might be the implementation of a distributed radio for communication . in extracting information from a deployed sensor network , it may be beneficial for the nodes to cooperatively transmit information to a far away receiver @xcite . such an application would require that nodes across the network be well synchronized . as a result , a need for the synchronization of large distributed systems is very real and one that requires careful study to understand the fundamental performance limits on synchronization . the synchronization of large networks has been studied in fields ranging from biology to electrical engineering . the study of synchronous behavior has generally taken one of two approaches . the first approach is to consider synchronization as an emergent behavior in complex networks of oscillators . in that work , models are developed to describe natural phenomena and synchronization emerges from these models . the second approach is to develop and analyze algorithms that synchronize engineering networks . nodes are programmed with algorithms that estimate clock skew and clock offset to achieve network synchronization . however , both of these approaches have significant limitations . emergent synchronization properties in large populations has been the object of intense study in the applied mathematics ( @xcite ) , physics ( @xcite ) , and neural networks ( @xcite ) literature . these studies were motivated by a number of examples observed in nature : * in certain parts of south - east asia , thousands of male fireflies congregate in trees and flash in synchrony at night @xcite . * pacemaker cells of the heart , which on average cause 80 contractions a minute during a person s lifetime @xcite . * the insulin - secreting cells of the pancreas @xcite . for further information and examples , see @xcite , and the references therein . a number of models have been proposed to explain the emergence of synchrony , but perhaps one of the most successful and well known is the model of _ pulse - coupled oscillators _ by mirollo and strogatz @xcite , based on dynamical systems theory . consider a function @xmath1\to[0,1]$ ] that is smooth , monotone increasing , concave down ( i.e. , @xmath2 and @xmath3 ) , and is such that @xmath4 and @xmath5 . consider also a phase variable @xmath6 such that @xmath7 , where @xmath8 is the period of a cycle . then , each element in a group of @xmath9 oscillators is described by a state variable @xmath10 $ ] and a phase variable @xmath11 $ ] as follows : * in isolation , @xmath12 . * if @xmath13 then @xmath14 , and if @xmath15 then @xmath16 . * when @xmath17 for any of the @xmath18 s and some time @xmath19 , then for all other @xmath20 , @xmath21 @xmath22 where @xmath23 denotes an infinitesimal amount of time after @xmath19 . that is , oscillator @xmath18 reaching the end of a cycle causes the state of all other oscillators to increase by the amount @xmath24 , and the phase variable to change accordingly . the state variable @xmath25 can be thought of as a voltage . charge is accumulated over time according to the nonlinearity @xmath26 and it discharges once it reaches full charge , resetting the charging process . upon discharging , it causes all other charges to increase by a fixed amount of @xmath24 , up to the discharge point . for this model , it is proved in @xcite that for all @xmath9 and for almost all initial conditions , the system eventually becomes synchronized . for the network to converge into a synchronous state , one key assumption is that the behavior of every single oscillator is governed by the same function @xmath27 . this means that all oscillators must have the same frequency . from the literature , it appears that this requirement is nearly always needed . as far as we are aware , for a fully synchronous behavior to emerge , the oscillators need to have the same , or nearly the same , oscillation frequencies . the need for nearly identical oscillators presents a significant limitation for emergent synchronization . this emergence of synchrony is clearly desirable and it has been considered for communication and sensor networks in @xcite . however , whether or not these techniques can be adapted to synchronize networks with nodes that have arbitrary oscillator frequencies ( clock skew ) is still unclear . thus , in order to overcome this limitation and find techniques capable of synchronizing a more general class of networks , we turn to algorithms designed to estimate certain unknown parameters such as clock skew . there have been many synchronization techniques proposed for use in sensor networks . these algorithms generally allow each node to estimate its clock skew and clock offset relative to the reference clock . reference broadcast synchronization ( rbs ) @xcite eliminates transmitter side uncertainties by having a transmitter broadcast reference packets to the surrounding nodes . the receiving nodes then synchronize to each other using the arrival of the reference packets as synchronization events . tiny - sync / mini - sync @xcite and the timing - sync protocol for sensor networks ( tpsn ) @xcite organize the network into a hierarchial structure and the nodes are synchronized using pair - wise synchronization . in lightweight tree - based synchronization ( lts ) @xcite , pair - wise synchronization is also employed but the goal of lts is to reduce communication and computation requirements by taking advantage of relaxed accuracy constraints . the flooding time synchronization protocol ( ftsp ) @xcite achieves one - hop synchronization by having a root node broadcast timing information to surrounding nodes . these surrounding nodes then proceed to broadcast their synchronized timing information to nodes beyond the broadcast domain of the root node . this process can continue for multi - hop networks . the problem with each of these traditional synchronization techniques is that synchronization error will increase with each hop . since each node is estimating certain synchronization parameters , i.e. clock skew , there will be inherent errors in the estimate . as a result , a node multiple hops away from the node with the reference clock will be estimating its parameters from intermediate nodes that already have estimation errors . therefore , this introduces a fundamental _ scalability problem _ : as the number of hops across the network grows , the synchronization error across the network will also grow . current trends in network technology are clearly moving us in the direction of large , multi - hop networks . first , sensors are decreasing in size and this size decrease will most likely be accompanied by a decrease in communication range . thus , more hops will be required to traverse a network deployed over a given area . second , as we deploy networks over larger and larger areas , then for a given communication range , the number hops across the network will also increase . in either case , the increased number of hops required to communicate across the network will increase synchronization error . therefore , it is essential that we develop techniques than can mitigate the accumulation of synchronization error over multiple hops . to decrease the error increase in each hop , we need to decrease the estimation error . there are two primary ways of achieving this . first , each node can increase the amount of timing information it obtains from neighboring nodes . for example , from a received timing packet , the node may be able to construct a data point telling it the approximate time at the reference clock and the corresponding time at its local clock . using a collection of these data points , the node can estimate clock skew and clock offset . so instead of using , say , five packets with timing information , a node can wait for ten packets . more data points will generally give better estimates . the drawback to such an approach is the increase in the number of packet exchanges . the second way in which to reduce estimation error is to increase the quality of each data point obtained by the nodes . this can be achieved through improving packet exchange algorithms and time stamping techniques . however , we believe that there is one fundamentally new approach to improving data point quality that has not be carefully studied . this is to use spatial averaging to improve the quality of each data point . the motivation for this approach is very simple . assume that each node has many neighbors . if all nodes in the network are to be synchronized , then the neighbors of any given node will also have synchronization information . is it possible to simultaneously use information from all the neighbors to improve the quality of a timing observation made by a node ? furthermore , it would seem to make sense that with more neighbors , hence more available timing information , the quality of the constructed data point should improve . if this is indeed the case , then achieving synchronization through the use of spatial averaging will provide a fundamentally new trade - off in improving synchronization performance . network designers would simply be able to increase the number and density of nodes to obtain better network synchronization . the study of cooperative time synchronization using spatial averaging is the focus of this work . to obtain a model for developing cooperative synchronization in large wireless networks , we begin by looking at the signals observed by a node in a network with @xmath9 nodes uniformly deployed over a fixed finite area . to start , we assume propagation delay to negligible ( the general case is considered in section [ sec : timedelay ] ) . all nodes transmit a pulse @xmath28 and a node @xmath29 will see a signal @xmath30 which is the superposition of all these pulses , @xmath31 in this expression , @xmath28 is the basic pulse transmitted by each node ( assumed to be the same for all nodes ) . @xmath32 is the ideal pulse transmit time , but since we assume imperfect time synchronization among the nodes we have @xmath33 modelling random errors in the pulse transmission time . @xmath34 models the amplitude loss in the signal transmitted by the @xmath18th node . @xmath35 is the maximum magnitude transmitted by a node . we scale each node s transmission by @xmath9 so that as the network density grows , the total power radiated does not grow unbounded . this model thus describes the received signal seen at a node @xmath29 for a network with @xmath9 nodes and this holds for any @xmath9 . increasing @xmath9 will have two effects : ( a ) node density will increase since the network area is fixed and ( b ) node signal transmission magnitude will decrease due to the @xmath36 scaling . therefore , by increasing @xmath9 this model allows us to study the scalability of networks as node density grows and node size decreases . given that these are the signals observed at each node , we ask : is it possible for @xmath30 to encode a time synchronization signal that will enable all nodes in the network to synchronize their clocks with bounded error , as @xmath37 ? the answer is yes , and the key to proving all our results is the law of large numbers . our key idea is the following . if all nodes were able to determine when time @xmath32 ( in the reference time ) arrives , then by transmitting @xmath28 at time @xmath32 , the signal observed at any node @xmath29 would be @xmath38 , which is a suitably scaled version of @xmath28 centered at @xmath32 . in reality however , there will be some error in the determination of @xmath32 , which we account for by allowing for a node - dependent random error @xmath39 . but , if the distribution of @xmath39 satisfies certain conditions , then the effects of that timing error can be averaged out . a pictorial representation of why this should be the case is shown in fig . [ fig : why - sync - holds ] . therefore , intuitively we can see how the technique of _ cooperative time synchronization _ using spatial averaging can average out the inherent timing errors in each node . even though there is randomness and uncertainty in each node s estimates , by using cooperation among a large number of nodes it is possible to recover _ deterministic _ parameters from the resulting aggregate waveform ( such as the location of certain zero - crossings ) in the limit as node density grows unbounded . thus more nodes will give us better estimates . this is because the random waveform converges to a deterministic one as more and more nodes cooperatively generate an aggregate waveform . at the same time , the average power required by each node will decrease since smaller nodes send smaller signals . therefore , by programming suitable dynamics into the nodes , in this paper we show how it is possible to generate an aggregate output signal with _ zero - crossings in the limit of asymptotically dense networks . thus , the detection of these zero - crossings plays the same role as that of an externally generated time reference signal based on which all nodes can synchronize . we develop this synchronization technique in three main steps . one , we set up the model for @xmath30 . two , we specify characteristics of the model ( i.e. the distribution of @xmath33 ) that allow us to prove desirable properties of the aggregate waveform ( such as a center zero - crossing at @xmath32 ) . three , we develop the estimators needed for our synchronization technique and show that the estimators give us the desired characteristics . the main contributions presented in this paper are the following ; * the definition of a probabilistic model for the study of the time synchronization problem in wireless networks . this model does contain the classical mirollo - strogatz model as a special case , but its formulation and the tools used to prove convergence results are of a completely different nature ( purely probabilistic , instead of based on the theory of dynamical systems ) . * using this model , we provide a rigorous analysis of a new cooperative time synchronization technique that employs spatial averaging and has favorable scaling properties . as the density of nodes increases , synchronization performance improves . in particular , in the limit of infinite density , deterministic parameters for synchronization can be recovered . * we show that cooperative time synchronization works perfectly for negligible propagation delay . when propagation delay is considered , we find that asymmetries at the boundaries reveal some limitations that need to be carefully considered in designing algorithms that take advantage of spatial averaging . in analyzing the proposed cooperative time synchronization technique , our goal is to show that the proposed technique can average out all random error and provide deterministic parameters for synchronization as node density grows unbounded . this asymptotic result can be viewed as a _ convergence in scale _ to synchrony . the result serves as a theoretical foundation for allowing a new trade - off between node density and synchronization performance . in particular , higher node density can yield better synchronization . the rest of this paper is organized as follows . the general model is presented in section [ sec : systemmodel ] . of particular interest here is section [ sec : specialcase ] , where we show how our model contains the model of mirollo and strogatz for pulse - coupled oscillators as a special case @xcite . in section [ sec : timesync - setup ] we specialize the general model for our synchronization setup and develop waveform properties that will be used in time synchronization . in section [ sec : timesynchronization ] we develop the cooperative time synchronization technique for no propagation delay . we extend the cooperative synchronization ideas to the case of propagation delay in section [ sec : timedelay ] . the paper concludes in section [ sec : conclusion ] with a detailed discussion on the scalability issue and how the technique proposed in this work lays the theoretical foundation for a general class of cooperative time synchronization techniques that use spatial averaging . we consider a network with @xmath9 nodes uniformly distributed over a fixed finite area . the behavior of each node @xmath18 is governed by a clock @xmath40 that counts up from @xmath41 . the introduction of @xmath40 is important since it provides a consistent timescale for node @xmath18 . by maintaining a table of pulse - arrival times , node @xmath18 can utilize the arrival times of many pulses over an extended period of time . the clock of one particular node in the network will serve as the reference time and to this clock we wish to synchronize all other nodes . we will call the node with the reference clock node @xmath42 and the clocks of other nodes are defined relative to the clock of node @xmath42 . we never adjust the frequency or offset of the local clock @xmath40 because we wish to maintain a consistent time scale for node @xmath18 . the clock of node @xmath42 , @xmath43 , will be defined as @xmath44 where @xmath45 . taking @xmath43 to be the reference clock , we now define the clock of any other arbitrary node @xmath18 , @xmath40 . we define @xmath40 as @xmath46 where * @xmath47 is an unknown offset between the start times of @xmath40 and @xmath43 . * @xmath48 is a constant and for each @xmath18 , @xmath49 $ ] where @xmath50 are finite . this bound on @xmath51 means that the frequency offsets between any two nodes can not be arbitrarily large . * @xmath52 is a stochastic process modeling random timing jitter . thus , this model assumes that there is a bounded constant frequency offset between the oscillators of any two nodes as well as some random clock jitter . it is important to note that node @xmath42 does not have to be special in any way ; its clock is simply a reference time on which to define the clocks of the other nodes . this means that our clock model actually describes the relative relationship of all the clocks in the network by using an arbitrary node s clock as a reference . from section [ sec : intromodel ] , we see that we are interested in studying the aggregate waveform observed at a node @xmath29 . as a result , we are only concerned with the aggregate signal magnitude and do not care about the particular signal contribution from each surrounding node . with this in mind , we can develop a random model for pathloss that , for dense networks , gives the appropriate aggregate signal magnitude at node @xmath29 . such a model is ideal for our situation since we are studying asymptotically dense networks . we start with a general pathloss model @xmath53 , where @xmath54 for all distances @xmath55 , is non - increasing and continuous . @xmath53 is a fraction of the transmitted magnitude seen at distance @xmath56 from the transmitter . for example , if the receiver node @xmath29 is at distance @xmath56 from node @xmath18 , and node @xmath18 transmits a signal of magnitude @xmath57 , then node @xmath29 will hear a signal of magnitude @xmath58 . we derive @xmath53 from a power pathloss model since any pathloss model captures the _ average _ received power at a given distance from the transmitter . this average received power is perfect for modelling received signal magnitudes in our problem setup since we are considering asymptotically dense networks . due to the large number of nodes at any given distance @xmath56 from the receiver , using the average received magnitude at distance @xmath56 as the contribution from each node at that distance will give a good modelling of the amplitude of the aggregate waveform . the random pathloss variable @xmath59 will be derived from @xmath53 . to understand how @xmath59 and @xmath53 are related , we give an intuitive explanation of the meaning of @xmath59 as follows : the @xmath60 $ ] is the fraction of nodes at distances @xmath56 from node @xmath29 such that @xmath61 , where @xmath62 is a small constant . this means that , roughly speaking , for any given scaling factor @xmath63 , @xmath64 is the fraction of received signals with magnitude scaled by approximately @xmath65 , where @xmath66 is the probability density function of @xmath59 . thus , if we scale the transmit magnitude @xmath57 from every node @xmath18 by an independent @xmath59 , then as the number of nodes , @xmath9 , gets large , node @xmath29 will see @xmath67 signals of approximate magnitude @xmath68 , and this holds for all @xmath65 in the range of @xmath59 . this is because taking a large number of independent samples from a distribution results in a good approximation of the distribution . thus , this intuition tells us that scaling the magnitude of the signal transmitted from every node @xmath18 by an independent sample of the random variable @xmath59 gives an aggregate signal at node @xmath29 that is the same magnitude as if we generated the signal using @xmath53 directly . even though the signals from two nodes at the same distance from a receiver have correlated magnitudes , we do not care about the signal magnitude from any particular node but only that an appropriate number of all possible received signal magnitudes contribute to the aggregate waveform . for a receiving node @xmath29 , we choose therefore to work with the random variable @xmath59 instead of directly with @xmath53 because , for the goals of this paper , doing so has two major advantages : ( a ) we can obtain desirable limit results by placing very minimal restrictions on the distribution of the @xmath59 s ( and hence on @xmath53 ) and ( b ) we can apply tools from probability theory ( basically , the strong law of large numbers ) to carry out our analysis . from the above intuition we can define the cumulative distribution function of @xmath59 as @xmath69 \\ 1 & k\in(1,\infty ) \end{array } \right.\ ] ] where * @xmath70 is the total area of the network , * @xmath71 is the area of the network contained in a circle of radius @xmath72 centered at node @xmath29 , * @xmath73 . from the above discussion we see that the distribution of @xmath59 is only a function of node @xmath29 , the receiving node . we illustrate the relationship among node @xmath29 , @xmath53 , @xmath74 , and @xmath75 in fig [ fig : propmodel - illus ] . we sometimes write @xmath34 with @xmath18 used to index each node surrounding node @xmath29 . @xmath18 is thus indexing a sequence of independent random variables @xmath34 for fixed @xmath29 . therefore , for a given @xmath29 , @xmath34 s are independent and identically distributed ( i.i.d . ) with a cumulative distribution function given by ( [ eq : prop_cdf ] ) for all @xmath18 . we assume that @xmath59 has the following properties : * @xmath59 is independent from @xmath76 for all @xmath29 , @xmath77 , and @xmath78 . * @xmath79 , @xmath80 , and @xmath81 . the requirements on the random variable @xmath59 places restrictions on the model @xmath53 . any function @xmath53 that yields a @xmath59 with the above requirements can be used to model pathloss . in this section we develop a more complex model to simultaneously model propagation delay and pathloss . this leads to the joint development of the delay random variable @xmath82 and a corresponding pathloss random variable @xmath59 . since we want to develop a model for both pathloss and time delay , we start by keeping the pathloss function @xmath53 defined in section [ sec : prop - model ] . the general delay model assumes a function @xmath83 that models the time delay as a function of distance . @xmath83 describes the time in terms of @xmath43 that it takes for a signal to propagate a distance @xmath56 . for example , if node @xmath18 and node @xmath29 are distance @xmath84 apart , then a pulse sent by node @xmath18 at time @xmath85 will be seen at node @xmath29 at time @xmath86 . we make the reasonable assumption that @xmath83 is continuous and strictly monotonically increasing for @xmath55 . as with the pathloss only model , we want to define a delay random variable @xmath82 for each receiving node @xmath29 . recall that this means that for every node @xmath29 there is a random variable @xmath82 associated with it since , in general , each node @xmath29 will see different delays . there is a correlation between the delay random variable @xmath82 and the pathloss random variable @xmath59 . this correlation arises for two main reasons . first , since in section [ sec : prop - model ] we define @xmath53 to be monotonically decreasing and continuous , it is possible for @xmath87 for @xmath88 , @xmath89 . this might be the case for a multi - hop network . in this situation , there will be a set of nodes whose transmissions will never reach node @xmath29 ( i.e. infinite delay ) even though according to @xmath83 these nodes should contribute a signal with finite delay . second , a small @xmath59 value would represent a signal from a far away node . as a result , the corresponding @xmath82 value should be large to reflect large delay . therefore , keeping these two points in mind , we proceed to develop a model for both pathloss and propagation delay . we define the cumulative distribution function of @xmath82 as @xmath90 \\ a(x-\delta(r))+\frac{a(j , r)}{a_{t } } & x\in(\delta(r),\delta(r+\delta r ) ] \\ 1 & x\in(\delta(r+\delta r),\infty ) \end{array } \right.\ ] ] where @xmath91 , @xmath92 is a constant , @xmath93 , and @xmath94 recall that @xmath71 , defined in section [ sec : prop - model ] , is the area of the network contained in a circle of radius @xmath72 centered at node @xmath29 and @xmath70 is the total area of the network . note that @xmath95 can be infinite . using the delay random variable @xmath82 with the cumulative distribution function in ( [ eq : delay_cdf ] ) , we define @xmath59 as @xmath96 where @xmath97 is the deterministic pathloss function from section [ sec : prop - model ] and @xmath98 is the inverse function of the deterministic delay function @xmath99 . note that @xmath100 exists since @xmath99 is continuous and strictly monotonically increasing on @xmath101 . to understand the distribution of @xmath82 , we need to consider the definition of @xmath59 as well . recall that a signal arriving with delay @xmath82 is scaled by the pathloss random variable @xmath59 . let us consider the cumulative distribution in two pieces , @xmath102 $ ] and @xmath103 . the case for @xmath104 is trivial . first , for @xmath102 $ ] , the probability that @xmath82 takes a value less than or equal to @xmath105 is simply the fraction of the network area around node @xmath29 such that the nodes are at distances @xmath56 with @xmath106 . the intuition is the same as that for the development of @xmath59 in section [ sec : prop - model ] . second , for @xmath103 , the situation is more complex . note that a transmitted signal from a node at distance @xmath107 from @xmath29 will arrive at node @xmath29 with infinite delay since @xmath87 for @xmath107 . since any delay values in @xmath103 correspond to distances @xmath108 , the corresponding scaling value will be zero because @xmath59 and @xmath82 are related by ( [ eq : djkjconnection ] ) . as a result , it does not matter what delay values we assign to the fraction of the network area outside a circle of radius @xmath95 centered at node @xmath29 as long as their delay value @xmath105 is such that @xmath109 . thus , we can arbitrarily choose a constant @xmath110 value and construct a piecewise linear portion of the cumulative distribution function of @xmath82 on @xmath103 . the probability that @xmath111 will be the fraction of the network area outside a circle of radius @xmath95 around node @xmath29 . and since @xmath111 will have a corresponding @xmath59 value that is zero , this fraction of nodes will not contribute to the aggregate waveform at node @xmath29 . it is clear that the correlated @xmath82 and @xmath59 random variables work together to accurately model a signal arriving with both pathloss and propagation delay . an illustration of how @xmath53 , @xmath83 , node @xmath29 , and @xmath112 are related can be found in fig . [ fig : delaymodel - illus ] . we require that @xmath82 is bounded , has finite expectation , and has finite variance for all @xmath29 . note that @xmath113 by the requirement that @xmath114 . as well , since the cumulative distribution in ( [ eq : delay_cdf ] ) is continuous , and often absolutely continuous , we assume that @xmath82 has a probability density function @xmath115 . when we write @xmath116 , the @xmath18 indexes each node surrounding node @xmath29 . thus , the @xmath116 s are independent and identically distributed in @xmath18 for a given @xmath29 and have a cumulative distribution given by ( [ eq : delay_cdf ] ) . using the @xmath59 and @xmath82 developed in this section to simultaneously model pathloss and propagation delay , respectively , we will be able to closely approximate the received aggregate waveform at any node @xmath29 as @xmath37 . to summarize , we see that our choice of the pathloss and delay random variables will depend on what we want to model . if we only consider pathloss and not propagation delay , then we will use the random variable @xmath59 defined in section [ sec : prop - model ] . if we account for both pathloss and delay , then we will use the delay random variable @xmath82 in this section ( section [ sec : delay - model ] ) and the pathloss random variable @xmath59 defined by ( [ eq : djkjconnection ] ) . the exchange of pulses is the method through which the network will maintain time synchronization . each node @xmath18 will periodically transmit a scaled pulse @xmath117 , where @xmath118 is a constant and @xmath28 , in general , can be any pulse . we call the interval of time during which a synchronization pulse is transmitted a _ synchronization phase_. what each node does with a set of pulse arrival observations is determined by the pulse - connection function @xmath119 for node @xmath18 . the pulse - connection function is a function that determines the time , in the time scale of @xmath40 , when node @xmath18 will send its @xmath120th pulse . it can be a function of the current value of @xmath121 and past pulse arrival times . this function basically determines how any node @xmath18 reacts to the arrival of a pulse . the system model that we presented thus far is powerful because it is very general . in this section we show that it is a generalization of the pulse - coupled oscillator model proposed by mirollo and strogatz @xcite . as a result , the results presented in that paper will hold under the simplified version of our model . in setting up the system model , mirollo and strogatz make four key assumptions : * pathloss model : the first assumption that is made is that there is all - to - all coupling among all @xmath9 oscillators . this means that each oscillator s transmission can be heard by all other oscillators . thus , for our model we ignore pathloss , i.e. @xmath122 , to allow any node s transmission to be heard by each of the other @xmath123 nodes . * delay model : the second assumption is that there is instantaneous coupling . this assumption is the same as setting @xmath124 . in such a situation we would use our pathloss only model . * synchronization pulses : the third key assumption made in @xcite is that there is non - uniform coupling , meaning that each of the @xmath9 oscillators fire with strengths @xmath125 . we modify the parameters in our model by making node @xmath18 transmit with magnitude @xmath126 . they also assume that any two pulses transmitted at different times will be seen by an oscillator as two separate pulses . in our model , we may choose any pulse @xmath28 that has an arbitrarily short duration and each node will detect the pulse arrival time and pulse magnitude . * clock model : the fourth important assumption made by mirollo and strogatz is that the oscillators are identical but they start in arbitrary initial conditions . we simplify our clock model in ( [ eq : clock ] ) by eliminating any timing jitter , i.e. @xmath127 , and making the clocks identical by setting @xmath128 for @xmath129 . we leave @xmath47 in the model to account for the arbitrary initial conditions . we also assume that the phase variable in the pulse - coupled oscillator model increases at the same rate as our clock . that is , the time it takes the phase variable to go from zero to one and the time it takes our clock to count from one integer value to the next are the same . now that we have identical system models , what remains is to modify our model to mimic the coupling action detailed in @xcite . this is accomplished by defining a proper pulse - connection function @xmath119 . to match the coupling action in @xcite , we choose a pulse transmit time function @xmath130 that is a function of pulse receive times and also the time of node @xmath18 s @xmath131th pulse transmission time . @xmath132 is the time in terms of @xmath40 that node @xmath18 receives its @xmath65th pulse since its last pulse transmission at @xmath133 . in this case , @xmath119 will be a function that updates node @xmath18 s @xmath120th pulse transmission time each time node @xmath18 receives a pulse . let @xmath134 where it is node @xmath18 s @xmath120th pulse transmission time after observing @xmath65 pulses since its last pulse transmission . node @xmath18 will transmit its pulse as soon as @xmath135 where @xmath121 is node @xmath18 s current time . as soon as the node transmits a pulse at @xmath119 the function will reset and become @xmath136 . the node is now ready to receive pulses and at its first received pulse , the next transmission time will become @xmath137 . @xmath119 will thus be defined as @xmath138 , \qquad k>0\\ x_{n , i}^{c_{i}}(0 ) & = & x_{n-1,i}^{c_{i}}+1 \label{eq : pco_estimator2}\end{aligned}\ ] ] where the pulse received at @xmath132 is a pulse of magnitude @xmath139 and the function @xmath1\to [ 0,1]$ ] is the smooth , monotonic increasing , and concave down function defined in @xcite . equations ( [ eq : pco_estimator ] ) and ( [ eq : pco_estimator2 ] ) fundamentally say that each time node @xmath18 receives a pulse , node @xmath18 s next transmission time will be adjusted . this is in line with the behavior of the coupling model described by mirollo and strogatz since each time an oscillator receives a pulse , its state variable is pulled up by @xmath140 thus adjusting the time at which the oscillator will next fire . to see how equations ( [ eq : pco_estimator ] ) and ( [ eq : pco_estimator2 ] ) relate to the coupling model in @xcite , let us consider an example with two pulse coupled oscillators . consider two oscillators @xmath57 and @xmath141 illustrated in fig . [ fig : strogatzmodel ] . in fig . [ fig : strogatzmodel](a ) , we have that oscillator @xmath57 is at phase @xmath142 and oscillator @xmath141 is just about to fire . below the pulse - coupled oscillator model we have a time axis for node @xmath18 corresponding to our clock model going from time @xmath133 to @xmath143 . our time axis for node @xmath18 models the behavior of oscillator @xmath57 , that is , we want node @xmath18 to behave in the same way as oscillator @xmath57 under the influence of oscillator @xmath141 . if oscillator @xmath141 did not exist , then the phase variable @xmath142 will match our clock in that @xmath142 reaches @xmath42 at the same time our clock reaches @xmath144 and oscillator @xmath57 will fire at the same time our model fires . in fig . [ fig : strogatzmodel](b ) , oscillator @xmath141 has fired and has pulled the state variable of oscillator @xmath57 up by @xmath140 . this coupling has effectively pushed the phase of oscillator @xmath57 to @xmath145 and decreased the time before @xmath57 fires . in fact , the time until oscillator @xmath57 fires again is decreased by @xmath56 . we can capture this coupling in our model since we can calculate the lost time @xmath56 . the time at which oscillator @xmath141 fires is @xmath146 and it is clear that @xmath147 . thus , if the time that oscillator @xmath57 will fire again is decreased by time @xmath56 due to the pulse of @xmath141 , then we adjust our node firing time by decreasing the firing time to @xmath148 . this is exactly the expression in ( [ eq : pco_estimator ] ) for @xmath149 . this relationship between our model for calculating the node firing time and the pulse - coupled oscillator coupling model can be easily extended to @xmath9 oscillators . we can see then that the pulse - coupled oscillator model proposed by mirollo and strogatz in @xcite is a special case of our model . our model generalizes this pulse - coupled oscillator model by considering timing jitter , pulses of finite width , propagation delay , non - identical clocks , and an ability to accommodate arbitrary coupling functions . just as we could specialize our model to the pulse - coupled oscillator model of mirollo and strogatz , we now specialize the model for our proposed synchronization technique . we start under the assumption of no propagation delay and develop the synchronization technique for this case . propagation delay is considered in section [ sec : timedelay ] . we proceed in three steps . in section [ sec : signal - reception ] , we specify the model for @xmath150 , the received waveform at any node @xmath29 . second , in section [ sec : pulseproperties ] , we prove that given certain characteristics of the model , @xmath150 has very useful limiting properties . third , we show in section [ sec : timesynchronization ] that estimators ( i.e. , the pulse connection function ) developed for our synchronization technique give @xmath150 the desired properties . for our synchronization technique , we specialize the general model by making the following assumptions on @xmath51 and @xmath52 for @xmath151 : * a characterization of the @xmath152 is given by a known function @xmath153 with @xmath154 $ ] that gives the percentage of nodes with any given @xmath155 value . thus , the fraction of nodes with @xmath155 values in the range @xmath156 to @xmath157 can be found by integrating @xmath153 from @xmath156 to @xmath157 . we assume that @xmath158 , for some constant @xmath159 . we keep this function constant as we increase the number of nodes in the network ( @xmath37 ) . given any circle of radius @xmath95 that intersects the network , the nodes within that circle will have @xmath51 s that are characterized by @xmath153 . @xmath95 is the maximum @xmath56 such that @xmath160 . this means that the set of nodes that any node @xmath29 will hear from will have its @xmath51 s characterized by a known function . note that @xmath95 can be infinite , and in that case , any node @xmath29 hears from all nodes in the network . fundamentally , @xmath153 means that as we increase node density , the new nodes have @xmath155 parameters that are well distributed in a predictable manner . * @xmath52 is a zero mean gaussian process with samples @xmath161 , for any @xmath162 , and independent and identically distributed samples for any set of times @xmath163 $ ] , @xmath65 a positive integer . we assume @xmath164 and note that @xmath165 is defined in terms of the clock of node @xmath18 . we assume that @xmath52 is gaussian since the rms ( root mean square ) jitter is characterized by the gaussian distribution @xcite . we maintain the full generality of the pathloss model from section [ sec : prop - model ] . note that throughout this work we assume no transmission delay or time - stamping error . this means that a pulse is transmitted at exactly the time the node intends to transmit it . we make this assumption since there will be no delay in message construction or access time @xcite because our nodes broadcast the same simple pulse without worrying about collisions . also , when a node receives a pulse it can determine its clock reading without delay since any time stamping error is small and can be absorbed into the random jitter . for our proposed synchronization technique , the aggregate waveform seen by node @xmath29 at any time @xmath78 is @xmath166 where @xmath150 is the waveform seen at node @xmath29 written in the time scale of @xmath43 and @xmath167 for all @xmath18 . @xmath35 is the maximum transmit magnitude of a node . @xmath33 is the random timing offset suffered by the @xmath18th node , which encompasses the random clock jitter and estimation error . this model says that each node @xmath18 s pulse transmission occurs at the ideal transmit time @xmath32 plus some random error @xmath33 . in the next section , section [ sec : pulseproperties ] , we find properties for @xmath33 that will give us desirable properties in @xmath150 . then , in section [ sec : timesynchronization ] , we show that our proposed steady - state synchronization technique and its associated pulse - connection function will give us the desired properties . there are two comments about ( [ eq : timesync - aggwaveform ] ) that we want to make . first , note that even though we sum the transmissions from all @xmath9 nodes in ( [ eq : timesync - aggwaveform ] ) , we do not assume that node @xmath29 can hear all nodes in the network . recall from the pathloss model that if we have a multi - hop network , then there will be a nonzero probability that @xmath168 . thus , node @xmath29 will not hear from the nodes whose transmissions have zero magnitude . second , it may be possible that the nodes are told that there are @xmath169 nodes in the network while the actual number of functioning nodes is @xmath9 . in which case , each node will transmit with signal magnitude @xmath170 and ( [ eq : timesync - aggwaveform ] ) will have a factor of @xmath171 . other than for this factor , however , the theoretical results that follow are not affected . to model the quality of the reception of @xmath150 by node @xmath29 , we model the reception of a signal by defining a threshold @xmath172 . @xmath172 is the received signal threshold required for nodes to perfectly resolve the pulse arrival time . if the maximum received signal magnitude is less than @xmath172 then the node does not make any observations and ignores the received signal waveform . we assume that @xmath173 . in our work we will assume that @xmath28 takes on the shape @xmath174 where @xmath175 is expressed in terms of @xmath43 . we assume @xmath176 for @xmath177 , @xmath178 only on @xmath179 , @xmath180 , and @xmath181 is uniformly continuous on @xmath182 . thus , we see that @xmath28 has at most three jump discontinuities ( at @xmath183 ) . @xmath184 should be chosen large compared to @xmath185 , i.e. @xmath186 , where @xmath187 is the value of @xmath165 translated from the time scale of @xmath40 to @xmath43 . this way , over each synchronization phase , with high probability a zero - crossing will occur . for each node , the duration in terms of @xmath43 of a synchronization phase will be @xmath188 . note that we assume @xmath184 is a value that is constant in any consistent time scale . this means that even though nodes have different clocks , identical pulses are transmitted by all nodes . we define a pulse to be transmitted at time @xmath78 if the pulse makes a zero - crossing at time @xmath78 . similarly , we define the _ pulse receive ( arrival ) time _ for a node as the time when the observed waveform first makes a zero - crossing . a _ zero - crossing _ is defined for signals that have a positive amplitude and then transition to a negative amplitude . it is the time that the signal first reaches zero . for the exchange of synchronization pulses , we assume that nodes can transmit pulses and receive signals at the same time . this simplifying assumption is not required for the ideas presented here to hold , but simplifies the presentation . we mention a way to relax this assumption in section [ sec : nosimulttxrx ] . in ( [ eq : timesync - aggwaveform ] ) and in the discussions above , we have focused on characterizing the aggregate waveform for any one synchronization phase . that is , ( [ eq : timesync - aggwaveform ] ) is the waveform seen by any node @xmath29 for the synchronization phase centered around node @xmath42 s transmission at @xmath189 , @xmath32 a positive integer . we can , however , describe a synchronization pulse train in the following form , @xmath190 where @xmath191 is the integer value of @xmath78 at the @xmath192th synchronization phase , and @xmath193 is the error suffered by the @xmath18th node in the @xmath192th synchronization phase . we seek to create this pulse train with equispaced zero - crossings and use each zero - crossing as a synchronization event . an illustration of such a pulse train is shown in fig . [ fig : pulsetrain ] . for simplicity , however , most of the theoretical work is carried out on one synchronization phase . in this section , we characterize the properties of @xmath33 that give us desirable properties in the aggregate waveform . from ( [ eq : timesync - aggwaveform ] ) , the aggregate waveform seen at each node @xmath29 in the network has the form @xmath194 we have dropped the @xmath29 and @xmath43 for notational simplicity since in this section we deal solely with the received waveform at a node @xmath29 in the time scale of @xmath43 . as we let the number of nodes grow unbounded ( @xmath37 ) , the properties of this limit waveform can be characterized by theorem [ theorem : main ] . these properties will be essential for asymptotic cooperative time synchronization . as a note , in theorem [ theorem : main ] we present the case for gaussian distributed @xmath33 but similar results hold for arbitrary zero - mean , symmetrically distributed @xmath33 with finite variance . [ theorem : main ] let @xmath28 be as defined in equation ( [ eq : poft ] ) and @xmath195 with @xmath196 a constant and @xmath197 for all @xmath18 , @xmath141 a constant . also , let @xmath198 be defined as in section [ sec : prop - model ] and be independent from @xmath33 for all @xmath18 . then , @xmath199 has the properties * @xmath200 , * @xmath201 for @xmath202 , and @xmath203 for @xmath204 for some @xmath205 . * @xmath206 is odd around @xmath189 , i.e. @xmath207 for @xmath208 * @xmath209 is continuous . @xmath210 the properties outlined in theorem [ theorem : main ] will be key to the synchronization mechanism we describe . the specific value of @xmath211 will be determined by our choice of the pulse - connection function . before we prove theorem [ theorem : main ] in section [ sec : theoremproof ] we develop and motivate a few important related lemmas . at time @xmath212 , we have that @xmath213 where @xmath214 . we have the mean of @xmath215 being @xmath216 where @xmath217 is the gaussian pdf @xmath218 it is clear that the @xmath215 s , for different @xmath18 s , do not have the same mean and do not have the same variance since the two quantities depend on the @xmath51 value . since the @xmath51 s are characterized by @xmath153 ( defined in section [ sec : systemparameters ] ) , we write the gaussian distribution for @xmath8 as @xmath219 and @xmath215 is in fact a function of @xmath220 as well , denoted @xmath221 . using @xmath222 and @xmath221 , the notation makes it clear that we can average over the @xmath51 s that are characterized by @xmath153 . we use the results of lemmas [ lemma : polarity_positive ] and [ lemma : polarity_negative ] to prove the polarity result for @xmath206 in section [ sec : theoremproof ] . [ lemma : polarity_positive ] given the sequence of independent random variables @xmath215 with @xmath223 , @xmath224 , and @xmath225 . then , for all @xmath18 , @xmath226 @xmath227 for some constants @xmath228 , @xmath229 , and @xmath230 and @xmath231 almost surely , where @xmath232 [ lemma : polarity_negative ] given the sequence of independent random variables @xmath215 with @xmath233 , @xmath224 , and @xmath225 . then , for all @xmath18 , @xmath234 @xmath235 for some constants @xmath228 , @xmath229 , and @xmath230 and @xmath236 almost surely , where @xmath237 the results of lemma [ lemma : polarity_positive ] and lemma [ lemma : polarity_negative ] are intuitive since given that @xmath28 is odd and the gaussian noise distribution is symmetric , it makes sense for @xmath206 to have properties similar to an odd waveform . since the proofs of the two lemmas are very similar , we only prove lemma [ lemma : polarity_positive ] . the proof can be found in the appendix . knowing only the polarity of @xmath206 is not entirely satisfying since we would also expect that the limiting waveform be continuous . the proof of lemma [ lemma : continuity1 ] is once again left for the appendix . [ lemma : continuity1 ] using @xmath28 in ( [ eq : poft ] ) , @xmath238 is a continuous function of @xmath78 , where @xmath239 we can proceed in a straightforward manner to show that @xmath240 . for @xmath241 , @xmath242 where @xmath243 . since our goal is to apply some form of the strong law of large numbers , we first examine the mean of @xmath244 . we have that @xmath245 @xmath246 @xmath247 . furthermore , @xmath248 since @xmath249 is odd and @xmath217 is even because it is zero - mean gaussian . thus , @xmath250 . we next consider the variance of @xmath244 : @xmath251 where we have used the fact that @xmath252 and @xmath253 . from the preceding discussion we see that the @xmath244 s are a sequence of zero mean , finite ( but possibly different ) variance random variables . from stark and woods @xcite , we know that if @xmath254 , then we have strong convergence of the @xmath244 s : @xmath255 with probability-1 as @xmath37 . but it is easy to see that @xmath256 so the condition is satisfied . as a result , @xmath257 as @xmath258 . we have that @xmath206 is continuous from lemma [ lemma : continuity1 ] . thus , next we need to show that @xmath201 for @xmath259 , and @xmath203 for @xmath260 for some @xmath261 . we show the case for @xmath262 by simply applying lemma [ lemma : polarity_positive ] . since lemma [ lemma : polarity_positive ] holds for all @xmath223 , there clearly exists a @xmath263 such @xmath264 for @xmath259 . the case for @xmath265 comes similarly from lemma [ lemma : polarity_negative ] . lastly , it remains to be shown that @xmath206 is odd around @xmath189 . this , however , is evident from the form of @xmath266 . since @xmath222 is even in @xmath267 about @xmath41 and @xmath249 is odd about @xmath41 , it is clear that @xmath268 as a function of @xmath78 is odd about @xmath32 . thus , @xmath266 is odd around @xmath32 . this then completes the proof for theorem [ theorem : main ] . in this work we want to show that as we let @xmath37 then we can recover deterministic parameters that allow for time synchronization . such a result would provide rigorous theoretical support for a new trade - off between network density and synchronization performance . to simplify the study , we focus on the steady - state time synchronization properties of asymptotically dense networks . in particular , we develop a cooperative technique that constructs a sequence of equispaced zero - crossings seen by all nodes which allows the network to maintain time synchronization indefinitely given that the nodes start with a collection of equispaced zero - crossings . starting with a few equispaced zero - crossings allows us to avoid the complexities of starting up the synchronization process but still allows us to show that spatial averaging can be used to average out timing errors . if we are able to maintain indefinitely a sequence of equispaced zero - crossing using cooperative time synchronization , then it means that spatial averaging can average out all uncertainties in the system as we let node density grow unbounded . this recovery of deterministic parameters is our desired result . here , we overview the estimators needed for cooperative time synchronization . let @xmath269 be the time , with respect to clock @xmath270 , that the @xmath18th node sees its @xmath120th pulse . in dealing with the steady - state properties , we start by assuming that each node @xmath18 in the network has observed a sequence of @xmath271 pulse arrival times , @xmath272 , that occur at integer values of @xmath78 , @xmath271 is an integer . recall that @xmath272 is defined as a set of @xmath271 pulse arrival times in the time scale of @xmath40 . therefore , even though @xmath272 occur at integer values of @xmath78 ( the time scale of @xmath43 ) , these values are not necessarily integers since they are in the time scale of @xmath40 . note also that in our model the pulse arrival time is a zero - crossing location . using these @xmath271 pulse arrival times , each node @xmath18 has two distinct , yet closely related tasks . the first task is time synchronization . to achieve time synchronization , node @xmath18 wants to use these @xmath271 pulse arrival times to make an estimate of when the next zero - crossing will occur . if it can estimate this next zero - crossing time , then it can effectively estimate the next integer value of @xmath78 . this estimator can then be extended to estimate arbitrary times in the future which gives node @xmath18 the ability to synchronize to node @xmath42 . the second task is that node @xmath18 needs to transmit a pulse so that the sum of all pulses from the @xmath9 nodes in the network will create an aggregate waveform that , in the limit as @xmath37 , will give a zero - crossing at the next integer value of @xmath78 . this second task is very significant because if the aggregate waveform gives the exact location of the next integer value of @xmath78 , then each node @xmath18 in the network can use this new zero - crossing along with @xmath273 to form a set of @xmath271 zero - crossing locations . this new set can then be used to predict the next zero - crossing location as well as node @xmath18 s next pulse transmission time . recall that determining the pulse transmission time is the job of the pulse - connection function @xmath119 . with such a setup , synchronization would be maintained indefinitely . the zero - crossings that always occur at integer values of @xmath78 would provide node @xmath18 a sequence of synchronization events and also illustrate how cooperation is averaging out all random errors . the waveform properties detailed in theorem [ theorem : main ] play a central role in accomplishing the nodes task of cooperatively generating an aggregate waveform with a zero - crossing at the next integer value of @xmath78 . from ( [ eq : waveformgeneralform ] ) , if the arrival time of any pulse at a node @xmath29 is a random variable of the form @xmath274 , where @xmath32 is the next integer value of @xmath78 and @xmath33 is zero - mean gaussian ( or in general any symmetric random variable with zero - mean and finite variance ) , then theorem [ theorem : main ] tells us that the aggregate waveform will make a zero - crossing at the next integer value of @xmath78 . this idea is illustrated in fig . [ fig : example ] . thus , for achieving time synchronization in an asymptotically dense network we need to address two issues . first , we need to develop an estimator for the next integer value of @xmath78 given a sequence of @xmath271 pulse arrival times that occur at integer values of @xmath78 . we will call this the _ time synchronization estimator _ and let us write @xmath275 as the time synchronization estimator that determines the time , in the time scale of @xmath40 , when node @xmath18 predicts it will see its @xmath120th zero - crossing . two , we need to develop the pulse - connection function @xmath119 such that node @xmath18 s transmitted pulse will arrive at a node @xmath29 with the random properties described in theorem [ theorem : main ] . here we establish the conditions for estimating the next pulse arrival time , or equivalently the next integer value of @xmath78 , given @xmath271 pulse arrival times . these conditions apply most directly to the time synchronization estimator @xmath275 since we want to synchronize in some desired manner . the problem of synchronization is the challenge of having the @xmath18th node accurately and precisely predict when the next integer value of @xmath78 will occur . in our setup , the reception of a pulse by node @xmath18 tells it of such an event . let us explicitly model the time at an integer value of @xmath78 in terms of the clock of node @xmath18 . assume @xmath32 is an integer value of @xmath78 and at this time , node @xmath18 will observe its @xmath120th pulse . thus , from ( [ eq : clock ] ) we have that @xmath276 the equation makes use of the clock model of node @xmath18 ( [ eq : clock ] ) to tell us the time at clock @xmath40 when node @xmath42 is at @xmath32 , where @xmath32 is an integer in the time scale of @xmath43 . we are also starting with the assumption that the zero - crossing that occurs at an integer value of @xmath78 is observed by node @xmath18 at this time . from ( [ eq : state - eqns ] ) we see that the pulse receive time at node @xmath18 , @xmath277 , is a gaussian random variable whose mean is parameterized by the unknown vector @xmath278 $ ] . thus , to achieve synchronization node @xmath18 will try to estimate the random variable @xmath277 using a series of @xmath271 pulse receive times as observations ( recall that @xmath271 is known ) . note that the observations are also random variables with distributions parameterized by @xmath279 . we want the time synchronization estimator of node @xmath18 to make an estimate of @xmath277 , denoted @xmath280 which is a function of past observations @xmath281 , that meets the following criteria : @xmath282 = e_{\vartheta}(t_{n , i}^{c_{i } } ) \label{eq : opt1}\ ] ] @xmath283 \label{eq : opt2}\ ] ] for all @xmath279 . the subscript @xmath279 means that the expectation is taken over the distributions involved given any possible @xmath279 . the first condition comes from the fact that given a finite @xmath271 , it is reasonable to want the expected value of the estimate to be the expected value of the random variable being estimated for all @xmath279 . as in the justification for unbiased estimators , this condition eliminates unreasonable estimators so that the chosen estimator will perform well , on average , for all values of @xmath279 @xcite . the second condition is the result of seeking to minimize the mean squared error between the estimate and the random variable being estimated for all @xmath279 . for the time synchronization estimator , node @xmath18 will seek to estimate @xmath277 given @xmath272 . from ( [ eq : state - eqns ] ) , we see that @xmath284^{t}$ ] is a jointly gaussian random vector parameterized by @xmath279 . recall that we assume @xmath52 is a zero mean gaussian process with independent and identically distributed samples @xmath285 , for any @xmath78 . also , since we re assuming that the zero - crossings at node @xmath18 occur at consecutive integer values of @xmath78 , the random variable @xmath286 is gaussian with @xmath287 for some @xmath288 $ ] . we also notice that @xmath289 since each noise sample is independent , we see that the distribution of @xmath290 parameterized by @xmath279 can be written as @xmath291 where @xmath292\ ] ] and @xmath293 . as a result , for any @xmath271 consecutive observations , we can simplify notation by using the model @xmath294 where @xmath295^{t } = [ t_{n - m , i}^{c_{i } } \quad t_{n - m+1,i}^{c_{i } } \dots t_{n-1,i}^{c_{i}}]^{t}$ ] and @xmath296 = \left [ \begin{array}{c } \alpha_{i}(\tau_{0}-m-\bar{\delta}_{i})\\ \alpha_{i } \end{array } \right]\ ] ] with @xmath297^t\ ] ] and @xmath298^{t}$ ] . since @xmath52 is a gaussian noise process , @xmath299 with @xmath293 . using the simplified notation in ( [ eq : simple - obs ] ) , we want to estimate @xmath300 , where @xmath300 is jointly distributed with @xmath301 as @xmath302 \sim { \mathcal n } ( \left [ \begin{array}{c } { \mathbf m}\\ \theta_{1}+m\theta_{2 } \end{array } \right ] , \left [ \begin{array}{cc } \sigma & 0\\ 0 & \sigma^{2 } \end{array } \right]).\ ] ] using this notation , we can rewrite the synchronization criteria as : @xmath303=e_{\theta}(y_{m+1})\ ] ] @xmath304,\ ] ] where @xmath305 is the estimator for @xmath300 . condition ( [ eq : opt1-simple ] ) implies that our estimate must be unbiased . condition ( [ eq : opt2-simple ] ) is equivalent to @xmath306.\ ] ] to see this equivalence , note that @xmath307 } \nonumber \\ & = & e_{\theta}\big[(\hat{y}_{m+1}(y_{1 } , y_{2 } , \dots , y_{m})-(\theta_{1}+m\theta_{2})-w_{m+1 } ) ^{2}\big ] \nonumber \\ & = & e_{\theta}\big[(\hat{y}_{m+1}(y_{1 } , y_{2 } , \dots , y_{m})- ( \theta_{1}+m\theta_{2 } ) ) ^{2}\big]+e\big [ w_{m+1}^{2}\big],\end{aligned}\ ] ] where the last inequality follows from the independence of @xmath308 from all other noise samples . since the distribution of of @xmath308 is independent of @xmath309 , @xmath310 } \nonumber \\ & = & \textrm{argmin}_{\hat{y}_{m+1 } } e_{\theta}\big[(\hat{y}_{m+1}(y_{1 } , y_{2 } , \dots , y_{m})- ( \theta_{1}+m\theta_{2 } ) ) ^{2}\big ] \nonumber.\end{aligned}\ ] ] with these two conditions , from @xcite we see that the desired estimate for @xmath300 will be the uniformly minimum variance unbiased ( umvu ) estimator for @xmath311 . using the above linear model , from @xcite we know the maximum likelihood ( ml ) estimate of @xmath309 , @xmath312 , is given by @xmath313 this estimate achieves the cramer rao lower bound , hence is efficient . the fisher information matrix is @xmath314 and @xmath315 . this means that @xmath312 is umvu . again from @xcite , the invariance of the ml estimate tells us that the ml estimate for @xmath316 is @xmath317 . first , it is clear that @xmath318 , where @xmath319 $ ] . as a result , we first see that @xmath320 so @xmath321 is unbiased . next , to see that @xmath321 is also minimum variance we compare its variance to the lower bound . @xmath322 the extension of the cramer rao lower bound in @xcite to a function of parameters tells us that @xmath323 with @xmath324 . in this case , @xmath325 $ ] so the lower bound to the mean squared error is @xmath326 as a result , we see that @xmath321 is umvu . since @xmath321 is the desired estimate of where the next pulse arrival time will be , it is the time synchronization estimator . thus , @xmath327 note that @xmath328 has a variance that goes to zero as @xmath329 . we now need to develop the pulse - connection function so that the conditions for @xmath33 in theorem [ theorem : main ] are satisfied . recall we are developing the synchronization technique under the assumption of no propagation delay , i.e. @xmath330 . given a sequence of @xmath271 pulse arrival times , the time synchronization estimator @xmath275 given in ( [ eq : timesync - estimator ] ) gives each node the ability to predict the next integer value of @xmath78 . what remains to be considered is the second part of the synchronization process : developing a pulse - connection function @xmath119 such that the aggregate waveform seen by a node @xmath29 will have the properties described in theorem [ theorem : main ] . let us first consider the distribution of @xmath275 . from ( [ eq : timesync - estimator - distribution ] ) , we have that @xmath331 using ( [ eq : clock ] ) , we can translate @xmath332 into the time scale of @xmath43 as @xmath333 which gives @xmath334 this means that @xmath335 under our assumption of @xmath330 , any transmission by node @xmath18 will be instantaneously seen by any node @xmath29 . as a result , the random variable @xmath336 will be seen as the pulse arrival time at node @xmath29 , in the time scale of @xmath43 . due to the assumption of no propagation delay , defining @xmath337 will give us the desired properties in the aggregate waveform . to see this , let us compare the distribution of @xmath338 to the assumptions of theorem [ theorem : main ] . since @xmath32 is the ideal crossing time in the time scale of @xmath43 , we have @xmath339 therefore , we see that @xmath340 where @xmath211 from theorem [ theorem : main ] is @xmath341 we have shown that using the pulse connection function @xmath337 satisfies the conditions of theorem [ theorem : main ] . thus , all the results of the theorem apply . as a result , we have established a time synchronization estimator @xmath336 and a pulse - connection function @xmath338 . in the case of @xmath330 , we have that @xmath337 , or in the time scale of @xmath40 , @xmath342 . when each node in the network uses the pulse - connection function @xmath343 we have a resulting aggregate waveform that has a zero - crossing at the next integer value of @xmath78 as @xmath37 . this fact follows from applying theorem [ theorem : main ] . thus , we have an asymptotic steady - state time synchronization method that can maintain a sequence of equispaced zero - crossings occurring at integer values of @xmath78 . an interesting feature of this synchronization technique is that no node needs to know any information about its location or its surrounding neighbors . before ending this section , let us comment on the assumption of simultaneous transmission and reception . one way to relax this assumption is to divide the network into two disjoint sets of nodes , say the odd numbered nodes and the even numbered nodes , where each set is still uniformly distributed over the area . then , the odd nodes and the even nodes will take turns transmitting and receiving . for example , the odd numbered nodes can transmit pulses at odd values of @xmath78 and the even numbered nodes will listen . the even numbered nodes will then transmit pulses at the even values of @xmath78 and the odd numbered nodes will listen . with such a scheme , nodes do not transmit and receive pulses simultaneously , but can still take advantage of spatial averaging . the odd numbered nodes will see an aggregate waveform generated by a subset of the even numbered nodes and the even numbered nodes will receive a waveform cooperatively generated by the odd numbered nodes . let us take a more detailed look at this scheme . in fig . [ fig : nosimulttxrx ] we assume that @xmath32 is an even integer value of @xmath78 and use @xmath344 . each even numbered node will use the aggregate signals occurring at @xmath345 , @xmath346 , and @xmath347 to estimate @xmath32 and cooperatively the even nodes will generate the aggregate signal at @xmath32 . the odd numbered nodes will then use the aggregate signals occurring at @xmath348 , @xmath349 , and @xmath32 to generate the aggregate signal at @xmath350 . therefore , the odd and even numbered nodes can take turns transmitting and receiving signals and nodes never need to simultaneously transmit and receive . of course , such a setup would require a modification of the estimators used by the nodes . nodes will receive a vector of @xmath271 observations @xmath351 with @xmath352 = \alpha_{i}(\tau_{0}+1 - 2(m - l)-\bar{\delta}_{i})+\psi_{i}$ ] for @xmath353 . with such a mechanism , the @xmath354 matrix in equation ( [ eq : simple - obs ] ) would change to @xmath355^t\ ] ] and @xmath309 becomes @xmath296 = \left [ \begin{array}{c } \alpha_{i}(\tau_{0}+1 - 2m-\bar{\delta}_{i})\\ \alpha_{i } \end{array } \right].\ ] ] to estimate the location @xmath32 in the time scale of @xmath40 , we can proceed as in section [ sec : timesyncestimator ] : @xmath356 will be distributed @xmath357 and @xmath312 is umvu . this leads to the umvu estimate @xmath358 , where @xmath359 $ ] , and @xmath360 . in this case , the variance of @xmath321 will be @xmath361 , and thus we have that @xmath362 converted to the time scale of @xmath43 we have @xmath363 comparing equations ( [ eq : comptonosimult ] ) and ( [ eq : nosimult ] ) , we see that they have the same form . as a result , we can again set @xmath342 and achieve cooperative time synchronization . we now extend the ideas of cooperative time synchronization to the situation where signals suffer not only from pathloss but also propagation delay . it turns out that the effect of propagation delay can also be addressed using the concept we have been using throughout this paper averaging out errors using the large number of nodes in the network . in this section , we use the pathloss and propagation delay model detailed in section [ sec : delay - model ] . we introduce a time delay function @xmath83 . for generality , we explicitly model a multi - hop network where we have a @xmath53 function that is zero for @xmath56 greater than some distance @xmath95 , i.e. @xmath364 for @xmath365 . such a model implies that the aggregate signal seen at any node @xmath29 is influenced only by the set of nodes inside a circle of radius @xmath95 centered at node @xmath29 . with this we can effectively divide the network into two disjoint sets , a set of _ interior nodes _ and a set of _ boundary nodes_. an interior node @xmath29 is defined to be a node whose distance from the nearest network boundary is greater than or equal to @xmath95 . a boundary node is thus defined to be a node that is a distance less than @xmath95 away from the nearest network boundary . we make this distinction since the synchronization technique for each set of nodes is different . please note that if a pathloss function where @xmath364 for @xmath365 is unreasonable , then we simply choose @xmath95 to be infinite and consider all nodes in the network to be boundary nodes . using the propagation delay model , @xmath116 will obviously modify the general received aggregate waveform seen at any node @xmath29 . in fact , equation ( [ eq : timesync - aggwaveform ] ) will now be written as @xmath366 for @xmath9 large , this model will give an accurate characterization of the aggregate waveform seen at node @xmath29 . from equation ( [ eq : timesync - aggwaveform - withdelay ] ) , it is clear that the aggregate waveform will not have a zero - crossing at @xmath32 for every node @xmath29 because of the presence of the @xmath116 random variables . therefore , to average out propagation delay , the idea we employ is to have each node introduce a _ random _ artificial time shift that counteracts the effect of the time delay random variable . more precisely , we want to introduce another random variable @xmath367 such that @xmath368 will have zero mean and a symmetric distribution . at the same time , we assume each node knows @xmath97 and @xmath99 and will also introduce an artificial scaling factor @xmath369 to simplify the analysis of the aggregate waveform . this means that instead of using the scaling factor @xmath370 , each node @xmath18 will scale its transmitted pulse by @xmath371 . for the motivation in this section , let us assume that node @xmath29 is an interior node . to find the distribution of @xmath367 , we consider the following . @xmath82 has density @xmath115 and let @xmath372 be the density of @xmath367 . since @xmath82 and @xmath367 are independent , we know that the density of @xmath373 , @xmath374 , will be the convolution of @xmath115 and @xmath372 . therefore , by the properties of the convolution function , if we set @xmath375 , then we have that @xmath374 is symmetric , i.e. @xmath376 . as well , since @xmath82 has finite expectation , it is easy to see that @xmath377 . given a sequence of @xmath271 zero - crossings that we know to be occurring at integers of @xmath78 , we can still use @xmath336 ( from ( [ eq : timesync - estimator ] ) in the time scale of node @xmath42 ) as the time synchronization estimator . however , with propagation delay , the pulse - connection function will now be @xmath378 . with @xmath367 and @xmath379 included , we can rewrite equation ( [ eq : timesync - aggwaveform - withdelay ] ) as @xmath380 it is important to see that since @xmath82 has the same distribution for _ all _ interior nodes @xmath29 , equation ( [ eq : timesync - aggwaveform - withdelayandfix ] ) holds for every node @xmath29 that is an interior node . this means that for the network to cooperatively generate the waveform in ( [ eq : timesync - aggwaveform - withdelayandfix ] ) each transmit node @xmath18 needs to have the following additional knowledge : ( 1 ) the distribution of @xmath367 whose density is @xmath375 , where @xmath29 is an interior node , and ( 2 ) the functions @xmath97 and @xmath99 to generate @xmath379 . with this knowledge , we can use equation ( [ eq : timesync - aggwaveform - withdelayandfix ] ) to study the aggregate waveform seen at any interior node @xmath29 . in fact , we find that the aggregate waveform has limiting properties that are similar to those outlined in theorem [ theorem : main ] . these properties are described in theorem [ theorem : main - delay ] . [ theorem : main - delay ] let @xmath28 be as defined in equation ( [ eq : poft ] ) and @xmath195 with @xmath196 a constant and @xmath197 for all @xmath18 , @xmath141 a constant . @xmath34 and @xmath116 are defined as in section [ sec : delay - model ] and @xmath367 with density @xmath375 is independent from @xmath116 . @xmath369 and let @xmath116 , @xmath367 , and @xmath33 be mutually independent for all @xmath18 . then , for any interior node @xmath29 with @xmath381 as defined in ( [ eq : timesync - aggwaveform - withdelayandfix ] ) , @xmath382 has the properties * @xmath383 , * @xmath384 is odd around @xmath189 , i.e. @xmath385 for @xmath208 . @xmath210 the proof of theorem [ theorem : main - delay ] is left for the appendix . from the arguments so far , it seems that time synchronization with delay , at least for interior nodes , can be solved simply by modifying the pulse - connection function @xmath386 and changing the scaling factor to @xmath371 . theorem [ theorem : main - delay ] tells us that the limiting aggregate waveform makes a zero - crossing at the next integer value of @xmath78 and the waveform is odd . thus , we can use this zero - crossing as a synchronization event and maintain synchronization in a manner identical to the technique used in the situation without propagation delay . this , however , unfortunately is not the case . in order to implement the above concept , we need to find the random variable , @xmath387 , in the time scale of @xmath40 , that corresponds to @xmath367 such that @xmath388 this means that we need @xmath389 . however , each node @xmath18 can not find @xmath387 that satisfies this since it does not know its @xmath51 . since the @xmath18th node does not know its own value of @xmath51 , to do time synchronization with propagation delay we can have each node estimate its @xmath51 value . however , this estimate will not be perfect and we may no longer have the symmetric limiting aggregate waveform described by theorem [ theorem : main - delay ] . this means that the center zero - crossing might occur some @xmath140 away from @xmath32 , @xmath32 an integer value of @xmath78 . however , steady - state time synchronization can be maintained if the network can use a sequence of @xmath271 equispaced zero - crossings that occur at @xmath390 , where @xmath32 is an integer value of @xmath78 , to cooperatively generate a limiting aggregate waveform that has a zero - crossing at @xmath391 . in such a situation , the network will be able to construct a sequence of equispaced zero - crossings and maintain the occurrence of these zero - crossings indefinitely . the idea is the same as in the case without propagation delay , but the only difference here would be that the zero - crossings do not occur at integer values of @xmath78 . let us give a more formal description of this idea . using notation from section [ sec : timesyncestimator ] , we start with the assumption that each interior node @xmath18 has a sequence of @xmath271 observations that has the form @xmath392 where @xmath393 and @xmath140 is known . to develop the time synchronization estimator @xmath394 and the pulse - connection function @xmath343 , we consider the observations made by each node . if we assume that each node knows the value of @xmath140 , the vector of observations can be written as in ( [ eq : simple - obs ] ) @xmath395 where the matrix @xmath396 in this case is @xmath397^t.\ ] ] using this model , we can follow the development in section [ sec : timesyncestimator ] to find the the time synchronization estimator @xmath398 where @xmath399 $ ] . this estimator will give each node the ability to optimally estimate the next integer value of @xmath78 . note that the variance of the time synchronization estimator is @xmath400 using the time synchronization estimator , we can choose the pulse - connection function as @xmath401 where each time node @xmath18 makes the estimate @xmath402 it also estimates @xmath403 as @xmath404 @xmath405 $ ] . we find that @xmath406 . since , from section [ sec : propdelay_motivation ] , we know we want @xmath389 , we have set @xmath407 . notice that since @xmath387 is simply a realization of @xmath367 multiplied by node @xmath18 s estimate of @xmath51 , node @xmath18 can use the realization of @xmath367 and find @xmath369 . with our choice of @xmath343 in ( [ eq : pulse - connection - propdelay ] ) , we see that @xmath408 where @xmath409 , and @xmath410 . because of the random factor @xmath411 , we see that @xmath412 is no longer a symmetric distribution . as a result , the limiting aggregate waveform @xmath413 may not have a zero - crossing at @xmath189 . thus , if we can find an @xmath140 such that each node @xmath18 using a set of observations of the form ( [ eq : observation_form ] ) allows the network to cooperatively generate the waveform in ( [ eq : aggregate - waveform - timedelay ] ) that has its zero - crossing occurring at @xmath414 ( in the time scale of @xmath43 ) , then we have steady - state time synchronization . this is because the network would be able to use a sequence of @xmath271 observations to generate the next observation that gives the same information as any of the previous observations . thus , by always taking the @xmath271 most recent observations , the process can continue forever and maintain synchronization . each node @xmath18 would need to know distribution of @xmath367 , the value of @xmath140 , and the functions @xmath97 and @xmath99 . therefore , we find that steady - state time synchronization of the interior nodes is possible under certain conditions . as a note , no interior node needs to know any location information . before we consider the synchronization of boundary nodes , we note that the key requirement for each boundary node @xmath18 is to have a pulse - connection function given in equation ( [ eq : pulse - connection - propdelay ] ) . the reason that this must be the pulse - connection for every boundary node @xmath18 is because the analysis for the interior nodes assumes that the aggregate waveform seen by any interior node @xmath29 is created by pulse transmissions occurring at a time determined by ( [ eq : pulse - connection - propdelay ] ) . since the aggregate waveform seen by some interior nodes are created by pulse transmissions from boundary nodes , each boundary node must have the appropriate pulse - connection function . this requirement , however , proves to be extremely problematic and reveals a limitation of the elegant technique of averaging out timing delay when we come to boundaries of the network . the problem comes because @xmath415 already does not have a symmetric distribution if @xmath29 is a boundary node . recall that @xmath416 when @xmath29 is an interior node and @xmath417 when @xmath29 and @xmath77 are both interior nodes . however , @xmath418 when @xmath29 is an interior node and @xmath77 is a boundary node . as a result , @xmath415 is no longer symmetric if @xmath29 is a boundary node . in fact , it is clear that the distribution of @xmath415 is a function of node @xmath29 s location near the boundary . because of this additional asymmetry , let us assume for a moment that the sequence of zero - crossings observed by boundary node @xmath18 occur @xmath419 away from an integer value of @xmath78 . that is , if every node in the network , including the boundary nodes , transmitted a sequence of pulses where each pulse was sent according to ( [ eq : pulse - connection - propdelay ] ) , then boundary node @xmath18 would observe the sequence of observations @xmath420 where @xmath393 and @xmath419 is known . this boundary node @xmath18 could then use the time synchronization estimator given by ( [ eq : timesync - estimator - propdelay ] ) but where the matrix @xmath396 is now replaced with @xmath421 @xmath422^t.\ ] ] thus , for this boundary node @xmath18 we have @xmath423 in this case , however , the variance of the time synchronization estimator depends on @xmath419 @xmath424 the fact that the variance depends on @xmath419 is the root of the problem . the pulse - connection function @xmath425 is _ not _ the same as that given by ( [ eq : pulse - connection - propdelay ] ) . to correct for this , we can make the strong assumption that each boundary node @xmath18 knows is own @xmath51 . we address the reasoning behind this assumption in section [ sec : propdelay_assumption ] . if we use this assumption , then each boundary node @xmath18 can get an observation sequence of the form ( [ eq : observation_form ] ) simply by adding @xmath426 to each of the @xmath271 observations of the form given in ( [ eq : observation_form_boundary ] ) , where we assume that node @xmath18 knows both @xmath140 and @xmath419 . with such an observation sequence , boundary node @xmath18 will have the time synchronization estimator ( [ eq : timesync - estimator - propdelay ] ) and , more importantly , the pulse - connection function ( [ eq : pulse - connection - propdelay ] ) . thus , maintaining time synchronization for the case of propagation delay would be possible . what we have then is that boundary node synchronization would require only the boundary nodes to know their @xmath51 parameters . with this strong assumption only for the boundary nodes , the network is effectively synchronized . even though the boundary nodes do not see the same zero - crossing as the interior nodes , they can calculate this time and thus have all the required synchronization information . the assumption that each boundary node @xmath18 knows @xmath51 is a strong assumption . even though the fraction of nodes that are boundary nodes is small for multi - hop networks requiring many hops to send information across the network , we believe that the assumption is still very artificial . there are two reasons that we make the assumption for the presentation of results on time synchronization with propagation delay . first , the assumption allows us to give an elegant presentation of the main concept of this paper which is to use high node density to average out errors in the network . throughout this work we have used high node density to average out inherent errors present in the nodes . we were able to average out random timing jitter that is present in each node and provide the network with a sequence of zero - crossings that can serve as synchronization events . we then applied this technique to averaging out the errors introduced by time delay . to this end we were partially successful in that the interior nodes can average out these errors assuming the boundary nodes have additional information . but this is of interest since the goal of this paper is to understand the theory of spatial averaging for synchronization and discover its fundamental advantages and limitations . second , the problem encountered at the boundaries is one that opens up an entirely new area of study which is the target of our future work . the issue that we encounter is that the waveform seen by some nodes in the network will have a zero - crossing that is shifted from the ideal location . this implies that different nodes will observe different zero - crossings . furthermore , these zero - crossings will now evolve in time since we do not have the same observations over the entire network . this problem is similar to what we encounter if we consider finite sized networks . for finite @xmath9 , the zero - crossing location will be random and thus introduce another source of error . as well , different nodes will see different zero - crossing locations . therefore , we will turn our attention to the case of finite @xmath9 and develop a different set of tools that will be needed to understand what types of synchronization are achievable under the situation where zero - crossing locations evolve in time . using this understanding , we hope to return to the issue of propagation delay in asymptotically dense networks and characterize the behavior of the network . to conclude , we revisit the scalability issue under the light of work developed in this paper . in the introduction ( section [ sec : estimate - params ] ) , we mentioned that most existing proposals for time synchronization suffer from an inherent scalability problem . the problem with those existing proposals lies in the fact that synchronization errors accumulate : if node 2 can synchronize to node 1 with some small error , and node 3 can synchronize to node 2 with the same small error , these errors accumulate , and the synchronization of node 3 to node 1 is worse . therefore , synchronization error increases with the number of hops in the network , and this problem is especially apparent in the regime of high densities . to make these ideas precise , we first determine the maximum number of hops over which synchronization information must travel and then study how the error in a generic pairwise synchronization mechanism depends on this number of hops . to obtain an estimate for the maximum number of hops @xmath427 in a network in the regime of high densities ( fixed area , @xmath37 ) , we approximate the transmission range of a node by the minimum required transmission distance , @xmath428 , to maintain a fully connected network with high probability . from @xcite , we have that for @xmath9 nodes uniformly distributed over a @xmath429\times [ 0,1]$ ] square , the graph is connected with probability-1 as @xmath37 if and only if each node s transmission distance @xmath428 is such that @xmath430 for some @xmath431 . let us , therefore , approximate @xmath428 as @xmath432 thus , @xmath433 , and thus @xmath434 as @xmath37 . now , we assume there are @xmath435 nodes arranged in a linear ordering , numbered @xmath42 to @xmath435 . to synchronize , each node @xmath18 forms an estimate of its own @xmath51 , based on @xmath271 pulses transmitted from node @xmath436 . as before , node @xmath42 will have the reference clock @xmath44 . node @xmath42 starts by sending @xmath271 pulses at times @xmath437 for @xmath393 . as a result , node @xmath0 will get a vector of observations @xmath438 , where @xmath439 = \alpha_{2}(\tau_{1}-\bar{\delta}_{2})+\psi_{2}$ ] and the @xmath440th element of @xmath438 is @xmath441=\alpha_{2}(\tau_{1}-\bar{\delta}_{2})+l\alpha_{2}+\psi_{2}$ ] . this is similar to the situation we had in ( [ eq : simple - obs ] ) and we can therefore estimate @xmath442 using @xmath443 where @xmath444 $ ] . we find that @xmath445 . node @xmath0 will now transmit @xmath271 pulses at times , in terms of @xmath446 , @xmath447 , for @xmath393 . note that @xmath448 is now a fixed value since node @xmath0 has estimated @xmath442 . in terms of @xmath43 , these pulses occur at @xmath449 for @xmath393 , where @xmath450 . thus , if we translate these times into the time scale of @xmath451 , we will have the vector of observations , @xmath452 , made by node @xmath453 . we find that the @xmath440th element of @xmath452 is @xmath454 = \alpha_{3}((\tau_{2 } + l\frac{\hat{\alpha}_{2}}{\alpha_{2}}- \frac{\psi_{2}}{\alpha_{2}})-\bar{\delta}_{3})+\psi_{3 } \sim { \mathcal n}\bigg(\alpha_{3}(\tau_{2}-\bar{\delta}_{3})+l\alpha_{3}\frac{\hat{\alpha}_{2}}{\alpha_{2 } } , \sigma^{2}\big(\frac{\alpha_{3}^{2}}{\alpha_{2}^{2}}+1\big)\bigg).\ ] ] this vector of observations is of the form @xmath455 where @xmath456 = \left [ \begin{array}{c } \alpha_{3}(\tau_{2}-\bar{\delta}_{3})\\ \alpha_{3}\frac{\hat{\alpha}_{2}}{\alpha_{2 } } \end{array } \right]\ ] ] with @xmath297^t\ ] ] and @xmath298^{t}$ ] . @xmath457 with @xmath458 . with this vector of observations , we can use the estimator @xmath459 where @xmath444 $ ] . we find that @xmath460 if we continue this reasoning , we find that @xmath461 will be the estimate of node @xmath435 . from the above analysis , we see that each node @xmath18 s estimate suffers from jitter variance of the same form . however , there is an accumulation of error because node @xmath18 s estimate has a mean that is dependent on node @xmath436 s estimate . as a result , if node @xmath436 has some small error , then that error will propagate to the estimate of node @xmath18 . a good way to see this is if we consider the special case where @xmath462 . this is the case where the clock frequencies are the same , but nodes do not know this . in this case , we find that node @xmath435 s estimate can be written as @xmath463 where @xmath464 . this is intuitively obvious because node @xmath18 s estimate @xmath403 will be the mean of the gaussian random variable @xmath465 . therefore , it is obvious that the error variance grows linearly with the number of hops . in fact , this behavior is observed in experimental work . with reference broadcast synchronization ( rbs ) , from @xcite the authors find that the synchronization error variance of an @xmath435 hop path is approximately @xmath466 , where @xmath165 is the one hop error variance . therefore , we have that the synchronization error between our two nodes will grow linearly as @xmath467 , which is strictly monotonically increasing . as a result , as @xmath37 , we have that synchronization error will grow unbounded . this scalability problem , however , can potentially be avoided using cooperative time synchronization as @xmath37 . this is because in the limit of infinite density , the cooperative time synchronization technique allows every node in the network to see a set of identical equispaced zero - crossings . as a result , in steady - state the synchronization error does not grow across the network . this comes about by using the high node density to average out random timing errors . thus , we find that cooperative time synchronization has very favorable scalability properties in the limit as @xmath37 . the cooperative synchronization technique described in this paper provides us deterministic parameters that we can use for time synchronization in the limit as node density grows unbounded . in fact , as the node density grows , the observations that can be used for synchronization improve . this means that our cooperative synchronization technique provides an effective trade - off between network density and synchronization performance . such a trade - off has not existed before and will provide network designers an additional dimension over which to improve network synchronization performance . the fundamental idea behind cooperative time synchronization is that by using spatial averaging , the errors inherent in each node can be averaged out . by using observations that are an average " of the information from a large number of surrounding nodes , synchronization performance can be improved due to the higher quality observations . from this point of view , it is clear that the particular technique described in this paper is but one example of using spatial averaging to improve synchronization . other techniques can also be developed using spatial averaging . for example , nodes may not necessarily have to send odd - shaped pulses and use zero - crossing observations . even though this setup takes advantage of the superposition of pulses , it has its drawbacks . to keep the signals in phase , the jitter variance will limit the maximum frequency at which signals can be sent . instead , nodes may transmit ultra wideband pulses . if the nodes surrounding a particular node @xmath29 each transmit an impulse at their estimate of an integer value of @xmath78 , then due to timing errors in the surrounding nodes , node @xmath29 will see a cluster of pulse arrivals around this integer value of @xmath78 . node @xmath29 can then take the sample mean of this cluster of pulses and use that as an observation , just like we used the zero - crossing as an observation in this paper . this idea is illustrated in fig . [ fig : pulsetraincluster ] . such a technique based on ultra wideband pulses will also provide similar scalability properties . as a result , cooperative time synchronization really describes a class of techniques that can take advantage of spatial averaging to improve synchronization performance . with the goal of developing practical cooperative synchronization mechanisms , two keys areas of interest are cooperative synchronization in finite - sized networks and algorithm development . first , the analysis of performance for finite - sized networks is very important . determining when the asymptotic properties presented in this work are good predictors of performance in networks that may be large but still finite in size is important in terms of bridging the gap between our proposed ideas and practical systems . preliminary , simulation - based work along these lines can be found in @xcite . second , developing practical techniques for cooperative time synchronization is essential for implementing spatial averaging in real networks . along these lines , one area of interest is determining what types of pulses should be used , i.e. odd - shaped pulses or ultra wideband pulses . furthermore , the ideas in this paper suggest a few other areas of interest for future work . one is the issue of distributed modulation methods . if we have the ability to generate an aggregate waveform with equispaced zero - crossings , by controlling the location of these crossings we can modulate information onto this waveform and use it to communicate with a far receiver . preliminary work along these lines can be found in @xcite . another issue is to study how the idea of spatial averaging that is so prevalent in this work contributes to synchronization that is observed in nature . * proof of lemma [ lemma : polarity_positive ] . * to show ( [ eqn : cond3a ] ) , we consider @xmath468 since @xmath223 , we have that @xmath469 implying that @xmath249 is shifted to the left and the zero - crossing of @xmath249 occurs at a negative value . @xmath249 is odd about its zero - crossing and @xmath217 is symmetric about zero and strictly monotonically increasing on @xmath470 $ ] for all positive finite variance values . thus , it is clear that @xmath471 which makes @xmath472 . now , the expectation will vary with the variance of @xmath33 and the variance will range from a positive upper bound of @xmath473 to a positive lower bound of @xmath474 , where recall that @xmath211 is a value determined by our choice of the pulse connection function . if we consider @xmath475 to be a function of the variance of @xmath33 , then we see that it is bounded and continuous on the compact domain @xmath476 $ ] . since we showed in the previous paragraph that @xmath472 whenever @xmath33 has a nonzero finite variance , clearly @xmath472 when @xmath477 $ ] . thus , it is clear that @xmath228 and @xmath229 exist and ( [ eqn : cond3a ] ) is shown . to show ( [ eqn : cond3b ] ) , we consider @xmath478 where the second to last inequality follows from the fact that @xmath479 is upper bounded by @xmath42 . the last inequality follows since @xmath480 by the fact that @xmath481 . thus , we have shown ( [ eqn : cond3b ] ) . [ theorem : feller ] the convergence of the series @xmath484 implies that the strong law of large numbers will apply to the sequence of independent random variables @xmath215 . that is , again from @xcite , for every pair @xmath485 , @xmath486 , there corresponds an @xmath9 such that @xmath487 for all @xmath488 . @xmath210 we have shown ( [ eqn : cond3b ] ) so we have @xmath489 . thus @xmath490 and we have convergence by the direct comparison test . therefore , we can apply theorem [ theorem : feller ] and get that for any pair @xmath485 , @xmath486 , we can find an @xmath9 such that @xmath491 for all @xmath488 . by ( [ eqn : cond3a ] ) we have that @xmath492 . thus , we can clearly see that @xmath493 furthermore , since we keep the function @xmath153 constant as we increase the number of nodes in the network we get that @xmath494 converges to a constant @xmath495 given by @xmath496 the above expression comes from the fact that since each @xmath497 is a function of @xmath51 , @xmath494 will converge to the average of the @xmath498 over @xmath153 , the function that characterizes the set of @xmath51 s . therefore , given any @xmath140 , we can find an @xmath499 such that @xmath500 for all @xmath501 . note that since @xmath502 , we have that @xmath503 . since @xmath504 using ( [ eqn : slln ] ) and ( [ eq : means - converge ] ) we have @xmath505 for all @xmath488 , where @xmath506 . thus , we have @xmath507 almost surely . this completes the proof of lemma [ lemma : polarity_positive ] . @xmath210 * proof of lemma [ lemma : continuity1 ] . * first , we start by finding an analytical expression for @xmath508 . from the proof of lemma [ lemma : polarity_positive ] we have that @xmath509 therefore , @xmath508 can be written as @xmath510 f_{t}(\psi , s ) f_{\alpha}(s)d\psi ds| \\ & \leq & a_{max } \int_{\alpha_{low}}^{\alpha_{up}}\int_{-\infty}^{\infty } |p(t-\tau_{o}-\psi)-p(t_{o}-\tau_{o}-\psi)|f_{t}(\psi , s ) f_{\alpha}(s)d\psi ds \\ & = & a_{max } \int_{\alpha_{low}}^{\alpha_{up}}\int_{-\tau_{nz}+t_{o}-\tau_{0}-|t - t_{o}|}^{\tau_{nz}+t_{o}-\tau_{0}+|t - t_{o}| } f_{t}(\psi , s)f_{\alpha}(s ) d\psi ds,\end{aligned}\ ] ] where @xmath511 . the change in the limits of integration in the last equality comes from the fact that @xmath512 outside of @xmath513 $ ] . this is the maximum interval over which @xmath514 can be non - zero . there is no need to take the absolute value of @xmath222 and @xmath153 since they are always non - negative . our second step is to bound the inner integral . before doing so , we first show that the inside integral is in fact riemann integrable . for any given @xmath78 and @xmath515 , the inside integral is taken over a closed interval . over a closed interval , we know from strichartz @xcite that any bounded function that is continuous except at a finite number of points is riemann integrable . furthermore , also from @xcite we know that the sums and products of continuous functions are continuous . as well , if a function is continuous then the absolute value of that function is also continuous . @xmath28 has at most @xmath516 locations at which it is discontinuous and over any open interval not containing a discontinuity , @xmath28 is uniformly continuous since @xmath181 is uniformly continuous . @xmath222 has @xmath517 discontinuities in @xmath267 for an given @xmath220 since it is gaussian for any @xmath220 . and since @xmath518 $ ] , @xmath519 for all @xmath267 and @xmath220 ( @xmath520 occurring when @xmath521 and @xmath522 ) . thus , since @xmath28 and @xmath222 are continuous except at a finite number of points , we see that for given @xmath220 , @xmath78 , and @xmath162 @xmath523 is also continuous in @xmath267 except at a finite number of points ( at most @xmath524 points ) . this function is also bounded since the product of two bounded functions is bounded . as a result , we see that the integral is riemann integrable over any closed interval . we now proceed to bound from above the value of this integral by first bounding the maximum value of the integral assuming no discontinuities and then introducing another term that bounds the maximum area contributed by the discontinuities . if we ignore the discontinuities and assume @xmath28 is uniformly continuous , for any @xmath525 there exists a @xmath526 such that @xmath527 for all @xmath78 and @xmath515 . as a result , @xmath514 can be made as small as desired by choosing the proper @xmath120 thus giving us @xmath528 for all @xmath267 for an appropriate choice of @xmath120 . furthermore , we note that @xmath529 because @xmath253 and @xmath519 . the maximum possible jump at a discontinuity in the function @xmath530 is thus @xmath531 and for any @xmath532 , the maximum area contributed by each discontinuity is @xmath533 . as a result , for all @xmath524 discontinuities , the maximum area contribution will be no more than @xmath534 . for the third term we want @xmath546 this gives us @xmath547 since the only requirement is @xmath536 for @xmath120 chosen by any given @xmath525 , we can always choose @xmath532 as small as desired . thus , this condition can be satisfied . with the second term we want the condition @xmath548 which means that @xmath549 again , this condition can be satisfied since we can choose @xmath550 as large as we want and @xmath532 as small as we want as long as @xmath536 for a given @xmath550 . thus , for any @xmath542 , we first choose @xmath551 . then , we find an @xmath552 such that @xmath553 implies that @xmath554 for all @xmath78 and @xmath515 if we remove the discontinuities in @xmath28 . then , if necessary , @xmath555 is increased to @xmath120 so that @xmath536 implies that @xmath556 and @xmath557 . if no increase is necessary , then @xmath558 . with this choice of @xmath526 , @xmath559 . as a result , for any @xmath271 , we can find an @xmath120 such that @xmath536 implies that @xmath560 . thus , @xmath206 is continuous . * proof of theorem [ theorem : main - delay ] . * let us start by writing ( [ eq : timesync - aggwaveform - withdelayandfix ] ) as @xmath561 where @xmath562 . recall that the dependence on @xmath220 comes from the fact that the density of @xmath33 is a function of @xmath51 which is characterized by @xmath153 . this notation is analogous to the notation used in section [ sec : pulseproperties ] . following the steps in the proof of lemma [ lemma : polarity_positive ] , we can quickly show that the limiting aggregate waveform at node @xmath29 will take on the form @xmath563 where @xmath564 with @xmath565 . therefore , we can prove theorem [ theorem : main - delay ] in two steps : * to show that @xmath266 is odd about @xmath32 , we need to show that @xmath566 is odd in @xmath78 about @xmath32 , i.e. @xmath567 for @xmath568 . * to show a zero - crossing at @xmath32 , show that @xmath569 . we first show that @xmath567 for @xmath208 . using the fact that @xmath570 and @xmath571 , we have the following : @xmath572)\big ) \\ & \stackrel{\small{(a)}}{= } & -e\big(a_{max}g(-d_{fix } ) g(d_{j , i } ) p(-\xi+[t_{i}+d_{fix}+d_{j , i}])\big ) \\ & = & -a_{max}\int_{-\infty}^{\infty } \int_{-\infty}^{0 } \int_{0}^{\infty}g(-y)g(x)p(-\xi+[\psi+y+x])f_{d_{j}}(x ) f_{d_{fix}}(y ) f_{t}(\psi , s)dx dy d\psi \\ & \stackrel{\small{(b)}}{= } & a_{max}\int_{\infty}^{-\infty } \int_{\infty}^{0 } \int_{0}^{-\infty}g(z)g(-u)p(-\xi-[w+z+u])f_{d_{j}}(-u ) f_{d_{fix}}(-z ) f_{t}(-w , s)du dz dw \\ & \stackrel{\small{(c)}}{= } & -a_{max}\int_{-\infty}^{\infty } \int_{-\infty}^{0 } \int_{0}^{\infty}g(-u)g(z)p(-\xi-[w+u+z])f_{d_{j}}(z ) f_{d_{fix}}(u ) f_{t}(w , s)dz du dw \\ & = & -e\big(a_{max}g(-d_{fix } ) g(d_{j , i } ) p(-\xi-[t_{i}+d_{fix}+d_{j , i}])\big ) \\ & = & -e(\tilde{m}_{i}(\tau_{0}-\xi , s)),\end{aligned}\ ] ] where @xmath573 follows because @xmath574 and at @xmath575 we did a change of variables with @xmath576 , @xmath577 , and @xmath578 . @xmath579 follows from @xmath580 and @xmath581 . we thus have @xmath567 for @xmath568 . an - swol hu was born in new york state and grew up in california . he received his b.s . in electrical engineering from stanford university in 2002 . currently he is a ph.d . candidate in the school of electrical and computer engineering at cornell university . his research interests include applied statistics and statistical signal processing , with applications to sensor networks . sergio d. servetto was born in argentina , on january 18 , 1968 . he received a licenciatura en informatica from universidad nacional de la plata ( unlp , argentina ) in 1992 , and the m.sc . degree in electrical engineering and the ph.d . degree in computer science from the university of illinois at urbana - champaign ( uiuc ) , in 1996 and 1999 . between 1999 and 2001 , he worked at the ecole polytechnique federale de lausanne ( epfl ) , lausanne , switzerland . since fall 2001 , he has been an assistant professor in the school of electrical and computer engineering at cornell university , and a member of the field of applied mathematics . he was the recipient of the 1998 ray ozzie fellowship , given to `` outstanding graduate students in computer science , '' and of the 1999 david j. kuck outstanding thesis award , for the best doctoral dissertation of the year , both from the dept . of computer science at uiuc . he is also the recipient of a 2003 nsf career award . his research interests are centered around information theoretic aspects of networked systems , with a current emphasis on problems that arise in the context of large - scale sensor networks .

the study of accretion processes is of fundamental astrophysical importance , being ubiquitous on both the largest and smallest scales , i.e. ranging from accretion by supermassive black holes in the center of galaxies at one end to star and planet formation at the other . here we are interested in studying accretion phenomena in the presence of a large gravitational potential such as that observed in x - ray binaries ( xrbs ) and active galactic nuclei ( agn ) . in these sources , the fundamental timescales of interest , governing the accretion process , scale with the mass of the central object ( @xmath5 ) . hence , studies of the galactic xrb population provide an excellent laboratory for detailed examination of the process of accretion on humanly accessible timescales . the very - high , high - soft and low - hard states ( hereafter vhs , hss & lhs ) are the primary active accretion states observed in xrbs ( see the review by @xcite for a detailed description of accretion states in black hole binaries ) . despite being the most common mode of accretion in black hole x - ray binaries , the nature of the accretion flow in the low - hard state remains uncertain . emission in the lhs is characterized by a hard power - law spectrum ( @xmath0 1.4 1.7 ) and strong x - ray variability ( 30% 40% rms ) . correlated variability at radio / nir / optical wavelengths has also been observed from the black hole systems while in the hard state @xcite . the hard power - law component may be the result of comptonization by a hot optically thin plasma of soft seed photons from a thermal disc or magnetic structures ( through cyclo - synchrotron processes - see , e.g. , @xcite ) . however , the geometry of this plasma is not well understood . a popular model for the accretion geometry in the low - hard state was given by @xcite . in that model the standard thin accretion disc , that dominates in the spectrally soft states is radially truncated and replaced by an advection dominated accretion flow ( adaf : see @xcite ) . the fundamental assumption of a radially recessed accretion disc may find some support in the low disc reflection fractions which are sometimes measured in the hard state ( e.g. , @xcite ) . a number of models have been developed which do not require a recessed disc in the low - hard state . @xcite proposed that black hole states are driven by the height and bulk velocity of magnetic flares above a disc , which remains at the innermost stable circular orbit ( isco ) . these flares would serve to feed a mildly relativistically outflowing corona . low disc reflection fractions do not signal a recessed disc in this model , but result from mild beaming of the hard x - ray flux away from the disc . a similar model for the lhs , based on magnetically dominated coronae , has been proposed by @xcite . these outflowing coronae share some properties with jet based models for the hard component @xcite , in the sense that the latter also produce low reflection fractions @xcite without the need for a recessed accretion disc . a number of recent observations call into question the assumption of a recessed disc in the lhs . in particular gx 339 - 4 was observed during its 2004 outburst at a luminosity of @xmath6 by both _ rxte _ & _ xmm _ @xcite . the observations by _ xmm _ are critical here as they provide coverage at energies below 3 kev whereas _ rxte _ is limited to energies above this . it was found that a cool disc blackbody ( @xmath7 0.35 kev ) consistent with an optically thick geometrically thin accretion disc extending to the innermost stable circular orbit ( isco ) existed , in contrast to theoretical expectations . fits with reflection models revealed reflection fractions @xmath7 0.2 0.3 . previous observations of swift j1753.5 - 0127 yielded similar results for the cool disc component , in this case a disc temperature of @xmath7 0.2 kev was required @xcite ; however , no significant disc reflection features were detected . swift j1753.5 - 0127 was discovered by the _ swift _ burst alert telescope ( bat ) at x - ray and @xmath8-ray energies on 2005 may 30 ( @xcite ) . subsequent observations with the x - ray telescope ( xrt ) revealed a hard power - law spectrum @xcite , while pointed _ rxte _ observations detected 0.6 hz quasi periodic oscillations ( qpo ; @xcite ) . the system was also detected at uv @xcite and optical wavelengths @xcite . radio observations with the _ merlin _ array also detected a variable counterpart @xcite . observations at optical wavelengths by @xcite have also detected a significant modulation with a period of 3.2hrs , which they identify as a superhump period slightly larger than the actual orbital period . this would make swift j1753.5 - 0127 the black hole binary with the shortest known orbital period . in addition to the observations of @xcite above , a number of other high energy studies of this source have been published , which we summarize below ( see fig . [ asmlc ] ) . an analysis of _ rxte _ observations of the outburst was reported in @xcite . the x - ray spectrum was found to be consistent with a power - law ( @xmath9 ) , while low - frequency quasi - periodic oscillations ( qpos ) were also detected ( up to 0.9 hz ) . @xcite analysed simultaneous _ rxte _ & _ integral _ data , which were also obtained during the 2005 outburst . the combined spectrum ( 3 400 kev ) could be fit with a model consisting of thermal comptonization modified by disc reflection ( @xmath10 ) . qpos were also detected here , but at a lower frequency than during the outburst peak ( 0.24 hz ) . to place meaningful constraints on the accretion flow in the lhs one requires sensitivity to both the soft x - rays ( @xmath11 2 kev ) , in order to detect the cool accretion disc , and higher energies in order to detect the most prominent disc reflection features ( 5 7 kev and 20 30 kev ) . _ suzaku _ with its large bandpass and low background , is ideally equipped to carry out these observations . in this paper , we describe observations undertaken with the _ suzaku _ x - ray observatory in 2007 , while swift j1753.5 - 0127 was in the low - hard state . in 2 , we describe our observation and extraction of source spectra and lightcurves . we proceed to analyze the data in 3 , where both phenomenological and more physically motivated models are considered . the broadband spectrum ( 2 250 kev ) is consistent with a simple power - law model , although there are also contributions from the accretion disc . in 4 , these results are discussed in the context of models for the accretion flow in the low - hard state , and finally our conclusions are presented in 5 . swift j1753.5 - 0127 was observed while still in the low - hard state by _ suzaku _ @xcite from 2007 september 19 20:36 ut until september 22 10:30 ut ( obsid:402088010 , pi : homan , see fig [ asmlc ] ) . data were acquired over a broad spectral range ( 0.2 600 kev ) , with the x - ray imaging spectrometer ( xis : @xcite ) and the hard x - ray detector ( hxd : @xcite ) . the source was observed at the xis nominal position for total uncorrected exposure times of @xmath7 95 ks & 82 ks respectively . all data reduction and analysis takes place within the heasoft 6.4.1 environment , which includes ftools 6.4 , suzaku 8.0 and xspec 12.4.0x . the latest versions of the relevant _ suzaku _ caldb files were also used . the xis is installed at the focal plane of the four x - ray telescopes ( xrt : @xcite ) and currently consists of 3 functioning detectors xis0 , xis1 and xis3 . xis0 & xis3 are front illuminated and provide coverage over the energy range 0.4 - 12 kev whereas xis1 is back illuminated in an effort to provide greater sensitivity at lower energies , 0.2 12 kev . the xis has a field of view of @xmath7 18 x 18 ( 1024@xmath12 pixels ) and was operated in 5x5 , 3x3 readout mode . in addition the data were taken in 1/4 window mode in an effort to minimize possible photon pile - up , giving a time resolution of 2s . as the downloaded data products had been processed via the _ suzaku _ pipeline v2.1.6.15 , the data were reprocessed from the raw telemetry files as recommended in the _ suzaku _ data reduction guide ( abc guide ) . standard screening was applied , in particular , we extracted _ asca _ event grades 0:0 , 2:4 , 6:6 with the data filtered to be taken outside of the south atlantic anomaly ( saa ) and where the earth elevation angle was greater than 5@xmath13 . good time interval ( gti ) events were extracted using xselect , where the 3x3 and 5x5 observation mode data for each detector were extracted simultaneously . science images , spectra and backgrounds were then extracted from these event files . even though the observations were carried out in 1/4 window mode , we nonetheless suffered from pileup at the source position . hence the spectra were extracted using an annular extraction region extending from 30 to 250 pixels from the source position . this extraction region size was chosen so as to extract @xmath14 99@xmath15 of the point source flux . as the outer radius of this annulus is larger than the window size ( @xmath7 280 , 295 , 280 pixels respectively ) , the effective extraction region is the intersection of window and the annulus . the resulting extraction region has an area equivalent to 62% of the 250 pixels outer radius , hence , we expect to have detected a commensurate percentage of the total source flux . background spectra were extracted from a source free region of the detector , these are automatically scaled to match the data during the spectral analysis . response files were generated using the tasks xisrmfgen and xissimarfgen . the background and response files were then grouped with the science spectrum for analysis in xspec . the hxd covers the energy range from 10 600 kev , consisting of two separate detectors , ( i ) pin : silicon pin photodiodes covering the energy range 10 70 kev and , ( ii ) gso : gso / bgo phoswich scintillators covering the energy range 40 - 600 kev . due to the arrangement of the instrument , with the pin diodes residing in front of the gso scintillator in each of the 16 detector units that make up the hxd , the raw data do not differentiate between the pin & gso , with this distinction instead being made during the extraction process . as at the time of writing , the gso analysis procedure had not been included in the official _ data reduction pipeline , all of the hard x - ray detector data were reprocessed following the prescription in the abc guide and the 7-step guide . gti science events were extracted using xselect with the appropriate gti files and filter options i.e. , pointing elevation @xmath14 5@xmath13 above earth and excluding data taken near the saa . the relevant events for each detector were then extracted by requiring det_type=1:1 and det_type=0:0 for the pin and gso respectively . the relevant background and response ( pin / gso * 20080129.rsp ) files were obtained from the _ suzaku _ website . a gti between the data and the background file was then created using the ftool mgtime , which was then used to to extract the spectra in xselect . standard corrections were applied to the data , i.e. deadtime for science data , pin background exposure time . as the hxd background files do not include the contribution from the cosmic x - ray background ( cxb ) , the expected cxb was simulated following the recipe detailed in the abc guide , this was then added to the background file . as a check on the accuracy of the background , a separate background estimate was made using the earth occulted data ( earth elevation = 0 ) . this was found to be consistent with the above background . a similar procedure is followed in the case of the gso spectral extraction with the following caveat : the background file exposure time does not need to be corrected , instead one must rebin the science spectrum to match the provided background . the sensitivity of the gso detector is background dominated , hence the high energy detection threshold is determined not by the statistical error but by the reproducibility of the background , here we conservatively estimate the background reproducibility to be 3% ( e.g @xcite ) . in fig . [ xispingso_back ] , we plot the extracted science & background spectra . swift j1753.5 - 0127 is detected out to an energy of 250 kev source and background lightcurves were extracted from the event files using xselect and the appropriate good - time - interval events , after the application of the appropriate baryocentric correction using the ftool aebarycen . the flux from swift j1753.5 - 0127 is observed to remain constant throughout our observation . the mean count rates for each individual detector are approximately : 16.2 , 20.8 , 17.1 , 2.6 , 2.2 counts s@xmath16(xis0 , xis1 , xis3 , pin , gso ) . there is no evidence for any periodic modulation , the lightcurves across all detectors are characterized by rapid variability of a stochastic nature as is expected from accretion processes in the vicinity of a black hole @xcite . this is consistent with rxte power spectra bracketing this data that revealed a power - law slope , @xmath17 . we initially choose to fit the spectra with a number of phenomenological models in an effort to provide a broad characterization of the data . initial spectral fits were made to the entire dataset consisting of 5 spectra in total ( xis0 , xis1 , xis3 , pin , gso ) , spanning the energy range 0.6 250 kev over which swift j1753.5 - 0127 is reliably detected . explicitly , the spectra provide us with data in the following ranges : xis0 , xis3 0.4 10 kev ; xis1 0.2 10 kev ; pin 12 70 kev and gso 50 250 kev . unfortunately , the low energy response of the xis detectors contains a number of uncertain residuals , hence we additionally ignored the regions below 0.6 kev and between 1.7 1.9 kev and 2.1 2.4 kev in all further modelling . initially , the column density was fixed at a value consistent with previous detections ( @xmath18 : @xcite ) , while the normalization was allowed to vary independently for each spectrum . the resulting normalizations were found to be consistent with those expected for the _ suzaku _ detectors . the spectra were fit with a model consisting of an absorbed power - law ( pha*po ) , which provided an acceptable fit except at the lowest energies ( @xmath19 2 kev ) . as the main residual was present at energies below 2 kev , this region was ignored in further fitting . re - applying the power - law fit results in a value for the spectral index of @xmath20 ( @xmath21 ) . it is immediately apparent that a simple phenomenological power - law model suffices to describe the spectrum from 2 250 kev . the data was also fit with a cutoffpl model for comparison . the fit does not indicate the presence of a cut - off in the spectrum with the cutoffpl model fit being significantly inferior to a simple power - law , i.e we find @xmath22 = 6898/6463 ( 6968/6456 ) and @xmath23 = 6962/6462 ( 7048/6455 ) for the @xmath24 models respectively . in both cases the best fit cut - off power - law model requires the high energy cut - off to be 500 kev , i.e the intrinsic upper limit of the cutoffpl model . ignoring the data above 150 kev does not improve the quality of the cutoffpl fit relative to the po fit . as the gso background contains a large feature in the energy range 150 180 kev ( see fig . [ xispingso_back ] ) , we decided to ignore the data beyond 150 kev in all further fitting . furthermore , as there is clearly an additional component contributing at energies below 2 kev , this region is ignored while constraints are placed on the hard x - ray emission . repeating the above power - law fit , we find @xmath25 ( @xmath26 ) unless otherwise explicitly stated ] . inspection of the residuals reveals any contribution from disc reflection to be small , although there is evidence for curvature in the pin spectrum . to investigate the possible contribution due to disc reflection , the best fit power - law from above was convolved with the reflection model of @xcite ( reflect*po ) . the inclination was held fixed ( cos_i _ = 0.45 ) , while the abundances of metals were frozen at the default values ( @xmath27 = 1 ) . the best fit model reveals a slightly softer power - law ( @xmath28 ) in addition to a highly significant ( @xmath29 as determined via an f - test ) reflection fraction , @xmath30 ( @xmath31 ) . as the measured spectral parameters depend on the inclination angle , these and subsequent fits were repeated at a lower inclination angle of 30 degrees . these fits revealed a lower reflection fraction , f @xmath7 0.14 ( see table [ specfit_params ] ) . closer inspection of the xis spectra residuals reveals a slight excess consistent with the presence of a broad fe k line . to place a constraint on the size of any possible line , a gaussian was added to this model . the best energy of this line is @xmath7 6.4 @xmath32 0.1 kev , in agreement with that expected from neutral fe k@xmath33 . this is consistent with that expected from reflection from neutral matter ( @xmath34 ) as assumed in the reflect model . the gaussian component was then replaced with relativistic line model ( laor : @xcite ) , which more accurately represents the expected line profile in the inner disc region . the inclination of this line is fixed at the same value as the the reflection component , the emissivity profile of the disc is fixed at @xmath35 and the outer radius of the emitting region is fixed at 400 r@xmath36 . we find the data require the presence a broad weak iron line ( @xmath37 significant as determined via an f - test ; ew = 73@xmath38 ) . allowing the inner disc radius to vary , we find @xmath39 . the iron line inner radius has a strong inclination dependence with a best fit @xmath40 at 30@xmath41 . the above inner radii are inconsistent with the isco for a schwarzschild black hole , if we extend the confidence intervals we place the following limits on the inner radius @xmath42 ( 3@xmath43 level ) for inclinations of 63.256@xmath41 & 30@xmath41 ( cos_i _ = 0.45 , 0.8660 ) respectively . including the lower energy flux in the fit once again , results in a chi - squared value of @xmath44 , we plot this in fig . [ soft_excess ] . a soft excess is present consistent with previous observations , i.e. @xcite . a simple blackbody accretion disc component ( diskbb : @xcite ) was added to this model to account for the excess soft x - ray flux , while the power - law index , @xmath45 , and the reflection fraction , f , were frozen at their previous best fit values . the disc component is strongly required by the data ; the best fit is achieved for a disc temperature of kt = 0.20 kev ( @xmath46 ) . we measure the associated 0.6 10 kev ( 2 150 kev ) unabsorbed flux to be @xmath47 ( @xmath48 ) . the column density is crucial here : although it has little impact on the spectrum above energies @xmath7 2 kev , it will have a significant impact on the shape of the spectrum and the measured flux below this value , e.g. fig . [ soft_excess ] . in order to test the effect of different values for @xmath49 the fitting was repeated at a number of different values for the interstellar column density ranging from @xmath50 @xcite to @xmath51 @xcite . we find the best fit as measured by _ suzaku _ was found to be @xmath52 . in table [ specfit_params ] , we list the parameters for the best fit model and also those for the model corresponding to the value of @xmath49 measured previously by _ xmm_. the normalization of the diskbb model is proportional to the inner disc radius , norm = ( @xmath53 [ km])/d [ 10kpc])@xmath12cos@xmath54 . in fig . [ inner_radius ] , we plot the inner disc radius corresponding to the best fit diskbb component for various values of the column density , where we have corrected for spectral hardening and the inner disc radius ( see [ thin_disc_isco ] ) . we find that for reasonable values of the column density and inclination the inner disc radius may reside close to the isco , in agreement with recent work , which has provided evidence that the cool disc may reside at or near the isco , even at the low luminosities typically observed in the lhs @xcite . we also modelled the soft excess using the ezdiskbb and diskpn models in xspec . in both cases , the observed blackbody component is found to be consistent with that obtained using diskbb . we also experimented with using the kdblur kernel to relativistically blur the above continuum model ; however , this did not result in an improved fit . we also carried out fits to the individual telemetry segments of the xis , pin & gso spectra , to check for possible variability . the fits to each of the 3 segments were found to be consistent with each other , hence there is no evidence for spectral variability in agreement with the constant flux observed from the lightcurves . [ cols="<,^,^,^,^,^ " , ] we present _ suzaku _ broadband spectra ( 0.6 250 kev ) of the black hole candidate swift j1753.5 - 0127 while in the lhs . the observed spectrum is measured to be consistent with an unbroken power - law , @xmath55 . during our observation the flux from the source was observed to be constant , with a 0.6 - 150 kev unabsorbed flux of @xmath56 ( @xmath57 ) . this is consistent with _ integral _ observations of gro j1655 - 40 in the lhs where unbroken power - law emission ( @xmath58 ) extending out to @xmath7 500 kev was detected at a luminosity of @xmath7 0.015 l@xmath59 ( @xcite ; although see @xcite ) . accurate determination of the interstellar column density is crucial due to the preferential absorption of soft x - rays , which in our case are consistent with being emitted by the cool accretion disc . measurement of @xmath49 is best achieved at soft x - ray energies . here _ xmm _ with its well studied low energy calibration should provide the most reliable determination of the interstellar column density @xmath60 @xcite . in contrast the best fit value as measured with _ suzaku _ is @xmath61 . however , we do note that the measured @xmath49 depends on the continuum model assumed , e.g. for the comptonizing corona model we find a best fit value @xmath62 . in order to investigate the instrumental dependence of our measured values of the column density , we also fit the best fit models above to a number of _ swift _ xrt spectra taken before and after our _ suzaku _ observation , with the closest occurring @xmath7 60 days beforehand ( obsid : 00030090050 ) . the results from these fits are consistent with the results of our fits to the _ suzaku _ spectrum , with the caveat that the smaller energy coverage ( 0.6 10 kev versus 0.6 150 kev ) and exposure time ( @xmath7 2 ks ) necessarily result in larger confidence regions from fits to the _ swift _ spectra . these measurements are in agreement with numerous independent measures of the interstellar column , sourced both from direct measurements and galactic surveys . x - ray observations from _ swift _ measured a column density of @xmath63 @xcite , while optical measurements also require @xmath64 @xcite . additionally an estimate of the column density may be obtained from various radio ( n@xmath65 = @xmath66 @xcite , @xmath67 @xcite ) , and far - ir ( @xmath68 @xcite ) surveys . the radius of the cool disc component depends on the column density and inclination as illustrated in fig . [ inner_radius ] . from above , we see that for the directly measured values of @xmath49 towards swift j1753.5 - 0127 , the observed soft excess may originate from an accretion disc whose inner radius is consistent with the isco . the significant soft excess observed in swift j1753.5 - 0127 ( fig . [ soft_excess ] ) is consistent with an origin in a standard multi - colour blackbody accretion disc ( @xcite and see table [ specfit_params ] ) . the disc normalization may be used to estimate the inner radius of the accretion disc when knowledge of the distance and inclination are available , as norm @xmath69 . this estimate is subject to a number of corrections , ( i ) spectral hardening must be accounted for , typically this requires a multiplicative correction factor @xmath7 1.7 @xcite and , ( ii ) one must correct for the inner radius , where @xmath70 ( @xcite ) . there are also additional errors , i.e. the zero torque inner boundary condition @xcite and radiative transfer effects @xcite , which could contribute but are difficult to quantify . in fig . [ inner_radius ] , we plot the inner disc radius ( corrected for ( i ) & ( ii ) above ) measured from our best fit models in 3.2 . the best fit radius is consistent with the isco , for certain values of @xmath49 & inclination , although there are large uncertainties . previous observations of swift j1753.5 - 0127 at a 0.5 10 kev flux of 3.9@xmath71 detected a cool accretion disc with an inner radius consistent with the isco . the measurements presented herein , at an unabsorbed 0.6 10 kev flux of 6.8@xmath71 , are in agreement with the previous analysis of @xcite and furthermore atest to the stability of the cool disc component given that the observations were separated by almost 18 months ( see fig . [ asmlc ] ) . @xcite observed gx 339 - 4 at a luminosity of @xmath72 with _ xmm _ & _ rxte_. a significant excess of soft x - ray flux was detected in the _ xmm _ data alone , due to the absence of soft x - ray coverage for _ rxte_. modelling revealed this excess to be consistent with a cool accretion disc ( @xmath73 ) extending to the isco . subsequent observations utilizing both _ rxte _ & _ swift _ at luminosities of 0.023 and 0.008 @xmath74 also detected this excess soft x - ray component @xcite . together with the new observations presented in this paper , these results appear to confirm the presence of the inner radius of the accretion disc at the isco ( or at least at a radius far lower than expected ) for luminosities far below that at which the state transition from hss to lhs state occurs , in contrast to theoretical expectations ( e.g. @xcite ) . @xcite have re - analysed the _ xmm _ data on swift j1753.5 - 0127 & gx 339 - 4 focusing on the timing characteristics and variability inherent to the spectra . using a new analysis technique , they find significant evidence for the presence of the soft spectral component in agreement with @xcite . they interpret the soft x - ray flux as originating in variability intrinsic to the accretion disc . in particular they find that the intrinsic variability of the accretion disc is likely to be responsible for the low frequency lorentzian feature present in the power spectral density function of sources in the low - hard state . this places a limit on the disc truncation radius of @xmath11 20 @xmath75 . the _ xmm _ spectrum of swift j1753.5 - 0127 has also been re - analysed by @xcite . they fit the spectrum with a relativistically blurred reflection model and find an inner disc radius of @xmath76 . this in turn provides an measure of the spin of the putative black hole of @xmath77 . again , inconsistent with the idea of a large truncation radius for the accretion disc in the low - hard state . the magnitude of the reflection features is observed to be quite low in our data . although the presence of a broad relativistic line is required by the data ( @xmath78 significant , ew @xmath79 ) , it is intrinsically weak ( norm @xmath80 ) . the detection of a relativistic iron line is consistent with the measurements of @xcite who also measured a redshifted line at an energy of 6.2 kev during pointed _ rxte _ observations ( see fig . [ asmlc ] ) . the inner radius inferred from the laor line fits is larger than the isco at @xmath81 10 20 gravitational radii although much less than might be expected from the standard disc picture for disc truncation , where the inner radius is expected to be an order of magnitudes larger @xcite . the curvature present in the pin spectrum is also consistent with a low reflection fraction , f @xmath7 0.12 - 0.21 ( inclination @xmath7 30@xmath41 63.256@xmath41 ) . we note that the reflection fraction inferred from the self consistent cdid model is at the lower end of the values inferred from the phenomenological reflect*(diskbb+po)+laor model . it is known that the reflection features of black hole binaries are generally weaker in the low - hard state than in higher luminosity states ( i.e. @xcite ) . the relative weakness of these features in swift j1753.5 - 0127 is primarily due to the low luminosity nature of the source ( @xmath82 ) . in comparison , the x - ray luminosity in gx 339 - 4 ( @xcite ) was much higher when a strong relativistically smeared iron line ( @xmath83 significant , ew @xmath7 350 ev ) was detected ( @xmath84 ) . the value we measure for the reflection fraction in swift j1753.5 - 0127 is comparable to previous lhs observations where reflection fractions of @xmath7 0.2 0.3 were measured in cyg x-1 & gx 339 - 4 @xcite . in contrast , observations at higher luminosities require much higher reflection fractions . x - ray spectra of the black hole binaries xte j1650 - 500 and gx 339 - 4 in the vhs state revealed reflection fractions of approximately unity @xcite . the low reflection fractions inferred in the low - hard state may be interpreted as evidence for a truncated disc , e.g. @xcite . alternatively the disc may remain close to the isco with the low reflection fraction resulting from beaming of the hard x - ray flux away from the disc by a mildly relativistic outflowing corona , e.g. @xcite . a number of alternative explanations have been put forward to account for the presence of the soft excess observed in swift j1753.5 - 0127 & gx 339 - 4 . @xcite reanalyzed the data from @xcite and find that while it is consistent with containing a soft disc component , a number of alternative continuum prescriptions that allow the disc to be truncated are also valid ; however , see @xcite who outline a number of issues with this result . @xcite propose a scenario in which the disc is truncated at larger radii ; however , the inner edge of the disc is irradiated by the high energy comptonized photons in the corona . this irradiation will cause the truncated inner disc edge to be hotter and hence appear to lie at smaller radii than is actually the case . in the model of @xcite , the inner edge of the cool truncated disc is seen to overlap with the inner adaf region . here in the overlap region the cool disc is heated to temperatures mimicking those of a cool disc at the isco . it is also possible that what we observe is not the accretion disc but instead a ring of material that has condensed from the corona . such an idea has been explored in detail by @xcite and @xcite . they find that it is possible for enough material to condense from the corona to produce an inner ring of material extending from near the isco to a few tens of schwarzschild radii , while the accretion disc is truncated at larger radii . in particular , detailed modelling was able to reproduce the soft disc components observed by @xcite in gx 339 - 4 ( see @xcite ) . two additional black hole binaries have been observed by _ suzaku _ while in the low - hard state . @xcite presented observations of gro j1655 - 40 , while @xcite observed cyg x-1 . the observations took place while the systems were in the lhs at luminosities of approximately 0.007 l@xmath59 and 0.02 l@xmath59 respectively . we also modelled the spectra of swift j1753.5 - 0127 with a combination of 2 comptonizing components , to aid comparison of this data to the _ suzaku _ observations of the black hole binaries cyg x-1 @xcite & gro j1655 - 40 @xcite . for both of these systems , the observed spectrum was interpreted as being due to a 2 component comptonizing corona , i.e. a population of electrons with the same temperature ( @xmath85 ) but differing optical depths ( @xmath86 ) , modified by disc reflection , diskbb+compps+compps . the parameters of our best fit model are listed in table [ specfit_params1 ] , while the fit itself is displayed in fig . [ spec_ii_images_2 ] . in table [ obs_comparison ] , we display the parameters for our best fit comptonization model along with those for the best fit models for cyg x-1 and gro j1655 - 40 . the blackbody disc component is found to be cool ( kt @xmath7 0.1 kev ) and the inner radius of the accretion disc is consistent with the outer radius of the corona within the errors ( @xmath87 ) . the best fit radius for the iron line is also consistent with this value . the high value for the hydrogen column density ( @xmath88 ) returned by this model is inconsistent with the available data and casts doubt the comptonization scenario as a viable model for the observed _ suzaku _ spectrum of swift j1753.5 - 0127 . however , the presence of circumstellar matter or a local absorber could account for the excess absorption above that detected at radio wavelengths . we find an electron temperature cooler than that measured in both cyg x-1 and gro j1655 - 40 when the spectra are interpreted in terms of this comptonizing corona model . assuming the disc truncation model for the low - hard state to be correct , we would expect the electron temperature to decrease with increasing luminosity , e.g. @xcite . the value of @xmath89 we measure would imply that swift j1753.5 - 0127 was at a luminosity in excess of that at which cyg x-1 was observed ( @xmath90 ) . this is in contrast to various luminosity estimates . indeed , we measure a luminosity of @xmath91 ; for swift j1753.5 - 0127 to have a luminosity greater than that of cyg x-1 at the time of the _ suzaku _ observations would imply a distance greater than 8.5 kpc and or a black hole mass less than 10@xmath92 . @xcite have argued that the inner radius of the accretion disc does not extend to the isco in the lhs state based on the _ suzaku _ observations of cyg x-1 and gro j1655 - 40 ( see 4.3 ) . they base their argument around 2 main points ; firstly the low reflection fractions measured in the lhs state show that either the accretion disc does not intrude too deeply into the corona or some form of outflow must be formed . they consider the outflow case to be unlikely as the reflection fractions measured in both gro j1655 - 40 ( i @xmath93 ) and cyg x-1 ( i @xmath94 ) are similar , whereas if the reflection was due to an outflow it would be expected to have a strong inclination dependence , which is not observed . here , we note that the errors on the reflection fractions measured in the case of cyg x-1 ( f = 0.4@xmath95 ) , call into question the validity of this inclination based argument , given the quality of the currently available data . secondly it is argued that the inner radius determined from fits to the iron k line in cyg x-1 ( @xmath96 ) is inconsistent with the isco . this point is much more ambiguous as their reported value of the inner radius is @xmath97 . this value is consistent with the isco for a schwarzschild black hole . as such , it is clear from the currently available data that there is evidence for a cool disc component , with an inner radius that in some cases is consistent with the isco , whether or not the cool disc component actually extends to the isco is presently not clear . finally , it is important to note the high mass x - ray binary nature of cyg x-1 . here the accretion process , namely accretion via a stellar wind , is significantly different from that in gro j1655 - 40 ( roche lobe overflow ) and as such may preclude detailed comparison with the low mass x - ray binaries like swift j1753.5 - 0127 and gro j1655 - 40 . lcccccccccc system & l@xmath98 & n@xmath65 & kt@xmath99 & r@xmath100 & r@xmath101 & r@xmath102 & kt@xmath103 & @xmath104 & @xmath105 & f + & [ l@xmath59 ] & [ @xmath106 ] & [ kev ] & [ km ] & [ km ] & [ km ] & [ kev ] & & & [ @xmath107 ] + swift j1753.5 - 0127 & 0.016 & 0.31 & 0.1 & @xmath7 830 & @xmath7 205 & @xmath7 360 & 53 & 0.34 & 2.57 & 0 + gro j1655 - 40 & 0.007 & 0.74 & 0.2 & @xmath7 330 & @xmath7 26 & @xmath7 65 & 135 & 0.25 & 1.2 & 0.5 + cyg x-1 & 0.02 & 0.66 & 0.3 & @xmath7 103 & @xmath7 75 & @xmath7 200 & 100 & 0.4 & 1.5 & 0.4 + , to facilitate comparison with the seds displayed in @xcite . the optical / nir & x - ray data are quasi - simultaneous , see [ the_jet ] for details . ] our knowledge of role of the relativistic radio jets in x - ray binaries has advanced rapidly in the past 10 years . observations of numerous x - ray binaries have revealed the presence of a compact quasi - steady jet in the low - hard state , whereas in the high - soft state the jet is typically observed to disappear @xcite . these also point towards a significant contribution from the jet towards the overall energy budget of the binary system , with the jet luminosity being possibly equal to that observed in x - rays , e.g. @xcite . in addition , a significant correlation has been found between the x - ray and radio luminosities of accreting sources across a range of luminosity ( in eddington units ) , suggestive of a disc - jet coupling @xcite . @xcite have shown how one could , in principle , distinguish between different jet models based on the magnitude of the reflection features present in the spectrum . in particular , reflection fractions greater than 0.2 could not be reproduced by their jet models , reflection fractions between 0.10.2 could be due to synchrotron self - comptonization in the base of a jet , while a reflection fraction below 10% could be interpreted as evidence that the spectrum was dominated by synchrotron emission from the jet . if the base of the jet is confined within a few gravitational radii of the black hole , relativistic light bending will become important . this will act in the opposite direction to the beaming of the outflow leading to a slightly larger reflection fraction @xcite . these models are also consistent with the detection of an accretion disc near the isco , as we have likely detected with _ suzaku_. a break or cut - off in the high energy spectrum of black holes and neutron stars can be interpreted in terms of the electron temperature in a thermal comptonizing corona . the high energy spectrum of swift j1753.5 - 0127 extends to at least 200 kev without a strong break or cut - off in our observations . moreover , single thermal comptonization models fail to achieve acceptable fits . in the context of @xcite , the reflection fraction of @xmath7 0.12 0.21 measured in swift j1753.5 - 0127 is consistent with a disc that is illuminated by synchrotron self - comptonization produced in the base of a jet . a corona that is independent of a jet but which produces a non - thermal comptonization spectrum might be able to describe the data , but the base of a jet provides a natural context for the production of hard x - ray emission like that we have observed . in section [ spec_i ] , we showed how the data may be fit with a power - law modified by reflection , this model is supported by more detailed modelling in [ spec_ii ] . in both of these cases an incident power - law is required , the jet is a natural source of such a power - law . in fig . [ sed_jet ] , we plot the absorption corrected spectral energy distribution for swift j1753.5 - 0127 . the optical data is from @xcite and was taken during at approximately the same time as our _ suzaku _ spectra . in addition , we include v - band optical and j , h & k - band nir data points obtained in april and july 2009 at the magellan and mdm observatories . the v - band magnitude was found to be consistent with the previous measurements of @xcite , as such the nir data should be representative of the infrared portion of the sed at the time of our _ suzaku _ observations . the radio data points represent the radio detection shortly after the 2005 outburst and likely represent upper limits to the radio flux at the time of our observations @xcite , these are also consistent with the non - detection at 5 ghz & 8 ghz by @xcite . the solid line extending from the x - ray to the optical represents a power - law as might be expected from optically thin radio emission , i.e. @xmath108 . this is similar to the slope of the optically thin jet in gx 339 - 4 @xcite . extrapolating this power - law to optical / nir wavelengths shows that a spectral break is required at optical / uv frequencies . analysis of optical data of swift j1753.5 - 0127 in its current state by @xcite favour a case where any jet emission is insignificant at optical wavelengths consistent with our extrapolated power - law , this is in agreement with an independent observation / analysis by @xcite . @xcite has shown , based on analysis of optical data from xte j1118 + 480 , gx 339 - 4 & swift j1753.5 - 0127 , that the variability at optical wavelengths is inconsistent with being produced by reprocessing of x - rays in the accretion disc and instead favour a model where the emission in the low - hard state is dominated by the jet , e.g. @xcite . we have presented _ suzaku _ broadband x - ray observations of the candidate black hole x - ray binary swift j1753.5 - 0127 . the broadband spectrum ( 2 250 kev ) is observed to be consistent with a simple power - law model ( @xmath109 ) . confirming previous observations , we detect the presence of an excess at soft x - ray energies . in addition , a weak relativistic iron line and curvature consistent with a reflection hump at 20 30 kev are detected . these observations point towards the persistence of the accretion disc at a much lower radius than previously appreciated in the low - hard state . estimates of the disc inner radius with both a simple diskbb+po model and a detailed reflection model reveal values consistent with the isco ( @xmath110 ) for certain values of both the column density and inclination . in contrast modelling the spectra with a comptonization model , while revealing a truncated inner disc ( @xmath111 ) , implies a value for the hydrogen column density in disagreement with all previous estimates . this research has made use of data obtained from the _ suzaku _ satellite , a collaborative mission between the space agencies of japan ( jaxa ) and the usa ( nasa ) . this research has made use of data obtained from the high energy astrophysics science archive research center ( heasarc ) , provided by nasa s goddard space flight center . this research made use of the simbad database , operated at cds , strasbourg , france and nasa s astrophysics data system . caballero garcia m.d . , miller j.m . , kuulkers e. , diaz trigo m.d . , homan j. , lewin w.h.g . , kretschmar p. , domingo a. , mas - hesse j.m . , wijnands r. , fabian a.c . , fender r.p . , van der klis m. , 2007 , apj , 669 , 534

collective optical ( superradiant ) effects appear in ensembles in which the distances between single emitters are much smaller than the radiation wavelength with which they interact ( dicke limit ) . @xcite the spatial dependence of the electromagnetic field within such ensembles is negligible and thus all the systems effectively interact with common photon reservoir . this leads to formation of rapidly decaying ( superradiant ) and optically inactive ( subradiant ) states @xcite and consequently to the appearance of a vacuum - induced coherence effect which results in occupation trapping . @xcite although effects resulting from collective coupling of atoms to radiative environment have been known for nearly sixty years , are very well described @xcite and have been extensively investigated experimentally , @xcite they still attract much scientific attention . this is caused by the increasing variety of physical systems in which these phenomena may be observed , such as quantum dots ( qds ) , @xcite bose einstein condensate , @xcite superconducting qubits , @xcite ion coulomb crystals @xcite or depolaritons.@xcite the investigation of superradiant effects have been initiated and driven to a large extend by the promise which the short - living states show for optimization of lasers . @xcite recently the scientific interest has been focused on the concept of `` superradiant laser '' which allows to increase spectral purity of emitted light . @xcite the technological realization of such a device presented in ref . allows to improve accuracy of atomic clocks @xcite and thus measurements of gravity @xcite and fundamental constants . @xcite the experimental investigations of the collective effects were mostly restricted to the analysis of the superradiant states which appear spontaneously in the cascade emission and manifest themselves as a maximum in the intensity or photon emission rate . @xcite although the observation of the subradiance phenomena was also reported in the atomic ensembles , @xcite as the opposite of the superradiant states , the preparation of subradiant states is much more difficult , and therefore the possibilities they give were much less investigated experimentally . recently , the preparation of an optically inactive state was reported in a system of superconducting qubits @xcite and in a diatomic molecule in an optical lattice . @xcite the advantage of the subradiant states stems from their decoupling from the photon environment because of which they do not undergo radiative decoherence and thus may form decoherence - free subspaces . @xcite this makes them useful for quantum information processing especially for noiseless encoding of quantum information . @xcite the stable states also allow to construct a scalable quantum processor , @xcite quantum memories , @xcite nonlinear sign - shift gate @xcite and storing time - bin qubits @xcite for quantum cryptography . interesting group of dicke states is formed of single - excitation combinations belonging to the class of `` w '' states which have been widely considered for quantum information processing @xcite or optimization of the quantum clock synchronization . @xcite systems composed of two and more coupled qds attract much scientific focus due to the richness of their properties which pave the way to new technological applications . already pairs of quantum dots allow for long - time storage of quantum information,@xcite conditional optical control of carrier states , @xcite implementation of a two - qubit quantum gate , @xcite optical writing of information on the spin state of the dopant mn atom @xcite and construct quantum nanoantennas due to the collective phenomena . @xcite systems of three qds enable to realize two different kinds of entanglement , @xcite teleportation via superradiance @xcite , cnot gates@xcite and the control of spin blockade . @xcite moreover , in these systems the collective transport effects ( electronic dicke or kondo - dicke effect ) may be realized @xcite and lead to the enhancement of thermoelectric efficiency . @xcite arrays of qds allow to reduce the effect of pure dephasing on quantum information encoded in excitonic states . @xcite in this paper we analyze the collective optical effects in ensembles of three and four qds . compared to double qds in which only one optically inactive state may be realized , @xcite ensembles of three and more two - level systems allow to realize many stable states which occur at different exciton occupations of the single emitters . although the superradiant effects are well described in systems of identical atoms the description of such phenomena in qd ensembles requires taking into account characteristic for those system properties which distinguish them from natural ones . therefore we include in our model the fundamental energy mismatches , different dipole moments of single dots and coupling which induces excitation transfer between single emitters , but conserves the total population of the ensemble . in our previous works concerning double qds it has been shown that the collective optical effects are extremely sensitive to inhomogeneity of the fundamental transition energy which leads to the decay of the exciton occupation for the energy splitting much below the present technological feasibility . this decoherence effect may be strongly reduced by sufficiently strong coupling between the dots @xcite and fully overcome in double qds with different decay rates . @xcite in this paper we extend the results of two qds and specify conditions which allow to take advantage of the superradiant phenomena in fully inhomogeneous qd systems . we analyze the dynamics of one electron - hole pairs and biexcitons . we show how to adjust the system parameters in such a way that an arbitrary dark single - exciton combination may be blocked in a multiple qd , we also specify the conditions which allow to trap two excitons and prepare a system in a biexcitonic state which allows to recombine only one electron - hole pair . due to existence of two or more subradiant single - exciton eigenstates and coupling between the dots the occupation of individual dots oscillates while the population of the whole ensemble remains stable , we show that the oscillation amplitudes may be strongly reduced if the system is initially prepared in a biexciton state . the paper is organized as follows . in sec . [ sec : system ] we describe the system under study , define its model and describe a method used to study the system evolution , in sec . [ sec : results ] we present and discuss our results , we conclude the paper in sec . [ sec : conclusions ] . the investigated system consists of @xmath0 @xmath1 quantum dots ( qds ) in which only the ground - level exciton states with fixed spin polarizations are taken into consideration . due to the strong coulomb coupling and absence of the external electric fields we may restrict the discussion to ` spatially direct ' excitonic states , i. e. , states with electron - hole pairs residing in the same qd which in these conditions have much lower energy than the ` dissociated ' states referring to excitons formed of carriers residing in two different dots . @xcite these assumptions allow us to treat every qd as a two - level system which may be either empty or contain an exciton and thus describe the set of @xmath0 qds as an @xmath2-level system , with @xmath3 denoting the ground ( or `` vacuum '' ) state in which all @xmath0 dots are empty , single - exciton states @xmath4 , corresponding to one exciton localized in the @xmath5th qd , biexcitonic states @xmath6 referring to electron - hole pairs residing in the @xmath5th and @xmath7th qds , states @xmath8 with three qds , @xmath5th , @xmath7th , and @xmath9th ( @xmath10 ) , occupied with an exciton _ etc_. present manufacturing technology does not allow to produce on demand systems of qds with identical fundamental transition energies , therefore we assume different electron - hole binding energies of each dot defined as @xmath11 where @xmath12 is the average transition energy and @xmath13 is the energy mismatch of the @xmath5th qd . we impose @xmath14 and @xmath15 , such that @xmath16 is the mean square variation of the transition energies . we analyze the system in a rotating frame defined by the evolution operator @xmath17,\end{aligned}\ ] ] where @xmath18 are the annihilation ( creation ) operators for the exciton in the @xmath5th qd , respectively and @xmath19 is the standard free photon hamiltonian with operators @xmath20 and @xmath21 creating and annihilating radiation modes with wave vector @xmath22 and polarization @xmath23 , while @xmath24 is the corresponding frequency . in this frame the hamiltonian of the system is @xmath25 the first term describes electron - hole pairs residing in the qd system . we assume that the ground state @xmath3 corresponds to the zero energy level , so the excitonic hamiltonian is @xmath26 where @xmath27 are biexcitonic shifts due to the interaction of static dipole moments of @xmath5th and @xmath7th qd , @xmath28 is a deviation of energy caused by interaction of the dipole moments of three dots , _ etc_. in ensembles of qds one may distinguish two types of interaction between the emitters : dipole ( frster ) coupling which in the leading order decays as @xmath29 with the qd separation @xcite and short - range coupling resulting form a combination of tunneling ( wave function overlap ) and coulomb correlations . both types of interaction induce excitation transfer between the emitters but conserve the total exciton occupation of the ensemble . the short - range couplings allow to rebuilt enhanced emission in the energetically inhomogeneous ensembles while in the case of dipole interaction a similar effect is achieved if the coupling is enhanced artificially by a factor of @xmath30 . @xcite to overcome the destructive effect of the transition energy mismatch the coupling between the dots must be of the order of the energy splitting , i. e. , @xmath31 mev for technologically feasible systems . @xcite for a planar qd arrangement the distance between emitters is about @xmath32 nm ( the average value for the sample studied in ref . ) , and for such a distance the frster coupling drops to about @xmath31 @xmath33ev , which is not sufficient to stabilize collective effects . the frster coupling reaches a fraction of mev for qds separated by only a few nanometers,@xcite which can be achieved in vertically stacked qd systems , and where indeed it can stabilize the collective effects.@xcite in the present paper we describe a planar system and therefore we prefer to consider a short - range coupling , for which we choose an exponential model , @xmath34 $ ] , where the subscripts @xmath5 and @xmath7 refer to @xmath5th and @xmath7th qd , respectively , @xmath35 is a constant amplitude , @xmath36 is the distance between the qds and @xmath37 the spatial range of the interaction . remarkably , this simple model reproduced well recent experimental results.@xcite the eigenstates of the hamiltonian ( [ hamiltonians ] ) do not mix quantities associated with different exciton numbers , i. e. the eigenstates of the system are superpositions of the basis states restricted to a particular exciton number . the second term of the hamiltonian accounts for coupling between the qd system and quantum electromagnetic field @xmath38 where @xmath39 is a coupling constant for the @xmath7th qd . here @xmath40 is the inter - band dipole moment for the @xmath7th qd , @xmath41 is the unit polarization vector of the photon mode with polarization @xmath23 , @xmath42 is the vacuum dielectric constant , @xmath43 is the relative dielectric constant and @xmath44 is the normalization volume . we restrict our investigations to wide - gap semiconductors with electron - hole binding energies of the order of 1 ev which allows us to describe the photon modes within the zero - temperature approximation at any reasonable temperature . to describe the evolution of the carrier subsystem we use an equation of motion for the reduced density operator in the markov approximation . in the rotating frame it takes a form @xmath45+\mathcal{l}_{\mathrm{rad}}[\rho],\end{aligned}\ ] ] where @xmath46 is a reduced density matrix of the exciton subsystem and @xmath47 is a lindblad dissipator @xmath48 = \sum_{i , j=1}^{n}\gamma_{ij}\left[\sigma^{(i)}_{-}\rho\sigma^{(j)}_{+ } -\frac{1}{2}\left\{\sigma^{(j)}_{+}\sigma^{(i)}_{-},\rho\right\}\right ] , \label{superlindblad}\end{aligned}\ ] ] where @xmath49 with @xmath50 being the speed of light . since for @xmath51 the eq . ( [ gamma ] ) describes the spontaneous decay rates of single qds , @xcite the mixed ( off - diagonal ) decay rates ( [ gamma ] ) may be expressed in terms of single qd quantities , @xmath52 where @xmath53 and @xmath54 , here @xmath55 is an irrelevant phase and @xmath56 is a small angle between the dipole moments which depends of light - hole admixture . @xcite in the numerical simulations we assume constant energy mismatches with the parameters @xmath57 , @xmath58 , @xmath59 , @xmath60 and @xmath61 mev ( except for the fig . [ fig:3qd_angle ] ) . for the coupling amplitudes we take @xmath62 mev and @xmath63 nm . in sec . [ idealsingle ] , [ coupledsingle ] and [ sec : biexciton ] we assume parallel dipole moments , since the effect of the light - hole admixture in the absence of external electric fields is negligible . @xcite below we present an analysis of the collective effects which occur in multiple qds . we define the dicke states for an arbitrary number of emitters and perform numerical simulations for ensembles of three and four qds . in fig . [ fig:34qd ] we illustrate the numbering of the qds and their spatial arrangement . in sec . [ idealsingle ] and [ coupledsingle ] we focus on single - exciton states . in sec . [ idealsingle ] we show the evolution of uncoupled systems with identical fundamental transitions and parallel dipole moments of different amplitudes . then , in sec . [ coupledsingle ] we analyze the same effects in a system composed of three coupled energetically inhomogeneous dots . the dynamics of biexciton states is presented in sec . [ sec : biexciton ] . in sec . [ pinjunction ] we show possibilities of controlling the exciton occupation given by a p - i - n junction . the collective effects were described for the first time in ensembles of uncoupled identical atoms where all the emitters have the same transition energies and dipole moments . @xcite such optical effects , resulting from the interaction in the dicke limit , are also present in uncoupled systems , even with different dipole moments , if all the emitters have identical transition energies . for the purpose of this paper we define such systems as ideal qds . the coupling of excitons to their radiative environment described in the dicke limit by the hamiltonian ( [ hamiltoniansrad ] ) and the fermi s golden rule , according to which the probability of releasing a photon through a transition from the initial to the final states is @xmath64 , allow to define rapidly decaying ( superradiant ) and optically inactive ( subradiant ) states , both also known as dicke states . by definition the superradiant initial state ( @xmath65 ) corresponds to the maximum transition probability , whereas the subradiant states ( @xmath66 ) refer to the vanishing probability . due to the decoupling from the photon reservoir these are dark , optically inactive states . in the weak excitation limit , i. e. , for a single excitation in the system from which the sample may decay only to the ground state ( @xmath3 ) , the proportionality @xmath67 allows to write the short - living states in the form @xmath68 and express the stable superpositions as @xmath69 where the coefficients @xmath70 satisfy @xmath71 . irrespective of the number of emitters there is only one superradiant state in the system of a particular number of qds , whereas the only structure which realizes just one dark state is a double quantum dot . the systems of three and more qds allow to realize an arbitrary number of dark states of the form ( @xmath72 ) since there are may combinations of the parameters @xmath70 for which the transition matrix element @xmath73 . and coherences [ ( b ) and ( d ) ] of a system of three qds prepared initially in a bright state @xmath74 for @xmath75 ns@xmath76 , @xmath77 ns@xmath76 and @xmath78 ns@xmath76 . figs . ( a ) and ( b ) show the results for uncoupled qds ( @xmath79 ) with identical transition energies ( @xmath80 ) , while figs . ( c ) and ( d ) refer to energetically inhomogeneous ( @xmath81 ) and coupled ( @xmath82 ) system arranged in a lateral array of an equidistant triangle shape with the side length @xmath83 nm . the inset to fig . ( a ) shows the initial evolution of the exciton occupation of the qd number 2 . ] the consequence of the co - existence of rapidly decaying and stable states is the effect of spontaneous trapping of excitation . @xcite an arbitrary single - exciton state @xmath84 , where @xmath85 is the localization probability of the exciton on the @xmath5th qd and @xmath86 , may be expressed as a combination of the superradiant state ( [ bright ] ) and a dark state of type ( [ dark ] ) , @xmath87 where the dark state is @xmath88 } { \sqrt{\left(\sum\limits^{n}_{i=1}\gamma_{ii}\right ) \left[\sum\limits^{n}_{i=1}\gamma_{ii}-\left(\sum\limits^{n}_{i=1}c_{i}\sqrt{\gamma_{ii}}\right)^2\right]}}.\end{aligned}\ ] ] the derivation of the eqs . ( [ single ] ) and ( [ dark2 ] ) is done in the the appendix . the collective coupling to the radiative surrounding induces emission only from the superradiant state and thus the fraction of excitation initially spanned in the dark state , @xmath89 , remains unaffected . since the only single - exciton state which decays totally is the superradiant state , we define a state @xmath90 as being bright if its superradiant contribution does not vanish , i. e. , if a system prepared in that state partially recombines and only a part of the initial exciton occupation remains trapped . in fig . [ fig:3qd_comparison ] we show the dynamics of a single - excitation induced by a common photon reservoir in an ideal system of three uncoupled ( @xmath91 ) qds with equal electron - hole binding energies ( @xmath92 ) and parallel dipole moments ( @xmath93 ) of different magnitudes . in fig . [ fig:3qd_comparison](a ) we show the evolution of the exciton occupations of a system prepared initially in a bright state @xmath74 and in fig . [ fig:3qd_comparison](b ) we show the corresponding coherences . as expected , the coupling to the photon reservoir spans the excitation into the sub- and superradiant states according to eq . ( [ single ] ) and induces emission only from the short - living state @xmath94 [ fig . [ fig:3qd_comparison](a ) ] . the excitation dynamics takes place until occupations of all dots stabilize at certain levels corresponding to the dark state defined in the eq . ( [ dark2 ] ) which confirms that the state ( [ dark2 ] ) is indeed unaffected by the photon reservoir and , after the decay of the superradiant state , neither the total exciton occupation nor occupations of single dots ( @xmath95 ) change due to radiative environment . the emission from the above state induces decay of the total exciton occupation and excitation transfer which results in the redistribution of the occupations of single dots . since all of the localized single - exciton states @xmath4 contribute to the superradiant state ( [ bright ] ) , the collective coupling spans the initial excitation in all of the dots even if some of them were initially empty . therefore the population of initially unoccupied systems builds up spontaneously [ @xmath96 , green - dashed line in fig . [ fig:3qd_comparison](a ) ] . if the initial occupation of one of the dots is relatively small while the spontaneous decay rate from that system is sufficiently strong then the exciton occupation of that dot may vanish at same point and then restore due to the excitation transfer [ magenta - dotted line in fig . [ fig:3qd_comparison](a ) and the inset to fig . [ fig:3qd_comparison](a ) ] . during the emission process also the evolution of the off - diagonal density matrix elements is observed , the coherences related to the initially populated dots decay , while those corresponding to initially empty systems build up spontaneously due to the increasing occupations of those dots . when exciton dynamics in the system reaches population distribution corresponding to the optically inactive state also the off - diagonal density matrix elements stabilize at a certain non - zero level [ fig . [ fig:3qd_comparison](b ) ] , defined by the dark contribution to the initial state . technologically feasible qds forming multiple structures differ both in fundamental transition energies ( @xmath81 ) and dipole moments ( @xmath97 ) , and are coupled with each other ( @xmath82 ) . as shown in the previous section and , for a double qd in ref . , the superradiant character of the evolution of one exciton is present in the ideal systems ( @xmath80 ) , but with parallel dipole moments ( @xmath93 ) . the collective evolution is very sensitive to the energy mismatches and is destroyed in ensembles with energy splittings of the order of the transition line width . @xcite in such systems the localized eigenstates corresponding to different energies can not form delocalized superpositions which would also be the system eigenstates . this destructive effect may be overcome by coupling between the dots ( @xmath98 ) which delocalizes the system eigenstates and different dipole moments allowing the superradiant state to be a non - symmetric superposition of the localized states @xmath4 [ eq.([bright ] ) ] . the single - exciton eigenstates of the system depend on the energy mismatches and coupling between the dots , while the dike states are defined by the interplay of decay rates [ eqs . ( [ bright ] ) and ( [ dark ] ) ] . if the single - exciton decay rates [ eq . ( [ gamma ] ) for @xmath99 are adjusted in such a way that the superradiant state ( [ bright ] ) corresponds to one of the system eigenstates , then the inhomogeneous ensemble of qds interacts with its radiative environment in the `` collective regime '' i. e. allows many effects typically present only in systems with identical electron - hole binding energies . the amplitudes @xmath100 of a single - exciton state orthogonal to the superradiant state ( [ bright ] ) must satisfy the equation @xmath101 which implies the condition @xmath102 defining a subradiant state . therefore , if one of the eigenstates has a superradiant character then the other eigenstates of a system are optically inactive and thus defying the `` collective regime '' requires only specifying the superradiant eigenstate . in figs . [ fig:3qd_comparison](c ) and [ fig:3qd_comparison](d ) we show the evolution of a realistic group of three qds placed in the corners of an equilateral triangle , we assume non - equal fundamental transition energies ( @xmath61 mev ) , non - vanishing coupling between the systems ( @xmath82 ) and parallel dipole moments ( @xmath93 ) . we compare the results obtained for an inhomogeneous system coupled to the photon reservoir in the `` collective regime '' to the ideal case presented in figs . [ fig:3qd_comparison](a ) and [ fig:3qd_comparison](b ) , where the decay rates of individual dots take the same values as in figs . [ fig:3qd_comparison](c ) and [ fig:3qd_comparison](d ) . as can be seen in fig . [ fig:3qd_comparison](c ) , coupled ensembles with energy mismatches of the order of mev allow to trap the same fraction of excitation as ideal dots with the same decay rates [ red - solid lines in figs . [ fig:3qd_comparison](a ) and [ fig:3qd_comparison](c ) ] . here , as in the ideal case , any single - exciton eigenstate may be decomposed into sub- and superradiant component according to eq . ( [ single ] ) and also in this case the superradiant state is the only state which decays totally . for an arbitrary set of decay rates , which do not correspond to the superradiant eigenstate , the exciton occupation of a system prepared initially in a state of the form ( [ dark ] ) is quenched and the decay of the state ( [ bright ] ) is slowed down compared to the `` collective regime '' . although sub- and superradiant states may exist in the appropriately designed realistic systems , the dynamics of individual qd occupations differs considerably from the discussed in the previous section ideal case , when the dots interact only through the common radiative reservoir . the situation changes when the dots communicate with each other via short - range coupling which induces excitation transfer between dots and thus oscillations in the evolution of single qd populations . in a double qd system the oscillation amplitudes decrease and the occupations stabilize at levels corresponding to the only one dark state of the system . @xcite in the multiple qds composed of three or more emitters oscillations of the individual dot populations never vanish [ figs . [ fig:3qd_comparison](c ) , [ fig:3qd_comparison](d ) and fig . [ fig:3qd_oscill ] ] . the excitation is trapped in the system because of the existence of dark states , which in such systems may be realized by many different amplitude combinations and thus also the final population number may be realized in various ways . ) shown by the blue line in fig . [ fig:3qd_comparison](c ) . the time evolution is shown for three different time intervals at picosecond resolution scale . ] in fig . [ fig:3qd_oscill ] we zoom the exciton occupation @xmath103 shown in fig . [ fig:3qd_comparison](c ) ( blue line ) . as can be seen in fig . [ fig:3qd_oscill](a ) the initial oscillation pattern is relatively complicated which is caused by the existence of three energy gaps defined by the differences between the superradiant eigenstate and the subradiant ones , @xmath104 , and by the energy splitting between the two subradiant eigenstates , @xmath105 . the period of the envelope oscillations is a multiplication of the corresponding three periods , i. e. @xmath106 and @xmath107 . the period @xmath108 itself defines the fine oscillations of the occupation . due to the emission process , the superradiant contribution attenuates and interference pattern simplifies [ fig . [ fig:3qd_oscill](b ) ] . finally , when the short - living state decays and the total exciton occupation becomes trapped , the evolution of the single dot occupation shows single - mode pattern which repeats with the time @xmath108 and with amplitude depending of the initial occupation of individual dots [ fig . [ fig:3qd_oscill](c ) ] . defined in fig . [ fig:34qd ] ( a ) , on the energy mismatch ( b ) and the corresponding steady state occupations for the initial state @xmath109 ( c ) and ( d ) , respectively . for all figs . we assume @xmath110 nm and @xmath111 nm and a constant decay rate of the qd number 2 ( @xmath112 ns@xmath76 ) . ] the system of three and more qds allows many different planar arrangements of the emitters . since the coupling amplitudes ( @xmath98 ) depend on the distances between emitters , the eigenstates of the system , and thus the decay rates for which the ensemble interacts collectively with its radiative environment , also depend on the arrangement of the dots . using the geometry design defined in fig . [ fig:34qd](a ) we calculate the dependence of the decay rates @xmath113 and @xmath114 on the spatial arrangements of the system . we assume constant distances @xmath115 and @xmath116 and thus constant values of the coupling amplitudes @xmath117 and @xmath118 and change the angle @xmath119 from @xmath120 degrees to linear design , i. e. , we increase the distance between dots @xmath31 and @xmath121 . as seen in fig . [ fig:3qd_angle](a ) the values of the decay rates necessary to form the `` collective regime '' slightly decrease with increasing angle @xmath119 . the two similarly decreasing decay rates ( while the third one is constant ) , according to eq . ( [ single ] ) lead to increasing steady state ( final ) exciton occupation which is shown in fig . [ fig:3qd_angle](c ) . if the coupling between two out of three qds is much stronger than coupling of that dots with the third one ( e.g. @xmath122 , @xmath123 ) then one of the system eigenstates has large contribution from the localized state associated with the weakly coupled dot ( number @xmath124 ) and to achieve `` collective regime '' the decay rate @xmath125 must be much smaller from @xmath113 and @xmath114 . in the limiting case of vanishing couplings @xmath126 and @xmath118 the localized state @xmath127 becomes the system eigenstate and the corresponding decay rate ( @xmath125 ) vanishes . consequently the pair of coupled dots acts as a double qd while the third dot does not contribute to the evolution . in fig . [ fig:3qd_angle](b ) we show the dependence of the decay rates forming the `` collective regime '' on the energy mismatch . both calculated decay rates decrease with increasing energy separation , but one changes slowly while the second decays fast , this leads to decreasing ( eq . [ single ] ) steady state occupation shown in fig . [ fig:3qd_angle](d ) . for appropriately arranged ensemble of dots the decay rates may decrease in such a way that for smaller energy mismatches the final exciton occupation decreases with increasing energy mismatch , but after exceeding a critical point it increases . the similar effects in the dynamics of single excitons are observed for electron - hole pairs confined in ensembles of four and more dots . ( a ) superradiant initial state . [ ( b ) and ( c ) ] initial biexciton state which allows for a recombination of only one exciton . ( @xmath128 , @xmath129 , @xmath130 , @xmath131 , @xmath132 , and @xmath133 ) . the values of the decay rates of individual dot are the same for all figs . and take the values : @xmath134 ns@xmath76 , @xmath135 ns@xmath76 , @xmath136 ns@xmath76 , and @xmath137 ns@xmath76 . the biexction shifts are @xmath138 mev , @xmath139 mev , @xmath140 mev , @xmath141 mev , and @xmath142 mev with @xmath143 being an arbitrary parameter . the line types used in panel ( b ) are also valid for panel ( a ) . ] in multiple qds built out of three and four units more than one exciton may be delocalized and thus these systems allow for more complex collective effect . below we focus on biexciton states which in general are described by a vector @xmath144 , where @xmath145 . the realistic ensembles of qds ( @xmath81 , @xmath146 , @xmath82 and @xmath147 ) permit superradiance phenomena in the two - exciton subspace if the biexcitonic , as well as single - exciton , eigenstates correspond to the dicke states . as in the single - exciton case , the biexcitonic state is considered to be superradiant if the exciton occupation of a system prepared in this state decays totally . although eigenstates of the hamiltonian ( [ hamiltonians ] ) do not mix localized basis states associated with different exciton numbers , the biexcitonic superradiant state may be formed provided one of the single - exciton eigenstates has a superradiant character . thus the biexcitonic superradiant states occur only if collective effects are present in the single - exciton subspace . due to the equal number of single - exciton eigenstates and qds forming the ensemble the `` collective regime '' in the single - exciton domain may be achieved by adjusting only the decay rates ( dipole moments ) . as will be explained in detail below , the same rates @xmath148 ( [ gamma ] ) define the biexcitonic dicke states . thus , in order to achieve the collective effects in the two - exciton subspace one has to also appropriately adjust the spatial arrangement of the dots or energies . the superradiant two exciton superpositions may be spanned as well in ensembles of four emitters as in triple qds and take a form @xmath149 similarly to the single - exciton case , the biexcitonic superradiant states are defined by the maximum value of the transition probability ( @xmath150 ) , but this time from the initial biexciton state to the final single - exciton superradiant one ( [ bright ] ) . the form of the condition is governed by the coupling to the radiative environment [ hamiltonian ( [ hamiltoniansrad ] ) ] which induces decay of only one exciton at a time . thus total quenching of two excitons must occur through formation of single - exciton superradiant states . as can be seen in fig . [ fig : biex_fig](a ) , due to the decay of the biexciton superradiant state a part of initial excitation is initially transferred to the single - exciton state which reaches a maximum population and then is totally quenched , together with the biexciton state . the condition for the existence of the superradiant state ( [ brightbi ] ) implies that the transition to subradiant states ( [ dark ] ) vanishes . two excitons can be blocked in a system if a transition from the biexciton state to any single - exciton state is forbidden , i. e. the transition matrix element @xmath151 vanishes . due to infinite number of possible single - exciton states this condition reduces to the requirement @xmath152 . to simplify the description we define non - normalized amplitudes @xmath153 in such a way that the amplitudes of biexciton states @xmath154 . for a triple qd the condition leads to a system of three equations of the form : @xmath155 where every subscript @xmath5 , @xmath7 , and @xmath9 takes the values @xmath31 , @xmath124 , and @xmath121 , respectively . the system is satisfied only in the case of vanishing amplitudes @xmath156 which means that it is impossible to block two excitons in a triple qd . the coefficients of a stable biexcitonic state spanned in a system of four qds must satisfy the system of four equations of the form : @xmath157 where every subscript @xmath5 , @xmath7 , @xmath9 , and @xmath158 takes the values @xmath31 , @xmath124 , @xmath121 , and @xmath159 , respectively . the above equations lead to the condition for the @xmath153 numbers in a form @xmath160 although in a system of four qds many dark states may be spanned , in realistic systems the parameters of biexciton states may be adjusted in such a way that only one , particular ( for specified values @xmath161 and @xmath162 ) state is blocked . thus , as in the ideal systems and similarly to the double qds , @xcite the contribution of any pair of dots to the total biexcitonic population is constant in time ( @xmath163 ) . in realistic qds one may realize either superradiant or subradiant biexciton eigenstate but never both simultaneously as in the ideal system . in both cases the basis is supplemented by a third kind of state which allow for a recombination of one electron - hole pair and trapping of the second exciton . the evolution of a four qd system prepared in this state is shown in figs . [ fig : biex_fig](b ) and [ fig : biex_fig](c ) . while the population of the biexciton state decreases , the occupation of single - exciton state increases until it stabilize at the level corresponding to total trapping of one electron - hole [ green - dashed line [ fig : biex_fig](b ) ] . the third basis state must be orthogonal to the superradiant state ( [ brightbi ] ) , which equals to the requirement of vanishing transition matrix element between the biexciton state and the single - exciton superradiant state ( @xmath164 ) and means that the state @xmath165 has a subradiant character . the orthogonality to the subradiant biexcitonic state ( [ subbiex ] ) excludes contributions from the two - exciton dark states and thus the population trapping occurs only due to formation of a single - exciton subradiant state ( [ dark ] ) . since the transition to the state ( [ bright ] ) and thus the decay of single - exciton states is forbidden , one exciton is blocked in the ensemble . because the biexcitonic subradiant states can not be formed in triple qds , the requirement defining biexcitonic states which allow to block one exciton reduces to the orthogonality to the superradiant state , and the states take a form @xmath166 where @xmath167 . whereas , due to the orthogonality to the superradiant state the six amplitudes of the biexcitonic states spanned in ensembles of four qds take a form : @xmath168 where the coefficients must satisfy equation @xmath169 and , using further the orthogonality to the subradiant state relations , @xmath170 due to the existence of single and biexciton optically inactive states in ensembles of four qds the initial biexcitonic states allow for many combinations of final occupation . an arbitrary fraction of exciton occupation ( @xmath171 ) may be trapped by an appropriate combination of blocked single - excitons and biexcitons due to contribution to the initial state from dark states and basis states which allow to recombine only one electron - hole pair . irrespective of the initial number of excitons , ensembles of three qds allow to span only single - exciton subradiant states and thus block only single - excitons . it is important to emphasis that if the system was prepared in a bright biexcitonic state which leads to trapping of single - exciton occupation [ figs . [ fig : biex_fig](b ) and [ fig : biex_fig](c ) ] , then the pronounced oscillations due to coupling between the dots appear only in occupation of localized biexciton states , while the amplitudes of oscillations in the populations of single dots are negligible [ [ fig : biex_fig](c ) ] compared to single - exciton initial state [ fig . [ fig:3qd_comparison](d ) ] . this means that the biexciton initial state allows to achieve well defined stable single - exciton subradiant states , which is important for the application of quantum computation . the presence of optical collective effects in realistic qd systems requires high accuracy of a system parameters which may be controlled on the manufacturing stage or by external fields by , e.g. implementing the dots into the intrinsic region of a p - i - n junction . this structure provides a possibility of a separated injection of electrons and holes into qds from both sides of a sample and control of exciton dynamics and qd parameters through application of contacts on n- and p- type regions . it has been shown that controlled with a bias voltage carrier tunneling into a single qd in a p - i - n structure incorporated into a microcavity leads to regulated emission of single photons and pairs of photons.@xcite the ideas were followed by a technological realization of an electrically driven single - photon emitter with a layer of self - organized inas qds.@xcite the gate voltages constructed over dots allow to control energies of the dots and dipole moments , but the magnitudes of decay rates of single qds ( [ gamma ] ) depend on the average energy of the ensemble ( @xmath12 ) and thus operations on one dot change the decay rates of all qds in the system . nm and prepared initially in a single - exciton subradiant state . in regions i and iii the decay rates are the same as in fig . [ fig : biex_fig ] i. e. selected such that the systems interact collectively with the radiative surrounding . the initial state is @xmath172 . the red solid line corresponds to changes of the decay rates of individual qds , whereas the blue dotted line to the same changes plus changes of the relative angular orientation of the dipoles . in regions ii the decay rates and angles have been modified such that the ensemble does not interact collectively with its photon environment ( see text ) . ] in order to simulate this possibility we have calculated the time evolution of the system of four qds and we have changed the decay rates at some selected time points . the results are shown in in fig . [ fig : pin ] where the system is initially in the `` collective regime '' and has been prepared in a single - exciton subradiant state . we assume that the control electric fields are weak enough to exclude `` dissociated '' exciton states . as expected , initially the excitation is blocked in the system . at time 2.5 ns a change of parameters occurs which destabilizes the system and induces quenching of the excitation . the effect may be produced by a variation of the internal electric field in the p - i - n junction , which is simulated here by a change of the decay rates of each single dot , and shown by the red solid lines in fig . [ fig : pin ] . the decay rates have been changed as follows : @xmath173 in this case the quenching is relatively weak because of a small change in the ratio of the decay rates but the effect is visible . at time 7.5 ns the parameters of each dot are changed back and the system population is again stable . the decay may be enhanced by changing the orientation of the dipoles as shown by the blue dotted lines in fig . [ fig : pin ] . all dipoles are parallel in the regions i and iii , i. e. all angles @xmath93 . but now , in addition to the previous variations of @xmath148 the angles are also modified in the regions ii : @xmath174 radians . the misalignment of the dipole moments creates thus a stronger decay . the initial conditions may be restored at any time which may result in the trapping of a desired fraction of the initial occupation . we have studied the optical collective effects due to interaction of multiple quantum dots built of three and four emitters with the radiative surrounding . ensembles of three and more emitters allow to span many subradiant states which facilitate preparation of the system in an optically inactive single - exciton states and for ensemble of four emitters also in the biexciton subspace . we specified the conditions which allow the superradiance phenomena to occur in coupled inhomogeneous systems with different fundamental transition energies and dipole moments ( and thus decay rates ) . we discussed the dynamics of single electron - hole pairs and biexcitons . although many features typical for identical atoms , such as spontaneous trapping of excitation , may also occur in inhomogeneous qds there are differences in the dynamics of these systems . in principle , coupling between the dots induces excitation transfer between the dots which together with a possibility to define many dark states in ensembles of three and more dots lead to oscillations in the occupation of single dots . the amplitudes of these oscillates may be considerably reduced if the system is prepared initially in an appropriate biexction state which allows for trapping of one electron - hole pair . we envision that the presented collective effects may be controlled if the ensemble of dots is placed in the intrinsic region of a p - i - n junction with contacts constructed over the dots which due to sufficiently weak electric fields allow to control the dynamics of excitons and thus collective effects . this work was funded by the icelandic research fund ( rannis ) and by a polish ncn grant no . dec-2011/01/b / st3/02415 . a. s. acknowledges support within a scholarship for outstanding young scientists granted by the polish mnisw . the authors are grateful to pawe machnikowski for fruitful discussions . to express an arbitrary single - exciton state in terms of the superradiant state ( [ bright ] ) and a subradiant state ( [ dark ] ) we begin with a derivation of a formula for a localized state @xmath4 . the subradiant state which allows to cancel out all different from the state @xmath4 localized contributions to the superradiant state has a form @xmath175 and is orthogonal to the superradiant state for an arbitrarily chosen state @xmath4 . therefore the localized single - exciton states may be decomposed into a superposition of the superradiant ( [ bright ] ) and subradiant state defined in the formula ( [ subi ] ) according to the equation @xmath176 the above formula allows to define a sub- and superradiant component in every single - exciton state @xmath177 since there is only one superradiant state in any system of @xmath0 qds , the short - living contribution @xmath178 is proportional to the state ( [ bright ] ) with a weight factor @xmath179 , while the stable part @xmath180 is a combination of @xmath0 subradiant states ( [ subi ] ) , which as a sum of dark states remains optically inactive irrespective of a number of emitters . the component @xmath181 is proportional to the subradiant state defined by the formula ( [ dark2 ] ) with an amplitude @xmath182 . r. h. dicke , phys . rev . * 93 * , 99 ( 1954 ) . h. m. nussenzweig , _ introduction to quantum optics _ , ( gordon and breach , nowy jork , 1973 ) . g. s. agarwal , _ quantum statistical theories of spontaneous emission and their relation to other approaches _ , springer tracts in modern physics vol . * 70 * , edited by g. hhler ( springer , berlin , 1974 ) . m. gross and s. haroche , phys . rep . * 93 * , 301 ( 1982 ) . m. r. singh , w. lau , phys . a * 231 * , 115 ( 1997 ) . n. skribanowitz , i. p. herman , j. c. macgilvray , m. s. feld , phys . lett . * 30 * , 309 ( 1973 ) . d. pavolini , a. crubellier , p. pillet , l. cabaret , s. liberman , phys . lett . * 54 * , 1917 ( 1985 ) . m. scheibner , t. schmidt , l. worschech , a. forchel , g. bacher , t. passow , and d. hommel , nat . phys . * 3 * , 106 ( 2007 ) . k. baumann , ch . guerlin , f. brennecke , and tilman esslinger , nature * 464 * , 13011306 ( 2010 ) . j. m. fink , r. bianchetti , m. baur , m. gppl , l. steffen , s. filipp , p. j. leek , a. blais , and a. wallraff , phys . 103 * , 083601 ( 2009 ) . j. a. mlynek , a. a. abdumalikov , jr . , j. m. fink , l. steffen , m. baur , c. lang , a. f. van loo , and a. wallraff , phys . a * 86 * , 053838 ( 2012 ) . p. f. herskind , a. dantan , j. p. marler , m. albert , and m. drewsen , nature physics * 5 * , 494 - 498 ( 2009 ) . o. kyriienko , a. v. kavokin , and i. a. shelykh , arxiv:1211.0688v2 ( 2013 ) . a. a. belyanin , v. v. kocharovsky , vl . v. kocharovsky , and d. s. pestov , laser physics * 13 * , 61167 ( 2003 ) . d. meiser and m. j. holland , phys . rev . a * 81 * , 033847 ( 2010 ) . d. meiser and m. j. holland , phys . a * 81 * , 063827 ( 2010 ) . j. g. bohnet , z. chen , j. m. weiner , d. meiser , m. j. holland , and james k. thompson , nature * 484 * , 78 ( 2012 ) . a. d. ludlow , t. zelevinsky , g. k. campbell , s. blatt , m. m. boyd , m. h. g. de miranda , m. j. martin , j. w. thomsen , s. m. foreman , j. ye , t. m. fortier , j. e. stalnaker , s. a. diddams , y. l. coq , z. w. barber , n. poli , n. d. lemke , k. m. beck , and c. w. oates , science * 319 * , 1805 ( 2008 ) . y. y. jiang , a. d. ludlow , n. d. lemke , r. w. fox , j. a. sherman , l .- s . ma , and c. w. oates , nature photonics * 5 * , 158 ( 2011 ) . c. w. chou , d. b. hume , t. rosenband , and d. j. wineland , science * 329 * , 1630 ( 2010 ) . t. m. fortier , n. ashby , j. c. bergquist , m. j. delaney , s. a. diddams , t. p. heavner , l. hollberg , w. m. itano , s. r. jefferts , k. kim , f. levi , l. lorini , w. h. oskay , t. e. parker , j. shirley , and j. e. stalnaker , phys . 98 * , 070801 ( 2007 ) . s. blatt , a. d. ludlow , g. k. campbell , j. w. thomsen , t. zelevinsky , m. m. boyd , j. ye , x. baillard , m. fouch r. l. targat , a. brusch , p. lemonde , m. takamoto , f .- hong , h. katori , and v. v. flambaum , phys . . lett . * 100 * , 140801 ( 2008 ) . m. gross , c. fabre , p. pillet , and s. haroche , phys . lett . * 36 * , 10351038 ( 1976 ) . s. filipp , a. f. van loo , m. baur , l. steffen , and a. wallraff , phys . a * 84 * , 061805(r ) ( 2011 ) . y. takasu , y. saito , y. takahashi , m. borkowski , r. ciuryo , and p. s. julienne , phys . lett . * 108 * , 173002 ( 2012 ) . n. yamamoto , arxiv:1210.2632 ( 2012 ) . x . gong , x .- b . niu , j. li , y .- f . huang , and g .- c . guo , phys . a * 77 * , 042317 ( 2008 ) p. zanardi , m. rasetti , rev . lett . * 79 * , 3306 ( 1997 ) . p. zanardi , f. rossi , phys . rev . lett . * 81 * , 4752 ( 1998 ) . d. petrosyan and g. kurizki , phys . lett . * 89 * , 207902 ( 2002 ) . a. kalachev , phys . a * 76 * , 043812 ( 2007 ) . a. kalachev , v.v . samartsev , quantum electronics * 35*(8 ) , 679 ( 2005 ) . a. kalachev and s. krll , phys . a * 74 * , 023814 ( 2006 ) s. b. zheng , phys . a * 74 * , 054303 ( 2006 ) . p. agrawal and a. pati , phys . a * 74 * , 062320 ( 2006 ) . l. li and d. qiu , j. phys . a : math . theor . * 40 * , 10871 ( 2007 ) . h. situ and d. qiu , j. phys . theor . * 43 * , 055301 ( 2010 ) . radel ben - av and iaakov exman , phys . rev a * 84 * , 014301 ( 2011 ) . e. pazy , i. damico , p. zanardi , and f. rossi , phys . b * 64 * , 195320 ( 2001 ) . t. unold , k. mueller , c. lienau , t. elsaesser , and a. d. wieck , phys . lett . * 94 * , 137404 ( 2005 ) . e. biolatti , r. c. iotti , p. zanardi , and f. rossi , phys . lett . * 85 * , 5647 ( 2000 ) . m. goryca , t. kazimierczuk , m. nawrocki , a. golnik , j. a. gaj , p. kossacki , p. wojnar , and g. karczewski , phys . lett . * 103 * , 087401 ( 2009 ) . s. mokhlespour , j. e. m. haverkort , g. slepyan , s. maksimenko , and a. hoffmann , phys . b * 86 * , 245322 ( 2012 ) . w. dr , g. vidal , and j. i. cirac , phys . rev . a * 62 * , 062314 ( 2000 ) . y. n. chen , c. m. li , d. s. chuu , and t. brandes , new j. of phys . * 7 * , 172 ( 2005 ) . j. p. kestner and s. das sarma , phys . a * 84 * , 012315 ( 2011 ) . m. busl , r. snchez , and g. platero , phys . b * 81 * , 121306(r ) ( 2010 ) . p. trocha , and jzef barna , phys . b * 78 * , 075424 ( 2008 ) . e. vernek , p. a. orellana , and s. e. ulloa , phys . b * 82 * , 165304 ( 2010 ) . q. wang , h. xie , y .- h.nie , and w. ren , phys . b * 87 * , 075102 ( 2013 ) . a. grodecka and p. machnikowski , phys . b * 73 * , 125306 ( 2006 ) . a. sitek and p. machnikowski , phys . b * 75 * , 035328 ( 2007 ) . b. szafran , t. chwiej , f. m. peeters , s. bednarek , j. adamowski , and b. partoens , phys . b * 71 * , 205316 ( 2005 ) . b. szafran , acta phys . a * 114 * , 1013 ( 2008 ) . b. w. lovett , j. h. reina , a. nazir , and g. a. d. briggs , phys . b * 68 * , 205319 ( 2003 ) . m. kozub , . pawicki , p. machnikowski , phys . rev . * 86 * , 121305 ( 2012 ) . marlan o. scully , m. suhail zubairy , quantum optics ( cambridge university press , cambridge , 1997 ) j. danckwerts , k. j. ahn , j. frstner , and a. knorr , phys . b * 73 * , 165318 ( 2006 ) . m. r. singh , phys . rev . a * 75 * , 043809 ( 2007 ) . m. r. singh , ch . racknor , and d. schindel , appl . phys , lett . * 101 * , 051115 ( 2012 ) . o. benson , c. santori , m. pelton , and y. yamamoto , phys . * 84 * , 2513 ( 2000 ) . z. yuan , b. e. kardynal , r. m. stevenson , a. j. shields , c. j. lobo , k. cooper , n. s. beattie , d. a. ritchie , m. pepper , science * 295 * , 102 ( 2002 ) .

the idea behind the sp was put forward by efetov in 1979 in the context of layered superconductor with parallel magnetic field @xcite . it was suggested that the field can suppress the inter - layer josephson coupling so that the low energy properties of this 3d system can be described as being essentially of 2d character . later , in ref . @xcite it was shown that the frustration due to the magnetic field is not sufficient to fully suppress the coupling . in the context of quantum 1d chains ( equivalent to 2d classical layers by the virtue of the quantum to classical mapping ) the possibility of the decoupling between chains has also been explored @xcite . the main argument for such a decoupling is based on the scaling dimensions of the josephson coupling determined with respect to the luttinger liquid parameter in each chain : if it is larger than 2 , the coupling should become irrelevant @xcite . these proposals have been criticized in refs.@xcite where it was shown that the inter - chain tunneling is always relevant . further argumentation in favor of the sp and , actually , the name _ sliding phases _ have been proposed in refs.@xcite where the inter - layer gradient couplings between classical xy variables in each layer have been considered in addition to the josephson one . such gradient terms can independently control the scaling dimensions of the josephson coupling and of the vortex fugacity in each layer so that the first one can become irrelevant above some temperature @xmath0 ( of the _ dimensional reduction _ @xcite ) which is below the temperature of the berizinskii - kosterlitz - thouless ( bkt ) transition in the layers . thus , there is a range of temperatures where the sp are supposed to exist . this approach was also developed for the case of quantum 1d luttinger liquids coupled by both the josephson and the gradient terms @xcite which are the analog of the andreev - bashkin drag effect @xcite . it is important to note that the proposal of sp is based on applying the rg logic to compact variables characterized by global u(1 ) symmetry . while these early suggestions were more of a purely academic interest , expanding capabilities of ultra - cold - atoms techniques in recent years emphasize the importance of these suggestions especially in the context of possible new phases in composite lattices @xcite and in the presence of disorder @xcite . in more general terms , the question is if it is possible to realize a phase transition from a low- to higher- dimensional behavior . here we will analyze a simplest classical xy system characterized by the gradient interactions and the josephson coupling @xmath1 . the gradient terms are chosen in such a way that the sp is supposed to exist in some range where the renormalized value @xmath2 of @xmath1 scales to zero as layers size @xmath3 grows . we will present the results of the large scale monte carlo simulations of this system . our analysis is based on the dual formulation of the model in terms of the closed loops . the main result is that no sp state exists in such a system . instead , the value of @xmath2 is always finite . this behavior will be compared with the standard asymmetric xy layered model where no sp are expected to occur . we will also derive the analytical result for @xmath2 in the asymptotic limit when the intra - layer stiffness is much larger than @xmath1 . the numerical results have been found to be consistent with the analytical solutions for both models . our paper is organized as follows . in sec.[sec : i ] we introduce the bilayer model and provide the rg solution for sp . then , we construct the dual representation in sec . [ sec : dual ] . using the duality we have found the asymptotic analytical solution for the renormalized josephson coupling @xmath2 in sec . [ sec : as ] . the monte carlo simulations of the bilayer model are presented in sec.[sec : num ] . then , in sec.[sec : nz ] we present the results on a stack of bilayers along the same lines as for the bilayer . finally , in sec . [ sec : dis ] we discuss the implications of our analytical and numerical results and also provide an alternative model for the sp . here we introduce a model of two asymmetric parallel layers , each being a square lattice of linear size @xmath4 ( in terms of the inter - site shortest distance ) characterized by two fields @xmath5 and @xmath6 on the layers @xmath7 , respectively.the action can be written as h&= & - _ ij [ t_1 ( _ ij _ 1 - a_ij ) + t_2 ( _ ij _ 2 -g_2 a_ij ) + & + & a^2_ij ] - _ i u(_2(i)-_1(i ) ) [ 2n ] where @xmath8 and @xmath9 are parameters ; @xmath10 denotes summation over nearest neighbor sites within each layer ; @xmath11 ; @xmath12 is a bond vector field ( that is , @xmath13 ) oriented along the bond @xmath10 . it is introduced in order to generate the `` current - current '' interaction ( cf . @xcite ) consistent with the compact nature of the fields @xmath14 . this action is to be used in the partition function z = da d_1 d_2 ( - h ) [ zz2 ] where the temperature is absorbed into the the parameters @xmath15 . our focus is on verifying the applicability of the rg analysis to the renormalization of the josephson coupling @xmath1 . hence , we will not discuss physical origins of the variables and the parameters . in the approximation ignoring compact nature of the variables , the terms @xmath16 and @xmath17 are replaced by @xmath18 and @xmath19 , respectively . then , the gaussian integration over @xmath12 can be carried out explicitly in eq.([zz2 ] ) , so that ( [ 2n ] ) in terms of the remaining variables becomes h_0= _ ij k_ab _ ij _ b - _ i u(_2-_1 ) , [ 2n0 ] where the matrix @xmath20 is related to the original parameters as k_11&= & , k_22= , + k_12 & = & - . the stability of @xmath21 is guaranteed by k_11k_22 - k_12 ^ 2= > 0 . [ stab ] the condition for sp can be obtained along the lines of the logic @xcite which ignores the compactness of @xmath14 . specifically , introducing the variables @xmath22 and @xmath23 and , then , integrating out @xmath24 , the resulting partition function becomes z_0=d^-h _ , h _ = d^2 x , [ z_0 ] where the notation k= [ k ] is introduced and the long wave limit is considered so that the summation along the layers is replaced by the integration @xmath25 . as long as the compactness of @xmath26 is ignored , eqs . ( [ z_0 ] ) represent the standard sine - gordon model in 2d . the rg analysis predicts ( see in , e.g. , @xcite ) that at @xmath27 the renormalization renders the josephson coupling @xmath1 irrelevant in the thermo - limit @xmath28 . more specifically , the renormalized @xmath1 should flow to zero as @xmath29 . such a behavior is supposed to occur together with the persistence of the algebraic order along the planes . this requirement imposes further restrictions on the values of @xmath30 . without loss of generality let s assume @xmath31 and introduce the notations : @xmath32 as a measure of temperature , and @xmath33 . then , the condition @xmath34 for sp becomes t > t_d=. [ ssp ] in order to guarantee the algebraic order in each layer no bkt transition should occur in the layers . in order to determine possible types of vortices responsible for the transition , we examine the form ( [ 2n0 ] ) by `` reinstating '' the compactness of the variables in the limit @xmath35 ( which is supposed to renormalize to zero ) . in the presence of the gradient coupling a vortex may have a composite structure @xcite : @xmath36 circulations in the component 1 may be bound to @xmath37 circulations in the component 2 . finding the condition for the proliferation of such composite vortices can be achieved along the line of kosterlitz - thouless argument developed for simple xy - model in 2d . the free energy of such a complex is f_v= l -2 l. [ vort ] then , @xmath38 - 2 t $ ] , and the stability against the bkt transition is guaranteed by the positivity of @xmath39 or t < t_(q_1,q_2)= [ ( q_1 + xq_2)^2 + ( y - x^2)q^2_2 ] , [ tbkt ] where the minimization with respect to @xmath40 must be performed . proliferation of simple vortices @xmath41 corresponds to @xmath42 . it is , however , not always a minimal value as long as @xmath43 . let s also note that , since @xmath44 by definition , there are no solutions for @xmath45 if @xmath46 . let s look for a solution when @xmath47 is integer , that is , @xmath48 . then , keeping in mind the stability requirement ( [ stab ] ) , that is , @xmath49 , the minimal vortex corresponding to @xmath50 has free energy lower than that of the simple vortex if @xmath51 . in this case , the solution for @xmath52 exists if @xmath53 . proliferation of the composite vortex pairs corresponds to disordering of the original fields @xmath54 , while the composite field @xmath55 remains ( algebraically ) ordered . this mechanism constitutes the formation of thermally induced bound phases ( or using the language of superfluidity _ thermally paired superfluid _ , tps , @xcite ) . in other words , the system behaves in a such a way that the algebraic order persists only in the composite field @xmath56.[since @xmath57 we call such a composite phase as thermally bound superfluid ( tbs ) by analogy with the tps ] . this effect does not require that @xmath47 is necessarily integer . if @xmath47 is non - integer , its closest integer part will determine the power of @xmath58 . if @xmath59 , the lowest energy vortex is the simple one , and the solution for the inequality @xmath45 also requires @xmath53 . for @xmath60 , @xmath61 while @xmath62 as long as @xmath63 is kept constant . such a limit corresponds to the largest range of @xmath64 where sp are to be anticipated for the two - layer model . however , for practical purposes of simulations using too large @xmath47 leads to slower convergence . thus , we choose @xmath65 corresponding to a reasonably wide range where sp is anticipated to exist . taking into account eq.([k ] ) , the condition for the sp can be written as < t < . [ spt ] keeping in mind the chosen values @xmath66 , this becomes @xmath67 or @xmath68 . the results of the simulations will be conducted at the `` temperature '' @xmath64 in the middle of the interval @xmath69 , that is , @xmath70 . more specifically , @xmath71 . in terms of the original parameters @xmath72 , the relations @xmath73 , @xmath74 , @xmath75 and @xmath76 must be satisfied . [ such a fine tuning is of no concern , since the focus is on the consistency of the paradigm of sp rather than on physical origin of the values ] . concluding this section , the presented analysis based on the rg finds that it is possible to find a range of temperatures where the sequence of phases is as presented in fig . [ pd ] in the panel ( a ) : at @xmath77 the josephson coupling is relevant . we call this range as josephson phase in fig . [ pd ] . at @xmath78 there is the sp where the symmetry @xmath79 is promoted to u(1)@xmath80 u(1 ) . in the range @xmath81 the tbs phase is characterized by the composite field @xmath82 . thus , the broken symmetry is partially restored through the subgroup @xmath83 , where @xmath84 $ ] . at higher temperatures @xmath85 the composite field @xmath82 becomes disordered . -8 mm -5 mm the partition function @xmath86 can be evaluated by the high - temperature expansion method ( see e.g. in @xcite ) in terms of @xmath87 and the explicit integration over the variables . this approach allows obtaining @xmath86 in terms of the integer bond variables powers of the corresponding taylor series . since the resulting configurational space consists of closed loops of the bond currents , further simulations can be effectively performed by the worm algorithm @xcite . as will be also shown , the language of loops also allows obtaining analytic expressions for the renormalized josephson coupling @xmath2 which are exact in the asymptotic limit . we will be utilizing the villain approximation @xcite for the cosines to obtain the so called j - current version @xcite of eqs.([zz2]),([2n ] ) : z=_\{m_a , ij , m_i } dda e^-h_v , [ vill ] h_v&= & _ ij [ ( _ ij _ 1 - a_ij + 2m_1,ij)^2 + & + & ( _ ij _ 2 -g_2 a_ij + 2m_1,ij ) ^2 + a^2_ij ] + & + & _ i ( _ 2(i)-_1(i ) + 2m_i)^2 , [ 2nv ] where @xmath88 ( @xmath89 ) are integer numbers defined along bonds between two nearest sites @xmath90 and @xmath91 along the planes and @xmath92 is an integer assigned to a site @xmath90 and oriented from the layer 1 to the layer 2 . the villain approximation proves to be very accurate for establishing the transition points as well as in general if the effective constants @xmath93 are properly expressed in terms of the corresponding bare values @xmath87 ( see in ref.@xcite ) . the `` renormalization '' can be essentially ignored for @xmath94 , so that in what follows we will be using @xmath95 . similarly , for the josephson coupling @xmath96 one should take @xmath97 and , if @xmath98 , the corresponding relation is @xmath99 @xcite . after using the poisson identity for each integer and performing the integrations over @xmath100 and @xmath101 , the resulting expression becomes z=_\{j_1,ij},\{j_2,ij } , \{j_z , i } e^-h_j , [ z2 ] h_j= _ ij ; a , b ( k^-1)_ab j_a , ij j_b , ij + _ i j_i^2 , [ h_j ] where @xmath102 labels layers and @xmath103 is the matrix inverse to @xmath30 introduced in eqs.([2n0]),([ktg ] ) ; the summation runs over the integer bond currents @xmath104 ( oriented from site @xmath90 to site @xmath91 so that @xmath105 ) within each corresponding layer 1,2 as well as over the integer currents @xmath106 oriented along the bond connecting the site @xmath90 in the layer 1 to the site @xmath90 in the layer 2 . all the configurations are restricted by the kirchhoff s current conservation rule the total of all j - currents incoming to any site must be equal to the total of all outcoming currents from the same site . it is useful to note that eqs.([z2],[h_j ] ) can also be obtained directly from eq.([2n0 ] ) by reinstating the compact nature of the variables : @xmath107 and @xmath108 so that the matrix @xmath109 is viewed as being independent from the parameters in the action ( [ 2n ] ) . the system ( [ z2],[h_j ] ) features statistics of closed loops . if @xmath35 , there are two sorts of loops one in each layer . thus , each configuration is characterized by definite values of the windings @xmath110 in the @xmath111th layer along the @xmath112 directions of the planes . it is straightforward to show that statistics of these windings determine the renormalized values @xmath113 of the matrix @xmath30 along the line of the approach @xcite . more specifically _ ab = _ = , w_a , w_b , . [ kr ] this expressions are valid for periodic boundary conditions ( pbc ) . it is important to note that @xmath113 represents an exact linear response ( at zero momentum ) with respect to thouless phase twists . in other words , if there are externally imposed infinitesimal constant gradients @xmath114 ( violating the pbc ) of the phases @xmath14 , the free energy acquires the contribution @xmath115 . on the other hand , in the presence of the gradient the integrand of the partition function gets the factor @xmath116 . comparing both expressions leads to the relation ( [ kr ] ) . as a test of consistency , we have checked numerically that in the regime where the sp state is supposed to exist ( that is , @xmath117 ) , the deviations of @xmath113 from the bare values @xmath30 are within the statistical error less than 1% for all tested sizes of the layers @xmath118 . significant deviations are observed only as the system approaches fully disordered state that is , where the fields @xmath119 as well as the composite one @xmath82 become disordered . in this case , @xmath113 flow to zero as @xmath3 increases . the deviations remain small ( about 2 - 3% ) even in the regime where @xmath82 is the only ordered field . the emergence of the tbs is detected by observing that windings @xmath120 in the layers 1 and 2 are changing exactly by the increment @xmath121 , @xmath122 ( plus or minus ) , respectively . at finite values of @xmath1 the loops belong to both layers so that no separate windings can be introduced . however , the sums @xmath123 remain well defined and can be used to evaluate the rigidity @xmath124 of the fields along the layers . in a general case of @xmath125 symmetric ( with respect to the @xmath126 directions ) layers @xmath127 : & = & _ w^2 _ [ stif22 ] + w _ & = & _ ij , a=1,2, ... n_z j_a , ij , [ ww ] where for a given @xmath112 in ( [ ww ] ) the bond @xmath128 ( as well as @xmath129 ) is oriented along the direction @xmath130 . our focus here on the renormalized value @xmath2 of the josephson coupling @xmath1 in the sp regime . if the periodic boundary conditions are also imposed perpendicular to the layers ( along @xmath131-direction ) , the inter - layer response @xmath2 is given by windings @xmath132 along @xmath131-direction : u_r = w^2_z , w_z = _ i j_i , [ wu ] where the summation @xmath133 of the currents @xmath134 ( oriented along @xmath131-direction ) is performed over all sites of all layers . similarly to the cases ( [ kr ] ) and ( [ stif22 ] ) , eq.([wu ] ) represents the full linear response at zero momentum that is , the renormalized value @xmath2 of the josephson coupling @xmath1 . at this point , we should comment on how to interpret the pbc for two layers , @xmath135 . while in the case @xmath136 it is a natural procedure to link the @xmath137th layer to the first one , @xmath138 , by the josephson term , the case @xmath135 needs an auxiliary construction because the layers 1 and 2 are coupled already directly . the formal procedure , then , consists of adding a third layer , @xmath139 , with no rigidity along @xmath126 directions and coupled by the josephson term to both layers , @xmath7 . if the coupling @xmath140 between the layers 1 and 3 and the coupling @xmath141 between the layers 2 and 3 add up as @xmath142 , in the dual action ( [ h_j ] ) the sum in the last term can be extended to the layers @xmath143 , while the first term is still confined to the layers @xmath7 . the key to this procedure is the kirchhoff s rule : the j - current from a site @xmath144 along @xmath131-direction from the layer 2 to the layer 3 must be exactly the same as the current from the site @xmath144 in the layer 3 to the layer 1 . then , in the form ( [ h_j ] ) the same value @xmath145 can be used for the currents from the layer 1 to the layer 2 directly or through the layer 3 . as mentioned above , the dual representation allows obtaining analytically the asymptotic values for @xmath2 . let s begin with the trivial case of zero stiffnesses @xmath30 and arbitrary number of layers , @xmath146 . in this case , the action ( [ h_j ] ) trivially becomes h_a= _ i j^2_i , j_i=0,1 , 2 , ... [ a ] where the summation runs over all sites @xmath90 of only _ one _ layer , say , @xmath138 . in this expression the kirchhoff rule dictates that the current @xmath134 at a given site along @xmath131-direction must be the same for all values of @xmath131 . thus , such a current with @xmath147 constitutes one closed loop characterized by the winding @xmath148 . this allows constructing the partition function exactly as z_a=^l^2 [ za ] where @xmath149 is the number of sites in one layer . the stiffness ( [ wu ] ) can be found by taking into account that the total winding along z - direction is @xmath150 , where the summation runs over @xmath149 sites of only one layer . then , using statistical independence of different sites we find u_r=. [ ur1 ] this expression shows that , as long as @xmath125 is finite , the josephson coupling remains relevant even if there is no in - plane order . in the limit @xmath151 only the term @xmath152 is important , so that eq.([ur1 ] ) becomes u_r= . [ urr ] eq.([urr ] ) can be used even in the case when there is finite stiffness along the layers , with @xmath125 substituted by some effective value @xmath153 , that is , u_r= . [ genm ] this is easily justified in the `` ideal gas '' approximation of rare fluctuations of the j - currents in z - direction . in the case of @xmath154 the asymptotic asymmetric limit corresponds to @xmath155 ( so that the system is well above the disordering transition ) and @xmath156 . the loop proliferation can be viewed from the perspective of the worm algorithm @xcite where one open end of a string of j - currents walks randomly until it meets another open end so that the loop is formed . then , the most of the path is residing in a layer with only occasional jumps between neighboring layers . these jumps are controlled by the exponential smallness of @xmath157 . thus , the full action can be well approximated by @xmath158 , eq.([a ] ) , with @xmath159 . this leads to eq.([genm ] ) with @xmath160 : u_r= . [ urxy ] later we will present the numerical evidence that the stiffness perpendicular to the layers of a simple xy layered model in the asymptotic limit can well be described by the above equation . below we will show that for the case of the two asymmetric layers , the effective value of @xmath153 is @xmath161 in eq.([genm ] ) . here we present the results of monte carlo simulations of the bilayer in the regime of sp in the limit @xmath156 . the action ( [ h_j ] ) can be represented in the notations @xmath162 as h_j= _ ij+ _ i , [ h12 ] where @xmath163 and @xmath164 refer to the inplane bond currents in the layers 1 and 2 , respectively ; the actual values of the parameters used in the simulations have been discussed at the end of sec.[sec : rg ] : @xmath165 . the structure of the loops is determined by the energy of creating a j - current element along a given direction . a typical energy to create an additional j - current element in the plane 2 can be estimated as @xmath166 . thus , large loops with a typical values @xmath167 can exist in the plane 2 . in contrast , the energy to create an isolated element in the plane 1 ( with no @xmath168 currents along the same bond in the layer 2 ) costs much larger energy : @xmath169 . thus , the probability to create such an element is exponentially suppressed as @xmath170 , and no entropy contribution ( due to 4 optional directions along the plane ) can compensate for such a low value . this implies that no large isolated loops can exist in the layer 1 . the only option to create a large loop in the layer 1 is if each element @xmath163 is coupled to currents @xmath171 along the same bonds in the layer 2 . a typical energy of this combined element is @xmath172 . this strong asymmetry between the layers has immediate implication for the windings along @xmath131-direction the minimal length @xmath153 of the element @xmath134 must be @xmath161 in eq.([genm ] ) . thus , the stiffness @xmath2 in the limit @xmath98 becomes u_r = 4e^-1/u_v= u^2 , [ ur3 ] where the asymptotic expression @xmath173 @xcite has been used . the results of the simulations is shown in fig.[fign2 ] . -8 mm -5 mm the first striking feature to notice is that @xmath2 does not depend on the layers size @xmath3 over 2 orders of magnitude of @xmath3 and over 7 orders of @xmath145 ( which is actually @xmath174 of the bare coupling ) . second , the numerically found value @xmath2 follows the analytical result ( [ ur3 ] ) with high accuracy even for values @xmath175 . both features are in the striking conflict with the rg prediction stating that @xmath2 should decay as @xmath176 in the sp regime ( @xmath177 ) . it should be also noted that the stiffness along the layers ( [ stif22 ] ) remains finite and much larger than @xmath2 , that is , @xmath178 for all simulated sizes from @xmath179 to @xmath180 . this justifies the validity of eq.([ur3 ] ) even in the case @xmath175 . as it became evident from the previous analysis , no sp can occur in the double layer . referring to the sketch of the possibilities , fig . [ pd ] , the option ( b ) is realized . here we will address a possibility of sp in a @xmath125-layers setup . in other words , we will be looking for a behavior where the renormalized stiffness ( the inter - layer josephson coupling ) @xmath2 decays as a function of @xmath125 in the limit @xmath182 , while the stiffness along planes remains finite.[this would be a `` weaker '' version of the sp ] . we consider the pbc setup : the odd @xmath183 and the even @xmath184 layers are characterized by the inplane stiffnesses @xmath185 and @xmath186 , respectively , with the nearest layers coupled by the current - current term @xmath187 ( the same for all pairs of layers ) as well as by the josephson coupling @xmath188 , where @xmath189 is the xy variable defined on a site @xmath144 belonging to the layer @xmath131 . in the linearized with respect to @xmath190 approximation analogous to eq.([2n0 ] ) the model becomes h_0&= & _ z=1,3,5, ... \ { _ ij h_z;ij + & - & u _ x , y [ ( _ z+1-_z ) + ( _ z-1-_z ) ) ] } [ nnz ] where the summation runs over odd values of @xmath131 and the notation h_z;ij = ( _ ij _ z)^2 + ( _ ij _ z+1)^2 + + k_12_ij _ z(_ij _ z+1+_ij _ z-1 ) [ hnnz ] is used . if the compact nature of @xmath191 is ignored , the one - loop rg flow equation for @xmath2 reads = ( 2- ( _ z+1 - _ z)^2 _ s ) u_r [ rgn ] where @xmath192 refers to the rg shell integration over the inplane momenta @xmath193 , with @xmath194 being the cutoff and @xmath195 . we note that , due to the pbc along @xmath131-direction , the mean @xmath196 does not depend on @xmath131 . using discrete fourier representation along @xmath131 direction with doubled unit cell containing two layers ( the odd and the even ) with two sorts of phases @xmath197 and @xmath198 along odd and even layers , respectively , the part @xmath199 can be diagonalized and the correlator in eq.([rgn ] ) found . this gives eq.([rgn ] ) rewritten as = 2(1- ) u_r , [ rgn2 ] where t^-1_d=^(nz/2)-1_m=0 , [ tdnz ] and the wavevectors along @xmath131 take values dictated by the periodic boundary conditions @xmath200 . here we use the same notations @xmath201 introduced in sec.[sec : rg ] . thus , at @xmath177 rg predicts irrelevance of @xmath2 . the upper limit on @xmath64 can be obtained from the requirement of no free 2d vortices in the limit @xmath35 . there are , actually , two options : i ) looking for a composite vortex characterized by phase windings @xmath36 and @xmath37 in odd and even layers , respectively , forming a string of length @xmath125 perpendicular to the layers ; ii ) considering independent vortices @xmath202 only in odd layers ( characterized by smallest stiffness @xmath185 ) and characterized by the temperature @xmath203 . as the analysis shows , the option when composite vortices form finite strings ( say , @xmath204 in layers @xmath205 and @xmath206 in the layer @xmath207 by analogy with the @xmath135 case ) costs larger energy than in the case ii ) . the option i ) is characterized by the vortex energy @xmath208\propto n_z$ ] . the minimum for a @xmath209 integer is achieved at @xmath210 . thus , in order to compensate for the factor @xmath211 , the system must be very close to the instability @xmath212 . simulations in this region turn out to be problematic . thus , we will conduct simulations in the range @xmath213 , provided the condition @xmath214 holds in the limit where @xmath215 remains finite for @xmath216 . specifically , the condition @xmath217 reads ^nz/2 -1_m=0 > 8 . [ tdt ] it can surely be achieved for large enough @xmath47 in the limit @xmath218 . replacing the summation by integration in this limit and considering @xmath219 , eq.([tdt ] ) gives @xmath220 . for the simulations we have chosen @xmath221 and @xmath222 , which gives @xmath223 with @xmath224 chosen in the middle of the interval between @xmath225 and @xmath0 . the chosen value of @xmath0 corresponds to the limit @xmath226 , and for any finite @xmath125 , the actual @xmath0 from eq.([tdnz ] ) is below this value . at this point we note that for any finite @xmath1 the system is 3d and , strictly , speaking the notion of the bkt transition becomes inadequate for large enough @xmath125 : even at @xmath227 the odd layers would still have xy order due to the proximity to even layers . here , however , we presume that sp scenario holds and @xmath2 vanishes at large @xmath3 . practically , simulations are performed at finite @xmath1 and we always control that the xy stiffness ( helicity modulus ) along the layers remains finite and independent of @xmath3 . the dual formulation in terms of the closed loops of integer j - currents ( along bonds in and between the layers ) can be achieved similarly to the case @xmath135 by reinstating the compactness of @xmath191 in eqs.([nnz],[hnnz ] ) through the villain approach : @xmath228 along the planes and @xmath229 for josephson coupling , where @xmath230 refers to an arbitrary ( oriented ) integer defined on the bond @xmath231 belonging to the plane @xmath131 and @xmath232 stands for an integer on a bond connecting site @xmath90 in the plane @xmath131 to the same site in the plane @xmath233 . the partition function is obtained as a result of integration over all @xmath234 and summations over all bond integers . the j - currents enter through the poisson identity @xmath235 applied to each bond integer . this allows explicit integration over all phases @xmath191 as well as over the bond integers @xmath236 . there are two types of j - currents : inplane @xmath237 within each `` elementary cell '' ( along @xmath131 ) and between the planes @xmath238 . the label @xmath239 refers to j - current defined on the bond @xmath231 belonging to a plane with odd @xmath131 . accordingly , @xmath240 stands for the inplane current on the layer with even @xmath131 . then , @xmath238 denotes the current from the site @xmath90 from the plane @xmath131 to the plane @xmath233 . the integration over phases @xmath241 generates the kirchhoff constraint similarly to the bilayer case . finally the ensemble can be represented as z&= & _ \ { } ( - h_j ) , + h_j&= & _ ij ; z , zv_ab(z - z ) j^(a)_z , ijj^(b)_z,ij + & + & _ i , z j^2_z , i , [ hdual ] where the matrix @xmath242 is defined in terms of the matrix @xmath30 . it reflects the asymmetry between odd and even layers . explicitly , @xmath243 , for @xmath244 being even , describes the interaction between odd layers , and @xmath245 is defined between even layers ; @xmath246 $ ] refers to the interaction between odd and even layers ( that is , @xmath131 is odd ) , and v_22(z)= _ q_m , [ vzz22 ] with @xmath247 and the summation running over @xmath248 . the dual formulation for @xmath125 layers allows obtaining the asymptotic expression for @xmath2 within the same logic used for deriving eq.([ur3 ] ) . we will repeat it here . the loop formation can be viewed as a process of random walks of two ends of a broken loop exactly along the line of the worm algorithm @xcite . such a walk of each end is controlled by energetics of creating one bond element @xmath249 in a randomly chosen direction either along a given plane or perpendicular to it . very similar to the case of the two layers , the energy to create such an element alone along the odd layer costs energy @xmath250 , while the same element along an even layer costs energy @xmath251 . the only option for creating a loop in an odd layer is if its energy is compensated by parallel elements in the even plane . this feature is caused by the strong current - current interaction @xmath252 . thus , if the walk occurs along @xmath131-direction from some even layer @xmath131 toward the neighboring odd layer @xmath233 , the subsequent move along the odd layer will be too energetically costly so that the walker would either move further toward @xmath253 layer or will go back to the original layer @xmath131 . thus , the inter - layer elements are characterized by either @xmath254 or @xmath255 . the weight of such a process is either @xmath256 or @xmath257 , respectively . even if the walker makes a step or two along the layer @xmath233 ( which is a highly improbable event ) and then chooses to go toward the layer @xmath253 , the contribution to the partition function will be further reduced exponentially by the energy of the element @xmath258 along the odd plane . thus , such processes can be ignored , and we arrive at the conclusion that @xmath2 given by eq.([ur3 ] ) must be valid for arbitrary @xmath125 in the asymptotic limit . the model ( [ hdual ] ) has been simulated by the worm algorithm @xcite . the renormalized inter - layer stiffness @xmath2 was found for a wide range of layer sizes , @xmath259 and @xmath260 . the resulting data is presented in figs . [ fig3],[fig4 ] . as can be seen in fig . [ fig3 ] , the solution ( [ ur3 ] ) plays the role of the envelop for the family of the curves @xmath2 vs @xmath261 for various @xmath3 and @xmath262 . we note that the stiffness @xmath263 along the layers ( as determined by eq.([stif22 ] ) ) remains independent of the sizes and much larger ( @xmath264 ) than @xmath2 . this justifies the applicability of the asymptotic limit for eq.([ur3 ] ) . we have also controlled that the system is far enough from any possible composite phases @xcite state by measuring the lowest order correlator @xmath265 and observing that it exhibits long - range order for values @xmath266 in the limit @xmath267.[in the composite phase state such a correlator is short ranged ] . thus , the system is well in the putative sp state . its behavior , however , is drastically different from the rg prediction . -8 mm -5 mm -8 mm -5 mm -8 mm -5 mm -8 mm -5 mm at this point we should discuss the deviations of the numerical curves from the analytical result seen in fig.[fig3 ] . to some extent this behavior could have been interpreted as the evidence of sp . there is , however , one important observation : the value of @xmath268 below which such suppression begins _ decreases _ as ( u^*)^-1 = ( l^2/n_z ) , = 1.000.02 [ ugm2 ] for @xmath269 in the main logarithmic approximation . this behavior is demonstrated in fig.[fig4 ] , where the value @xmath270 corresponds the offset for @xmath2 being taken at 1/10 of the value given by the analytical expression ( [ ur3 ] ) . the suppression of @xmath271 in the limit @xmath28 indicates that it is a finite size effect . in other words , starting from small @xmath3 at some @xmath272 , the renormalized stiffness @xmath2 can be well below the asymptotic limit ( [ ur3 ] ) . as @xmath3 increases while @xmath145 is kept fixed , @xmath2 eventually approaches the limit ( [ ur3 ] ) . this feature is clearly in a stark contrast with the rg prediction of sp where @xmath2 is supposed to flow to zero for fixed @xmath145 as a power of @xmath182 . the deviation from the limit ( [ ur3 ] ) has a very simple explanation : it is essentially a consequence of the generic exponential suppression of any order in a quasi-1d limit @xmath226 . [ there is no such suppression in the case of @xmath135 because the loops along z - direction are independent from the inplane ones ] . in the context of our system it can be interpreted as the effect of zero modes ( in each layer ) fluctuations . indeed , excitations along the planes renormalize @xmath145 to @xmath2 at short distances . then , as long as @xmath273 the only remaining lowest energy degrees of freedoms are zero modes , that is , excitations with @xmath274 being independent from the @xmath126 positions along the planes ( and dependent on @xmath131 ) . then , the effective action becomes s_0= -u_rl^2 _ z[(^(0)_z+1 - ^(0)_z ) ] . [ zero ] it should be used in the partition function @xmath275 . its analytical evaluation gives u^(0)_r ~ ( -n_z/(u_rl^2 ) ) [ zero ] for the value of the stiffness along @xmath131-direction that is , the renormalised @xmath2 on large scale . keeping in mind that at short scales @xmath276 for @xmath156 , one arrives at the relation ( [ ugm2 ] ) with @xmath277 . it corresponds to the requirement @xmath278 for some @xmath268 so that eq.([zero ] ) describes the exponential suppression at @xmath279 . clearly , such a quasi-1d suppression is also present in the standard xy model ( where no sp are anticipated to exist ) . in order to demonstrate this explicitly we have also simulated a simple xy model given by the system z_xy&=&d _ z ( -h_xy ) , [ hxy ] + h_xy&=&- _ ij , z [ ( _ ij _ z ) + u ( _ z _ z ) ] , with some @xmath280 ( guaranteeing that no bkt transition occurs in each layer for @xmath35 ) , and @xmath281 . in the dual representation this system is described by h_xy _ xy = _ ij , z j_z , ij^2 + _ i , z j_i , z^2 , [ hxy2 ] where @xmath282 and @xmath238 are the same j - currents introduced above for the model ( [ hdual ] ) . the results of the simulations of this model are presented in fig.[fig1],[fig2 ] . in the asymptotic limit the inter - layer stiffness is described by eq.([urxy ] ) . then , according to the above discussion the value @xmath270 determining the start of the deviations is given by @xmath283 , that is , with the slope @xmath284 which should be compared with the numerical value @xmath285 in fig.[fig2 ] . thus , both models demonstrate essentially the same 3d behavior . the only difference is the slope of the renormalized josephson coupling @xmath286 vs its bare value @xmath145 . it is determined by the minimal length of the elementary j - current in the direction perpendicular to the layers . the rg approach to 2d systems proves to be every effective in many cases including 2d xy model when it can be mapped on the sine - gordon ( sg ) one @xcite . a successful implementation of the rg analysis to the josephson coupling was demonstrated in ref.@xcite where a single weak link can make one channel luttinger liquid insulating . the merit of rg , however , should be taken with caution when applied to the dimensional reduction situations in layered systems . in this case there is no exact mapping between xy and sg representations at finite inter - layer josephson coupling , and the approximation ignoring the compact nature of the variables becomes uncontrolled . as our analysis of one particular layered system shows , no sp exists in such a system despite the rg prediction : the system shows essentially the 3d behavior of the asymmetric xy model . while it is not obvious to us what is wrong with treating josephson coupling between 2d layers by rg in the original representation of fields , the dual formulation in terms of the closed loops gives a very important insight . specifically , the sp means that as layer size @xmath28 , a suppression of the josephson coupling between layers would require that the number of times elements of closed loops fluctuate between layers must scale slower than @xmath149 so that the density of such events is zero in the limit @xmath287 . the loops statistics , however , is controlled by local energies of creating finite elements and the entropy due to 6 directions in 3d vs 4 along layers . thus , as long as there is a finite energy to cross between neighboring layers , the entropy will lead to a finite density of crossings for large enough @xmath3 . similar argument can be applied to quantum wires in terms of the quantum to classical mapping where imaginary time is treated as an extra dimension . the dual approach and the argumentation along the line of the numerical algorithm @xcite , treating closed loops formation as a process of the worm head wondering around and eventually finding its tail , allowed us to expose the actual stages of the renormalization of the josephson coupling : i ) at small scales josephson coupling is controlled by exponentially suppressed random and independent ( in the asymptotic limit ) events of crossings between layers . it can be viewed as an ideal gas of j - currents between the layers . this stage leads to the renormalized coupling , in general , represented by eq.([genm ] ) with @xmath288 . ii ) if the number of layers @xmath125 increases , with @xmath3 being fixed , quasi 1d fluctuations further suppress the coupling exponentially as demonstrated in eq.([zero ] ) . here we have discussed a local model characterized by short range interactions between the inter - layer j - current elements . this feature in combination with the low density of such elements justifies the `` ideal gas '' approximation for them , which in its turn leads to finite values of the renormalized inter - layer josephson coupling . the question may be raised if a presence of long - range forces between the inter - plane j - currents @xmath134 can change the situation and lead to the sp or its weaker version where @xmath289 with the growing number of layers @xmath125 in the limit @xmath287 . in this respect we note that in order to realize this , fluctuations of the difference of the j - currents with positive and negative orientations must be macroscopically suppressed . in this case the fluctuation of the winding numbers in @xmath131-direction @xmath290 will scale slower than @xmath149 so that @xmath291 . this may be caused by interactions between the inter - layer j - currents decaying not faster than the second power of their separation . more specifically , the following additional repulsive term h_sp=_i , j u(_i -_j ) j_i j_j [ spx ] in the simple xy j - current model ( [ hxy2 ] ) with @xmath292 having the long range tail @xmath293 with @xmath294 will generate the energy contribution @xmath295 in terms of the windings in z - direction . consequently , the renormalized josephson coupling ( [ wu ] ) would scale as @xmath296 . in one particular example long - range forces are introduced into the inter - layer josephson in the standard xy model ( [ hxy ] ) by some effective gauge - type term @xmath297 , where @xmath298 refers to the derivatives along the layers of some soft mode @xmath299 , with @xmath300 being a parameter . the resulting interaction in the dual form ( [ spx ] ) becomes @xmath301 and , thus , it suppresses the inter - layer josephson as @xmath302 in the limit @xmath28 for fixed @xmath125 . one of us ( abk ) acknowledges helpful discussions of the sliding phases with charles kane and boris svistunov . he thanks kitpc ( the program `` precision many - body physics of strongly correlated quantum matter '' , may - june 2014 ) and aspen center for physics ( the program `` beyond quasiparticles : new paradigms for quantum fluids '' , august 2015 ) for hospitality when this work was in progress . this work was supported by the national science foundation under the grant phy1314469 , and by a grant for computer time from the cuny hpcc under nsf grants cns-0855217 and cns - 0958379 . 99 k. b. efetov , sov . jetp * 49 * , 905 ( 1979 ) . s. e. korshunov and a. i. larkin , phys . b * 46 * , 6395 ( 1992 ) . p. w. anderson , phys . lett . * 64 * , 1839 ( 1990 ) . x. g. wen , phys.rev . * b 42 * , 6623 ( 1990 ) . c. castellani , c. di castro , and w. metzner , phys . lett . * 69 * , 1703(1992 ) . m. fabrizio , and a. parola , phys.rev.lett . * 70*,226(1993 ) . c. s. ohern and t. c. lubensky , phys.rev.lett . * 80 * , 4345(1998 ) . c. s. ohern , t. c. lubensky , and j. toner , phys.rev.lett . * 83 * , 2745 ( 1999 ) ; s. l. sondhi , kun yang , phys . rev . * b63*,054430(2001 ) . r. mukhopadhyay , c. l. kane , and t. c. lubensky , phys . * b 63 * , 081103(r ) ( 2001 ) ; a. vishwanath and d. carpentier , phys . * 86 * , 676 ( 2001 ) ; a. f. andreev and e. p. bashkin , sov . jetp * 42 * , 164(1976 ) . m. a. cazalilla , a. f. ho , and t. giamarchi , new j. phys . * 8 * , 158 ( 2006 ) d. pekker , g. refael , and e. demler , phys . 105 * , 085302 ( 2010 ) ; t. c. lubensky , p. m. chaikin , principles of condensed matter physics , cambridge university press ( 2000 ) . e. babaev , nucl.phys . * b 686*,397 ( 2004 ) . v. m. kaurov , a. b. kuklov , and a. e. meyerovich phys . lett . * 95 * , 090403 ( 2005 ) . e. k. dahl , e. babaev , a. sudbo , phys . rev . * b78 * , 144510 ( 2008 ) . g. parisi , in _ statistical field theory , frontiers in physics _ , addison - wesley , reading , ma , 1988 . prokofev , b.v . svistunov , and i.s . tupitsyn , phys . a * 238 * , 253 ( 1998 ) ; jetp * 87 * , 310 ( 1998 ) ; n.v . prokofev and b.v . svistunov , phys . rev . lett . * 87 * , 160601(2001 ) . j. villain , j. phys . ( paris ) * 36 * , 581(1975 ) . m. wallin , e.s . sorensen , s.m . girvin , and a. p. young , phys . rev . * b 49 * , 12115 ( 1994 ) . w. janke and h. kleinert , nucl . phys . * b270 * , 135(1986 ) . e. i. pollock and d. m. ceperley , phys . rev . * b 36 * , 8343 ( 1987 ) . a.m. polyakov , _ gauge fields and strings _ , in contemporary concepts in physics , vol.3 , harwood academic publishers , chur - melbourne , ch.4 ( 1987 ) . c.l.kane , m.p.a . fisher , phys . . lett . * 68*,1220 ( 1992 ) .

in the last few years , the universality of the stellar initial mass function ( imf ) has been questioned by the results of theoretical studies , which consistently predict that the first stars ( hereafter pop iii stars ) had characteristic masses of 100 - 600 @xmath3 ( omukai & nishi 1998 ; abel , bryan & norman 2002 ; bromm , coppi & larson 2002 ; omukai & palla 2003 ) , about 100 times more massive than those observed today ( larson 2003 ) . indeed , the physical conditions in primordial star - forming regions appear to systematically favor the formation of very massive stars . in particular , ( i ) the fragmentation scale of metal - free clouds is typically @xmath4 ( abel , bryan & norman 2002 ; bromm , coppi & larson 2002 ) ( ii ) because of the absence of dust grains the radiative feedback from the forming star is not strong enough to halt further gas accretion ( omukai & palla 2003 ) . ( iii ) since the accretion rate is as large as @xmath5-@xmath6 , the star grows up to @xmath7 within its lifetime ( stahler , palla & salpeter 1986 ; omukai & nishi 1998 ; ripamonti et al . 2002 ) . in spite of the significant progresses made by numerical simulations and semi - analytic models , many aspects of primordial star formation remain to be fully understood , the more important of which are the radiative feedback effects which operate during protostellar accretion and which are likely to ultimately set the final stellar mass . by modeling the structure of the accretion flow and the evolution of the protostar , tan & mckee ( 2004 ) have recently shown that radiative feedback becomes dynamically significant at protostellar masses @xmath8 , and are likely to constrain the mass of first stars in the range @xmath9 . on the other hand , observations of present - day stellar populations show that stars form according to a salpeter imf with a characteristic mass of 1 @xmath3 , below which the imf flattens . thus , unless the current picture of primordial star formation is lacking in some fundamental physical process , a transition in the modes of star formation must have occurred during cosmic evolution . the physical processes responsible for this transition , as well as its cosmological consequences , are currently the subject of systematic investigation . in particular , following the early study by yoshii & sabano ( 1980 ) , it has been shown that a key element driving this transition is the metallicity of the star - forming gas ( omukai 2000 ; bromm et al . 2001 ; schneider et al . 2002 , 2003 ) . the fragmentation properties of the collapsing clouds change as the mean metallicity of the gas increases above a critical threshold , @xmath10 ( schneider et al . 2002 , 2003 ) . the fragmentation of clouds with @xmath11 proceeds only down to relatively massive ( @xmath12 ) cores , whereas in clouds with @xmath13 lower mass fragments can be formed . within the critical metallicity range , low - mass cores can form if a sufficient amount of metals are depleted onto solid dust grains , which provide an efficient cooling channel in the high density regime ( schneider et al . 2003 ; omukai et al . 2005 ) . on cosmic scales , the transition between an epoch dominated by the formation of massive pop iii stars to an epoch dominated by the formation of ordinary pop ii / i stars is controlled by the strength of the so - called _ chemical feedback _ : the explosions of the first massive supernovae starts to metal - enrich the gas out of which subsequent stellar generations form . in particular , in regions of the universe where the metallicity exceeds @xmath14 , low - mass stars can form . chemical feedback self - propagates through the metal enrichment process and therefore depends on many poorly constrained parameters , such as the star formation efficiency at high redshifts , the number of supernovae per stellar mass formed , the efficiency of metal ejection and mixing . in particular , the shape of the primordial imf governs the number of pop iii stars with masses in the pair - instability supernova range ( pisn , heger & woosley 2002 ) 140 @xmath3 - 260 @xmath3 ( which contribute to the metal enrichment ) relative to that of pop iii stars which end up as massive black holes ( which instead do not contribute to the metal enrichment as all the heavy elements are captured by the black hole ) . using a semi - analytic approach , scannapieco , schneider & ferrara ( 2003 ) have modeled the probability that a collapsing halo has been impacted by metal - enriched outflows from neighboring halos . exploring different chemical feedback strengths , they concluded that the transition between the pop iii and pop ii / i dominated star formation epochs is essentially controlled by the spatial distribution of metals . because metal enrichment is generally inefficient ( the filling factor of metals is always found to be @xmath15 ) , the transition epoch is extended in time , with coeval pop iii and pop ii / i star formation episodes , and pop iii stars still contributing to the global star formation rate at redshifts @xmath16 . these findings have very important implications for the detection of pop iii stars through their strong ly@xmath17 emission ( scannapieco et al . 2003 ) as well as the corresponding pair - instability supernova rate ( scannapieco et al . 2005 ) . in this work , we investigate the cosmic transition between pop iii and pop ii / i stars following a complementary approach . we use the code pinocchio developed by monaco , theuns & taffoni ( 2002 ) and taffoni , monaco & theuns ( 2002 ) to produce random catalogues of dark matter halos at different redshifts . we then propagate metal enrichment along the hierarchy of mergers of star - forming galaxies to predict the number and mass distribution of metal - free objects , which can host pop iii stars , at each redshift . we then couple these findings to the star formation history ( of pop iii and pop ii / i stars ) and to the history of cosmic reionization and quantify their interplay through radiative feedback effects which can contribute to the extinction of pop iii stars . a similar approach has been recently undertaken by furlanetto & loeb ( 2005 ) with the aim of assessing the plausibility of the double reionization scenario suggested by cen ( 2003 ) as a way to meet both the constraints on the gas neutral fraction implied by observations of gunn - peterson troughs in the spectra of distant quasars ( fan et al . 2002 , 2003 ) and on the gas ionized fraction derived by the high optical depth to thomson scattering measured by the wmap satellite ( spergel et al . 2003 ; kogut et al . 2003 ) . here we instead focus on the extinction epoch of pop iii stars which we believe to be regulated by feedback both of radiative and chemical types . we investigate the implications that complementary observations on galaxy number counts at redshifts @xmath1 ( bouwens et al . 2005 ) and on the history of cosmic reionization ( kogut et al . 2003 ) set on the transition epoch between pop iii and pop ii / i stars and on the nature of pop iii stars as sources of uv photons contributing to cosmic reionization . in particular , the detection of galaxies at the highest redshifts is challenging because of the redshifting of uv light into the infrared and the well - known limitations of current ir instruments . in spite of these difficulties , the search for high redshift galaxies , based on the dropout technique , has provided the first tentative detection of galaxies at redshifts beyond @xmath18 ( bouwens et al . 2004 ) , stimulating a re - analysis of the small amount of deep ir data . bouwens et al . ( 2005 ) looked at the prevalence of galaxies at @xmath19 by applying the dropout technique to the wide variety of deep fields that have been imaged with the near infrared camera and multi - object spectrometer ( nicmos ) on board of the hubble space telescope . using an appropriate selection criterion for high-@xmath20 sources , bouwens et al . ( 2005 ) concluded that the number of @xmath21 sources must be three or fewer . when combined with the high optical depth to thomson scattering inferred from the _ wilkinson microwave anisotropy probe _ ( wmap ) satellite ( kogut et al 2003 ) , these data already provide a demanding benchmark for theoretical models . the paper is organized as follows . in section 2 we describe the different feedback effects that control the transition from a pop iii to a pop ii dominated star formation epoch . we illustrate the adopted chemical feedback model , predicting the number and mass function of halos which can host pop iii stars at different redshifts . in section 3 , we couple these predictions to a star formation and reionization model . in section 4 we summarize the observations of galaxy counts at redshifts @xmath22 and discuss how these can be used to set important constraints on the model . finally , in section 5 we present the main results and in section 6 we discuss their implications . throughout the paper , we adopt a @xmath23cdm cosmological model with parameters @xmath24 , @xmath25 , @xmath26 , @xmath27 , @xmath28 and @xmath29 , which lie within the experimental errorbars of wmap experiment ( spergel et al . we also use the ab magnitude system , where @xmath30 is the spectral energy density within a given passband in units of erg s@xmath31 @xmath32 hz@xmath31 ] . at least three types of feedback effects can be relevant to the extinction of the first massive stars : ( i ) radiative feedback caused by h@xmath34 photo - dissociating radiation emitted by the first stellar sources , which can inhibit gas cooling and star formation inside small halos ( ii ) chemical feedback due to the release of metals and dust in the first pop iii supernova explosions , which pollute the gas out of which subsequent generations of pop ii stars form ( iii ) radiative feedback due to photo - heating filtering caused by cosmic reionization , the increased temperature of the cosmic gas suppresses the formation of galaxies below the jeans mass . a thorough discussion of each of these processes is beyond the scope of this paper and can be found in the recent extensive review by ciardi & ferrara ( 2005 ) . as a first step , we build up the hierarchy of mergers of dark halos hosting the star formation sites . we use pinocchio code ( monaco , theuns & taffoni 2002 ; taffoni , monaco & theuns 2002 ) to generate random catalogues of dark matter halos in a hierarchical universe . our reference run has @xmath35 particles within a volume of comoving side 6 mpc @xmath36 so that the smallest resolved dark matter halos have total mass @xmath37 ( in the merger trees only halos with at least 10 particles are considered ) . 1 summarizes the details of our reference pinocchio run and shows the redshift evolution of the largest and smallest mass present in the simulation box and of the minimum halo mass that is allowed to form stars . the latter corresponds , at each redshift , to halos with virial temperatures @xmath38k , m_min = 10 ^ 8 ( ) ^-3/2 [ eq : min ] so that stars do not form in halos which rely only on h@xmath34 cooling ( radiative feedback ) . the chosen resolution ensures that we can follow the formation of stars in the lowest mass halos at the highest redshift . due to radiative feedback , the first star - forming halos have a total mass @xmath39 and start to form at @xmath40 . this resolution is appropriate to follow the transition from a pop iii to a pop ii star formation mode around redshift @xmath1 , which is the main scope of this analysis . in fig . 2 we compare the resulting mass function of existing halos at redshifts 30 , 20 , 10 and 5 in the simulation box , to the analytic predictions of press & schechter ( 1974 ) and sheth & tormen ( 1999 ) . the errorbars are poisson errors on the counts in each mass - bin . the agreement is very good down to redshift @xmath16 showing that the chosen resolution ensures completeness in the simulation volume in the redshift range relevant to our analysis . at each redshift , existing halos which are allowed to form stars ( @xmath41 k ) are classified as hosting pop ii ( pop iii ) stars depending on whether the halo itself or any of its progenitors have ( have not ) already experienced an episode of star formation , leading to the formation of sn progenitors which then metal - enrich the gas . note that even a single 200 @xmath3 pisn can enrich the gas within a halo with mass @xmath42 to a metallicity level well above the critical value as , z = 0.02 z _ ( ) ^3/2 where , following heger & woosley ( 2002 ) , we have assumed that the mass of metals released in a pisn explosion is 45% of the initial progenitor mass . this way of taking into account chemical feedback is very conservative as ( i ) we are assuming that all metals synthesized during a star formation episode remain confined ( or are re - accreted ) within the host halo and/or ( ii ) even if a fraction of metals is ejected out of the host halo , the metals that are kept within the potential well are sufficient to pollute the gas within the host system or within any unpolluted halo that the system will later merge with , to a metallicity level larger than @xmath14 . on the other hand , we are neglecting metal pollution from neighboring galaxies through enriched - winds ( scannapieco et al . 2003 ; sigward , ferrara & scannapieco 2005 ) . 3 shows the fraction of halos which can host pop iii stars ( hereafter pop iii halos ) at different redshifts as a function of their mass . the three different curves correspond to different assumptions on the shape of the primordial imf . in particular , we randomly classify pop iii halos as sn - forming or bh - forming depending on whether they host stars with masses in the pisn range or outside this range . sn - forming halos are metal - enriched and pollute their descendants . bh - forming halos are not enriched in metals and therefore do not propagate chemical feedback along their subsequent hierarchy of mergers . the fraction of sn - forming halos , @xmath43 , is therefore a free parameter that controls the strength of chemical feedback . the three curves shown in each panel of fig . 3 correspond ( from bottom to top ) to @xmath44 . each curve represents an average over 50 realizations of the random classification procedure and the errors are poisson errors on the counts in each mass bin . when @xmath45 , all pop iii halos are sn - forming and chemical feedback is strong . the total fraction of pop iii halos decreases with time , being @xmath46 at @xmath47 and @xmath48 at @xmath49 ; at all redshifts , pop iii stars are confined to form in the smallest halo mass bins ( close to the minimum mass to form stars shown as the vertical dotted line in each panel ) . for @xmath50 ( 50% of pop iii halos host sn ) , the number of pop iii halos at each redshift is larger than the previous case ( @xmath51 at @xmath47 and @xmath52 at @xmath53 ) but pop iii halos are still confined to the smallest mass bins ( @xmath54 ) . finally , when @xmath55 the strength of chemical feedback is significantly reduced ( as only 10% of pop iii halos at each redshift host sn ) and therefore the fraction of pop iii halos is dominant at all redshifts , being @xmath56 at @xmath47 and @xmath57 at @xmath49 , not negligible even in halos with the largest mass present in the simulation box at each redshift . hereafter we only consider @xmath55 and @xmath58 as representative of weak and strong chemical feedback models . despite the strength of chemical feedback , pop iii halos continue to form stars even at redshifts as small as @xmath16 , in agreement with the results of scannapieco et al . thus , independently of whether it preferentially propagates through galaxy mergers or through metal - enriched winds , chemical feedback always leads to a smooth transition between a pop iii and a pop ii dominated star formation epoch . note that the above results depend only on the parameter @xmath43 and on the hierarchy of galaxy mergers , and are totally independent of the star formation efficiency or of the value of critical metallicity for the transition in fragmentation scales of collapsing pre - stellar clouds . [ cols="^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^,^",options="header " , ] so far , we have taken into account radiative feedback only through the choice of the minimum halo mass that can host stars . namely , we have assumed that mini - halos with virial temperatures less than @xmath59k do not significantly contribute to the cosmic star formation history . however , radiative feedback does affect larger halos , particularly when the universe is approaching the epoch of cosmic reionization . indeed , the increase in temperature of cosmic gas during reionization will suppress the formation of small galaxies with mass below the jeans mass . to quantify this effect , we follow the method proposed by gnedin ( 2000 ) and adopt the filtering mass as the characteristic scale below which the gas fraction in star forming objects , hence the fraction of stellar mass , is reduced with respect to the universal value . in particular , before reionization , the average baryon fraction ( the ratio of the mass in baryons over the total halo mass ) has the universal value @xmath60 while after reionization it can be approximated by the fitting formula ( gnedin 2000 ) , f_b = , [ eq : fil ] where the characteristic mass , @xmath61 , is the total mass of objects which , on average , retain 50% of their universal baryon mass , @xmath62 . this characteristic mass is very well approximated by the linear - theory filtering mass , @xmath63 , with m_f^2/3 = _ 0^a da^m_j^2/3(t , a^)[1-()^1/2 ] , where the jeans mass is given by m_j(t ) = 2.56 10 ^ 5 ( _ m h^2)^-1/2(t/)^3/2 ( 1+z)^-3/2 , with @xmath64 and @xmath65 the temperature and molecular weight of the cosmic gas . here we assume the temperature of ionized regions to be @xmath66 k and we modulate the increase in temperature of the cosmic gas due to reionization with the volume filling factor of ionized regions at the corresponding redshift @xmath67 ( see section [ sec : reio ] ) . thus , the characteristic mass depends on the thermal history of the gas and it accounts for the finite time required for pressure to influence the gas distribution . the redshift evolution of the filtering mass is shown in fig . 1 for a model where reionization occurs at redshift @xmath68 ( short - dashed line ) and @xmath69 ( long - dashed line ) . the filtering mass steeply increases with time and , as expected , its value in the redshift range @xmath70 depends strongly on the reionization history . at redshift @xmath1 , the filtering mass is @xmath71 for the late reionization case , and it is a factor ten larger for the early reionization case . as can be inferred from fig . 1 and 3 , following reionization , photoionization feedback can significantly suppress the formation of pop iii stars since halos with masses @xmath72 have evaporated at least 50% of their mass . following barkana & loeb ( 2001 ) , we compute the redshift evolution of the filling factor of hii regions as , q_hii(z ) = _ z^ dz^ ^f(z^,z ) , [ eq : reio ] where @xmath73 and @xmath74 are the present - day number densities of hydrogen and baryons ( @xmath75 is the hydrogen mass fraction ) , and @xmath76 is the production rate of ionizing photons that we can express as , = f_esc n _ _ , where @xmath77 is the escape fraction , the fraction of ionizing photons that is able to escape the host galaxy , @xmath78 is the time - averaged number of ionizing photons emitted per unit stellar mass formed , and @xmath79 is the comoving star formation density , _ = f_f_b _ m_min^ dm m ( m , z ) where @xmath80 is the star formation efficiency , @xmath81 is the baryon fraction given by eq . ( [ eq : fil ] ) , @xmath42 is the minimum halo mass to form stars ( see eq . [ eq : min ] ) and @xmath82 is the number density of halos ( pop ii and pop iii ) at redshift @xmath20 . the parameter @xmath77 also affects the relative strength of the stellar continuum and the ly@xmath17 line and nebular luminosity of each galaxy . the latter contributions are particularly important for massive pop iii stars . finally , the function @xmath83 takes into account the effect of recombinations . assuming a time - independent volume - averaged clumping factor @xmath84 , common to all hii regions , we can write , f(z^,z ) = - c [ f(z^)-f(z ) ] , and f(z)= , where @xmath85 is the hydrogen recombination rate . for a given reionization history , it is straightforward to compute the optical depth to electron scattering as , _ e(z)=_0^z dz^c _ t n_e(z^ ) , where @xmath86 is the thomson cross section and @xmath87 is the number density of free electrons at redshift @xmath20 . in this paper we only consider the evolution of the filling factor of hii regions ( @xmath88 ) . however , for the purpose of computing the optical depth to electron scattering , we assume that @xmath89 and we neglect the contribution of electrons coming from heiii regions ( see also haiman & holder 2003 ) . under these assumptions we can write @xmath90 where @xmath67 is given by eq.(6 ) . bouwens et al . ( 2005 ) looked at the prevalence of galaxies at @xmath19 by applying the dropout technique to the wide variety of deep f110w- and f160w - band ( hereafter @xmath91 and @xmath92 , respectively ) fields that have been imaged with the near infrared camera and multi - object spectrometer ( nicmos ) . the principal data set is constituted by two 1.3 arcmin@xmath93 deep nicmos parallels taken with the advanced camera for surveys ( acs ) hubble ultra deep field ( udf ; each has about @xmath94 orbits data ) . complementary shallower fields possessing similar @xmath95 imaging have been analyzed as well . the 5@xmath96 limiting magnitude for the udfs is @xmath97 in @xmath91 and @xmath98 in @xmath92 ( @xmath99 aperture ) . since these limiting magnitudes are reached only in the deepest regions of the two nicmos udfs , we adopt the conservative limiting magnitude @xmath100 ( bouwens , private communication ) . the primary selection criterion for high-@xmath20 sources is @xmath101 . using this criterion , bouwens et al . ( 2005 ) found eleven sources . eight of these are ruled out as credible @xmath21 sources , either as a result of detection ( @xmath102 ) blue - ward of @xmath91 or because of their colors red - ward of the break ( @xmath103 ) . the nature of the three remaining sources could not be assessed from the data , but this number appears consistent with the expected contamination from low - redshift interlopers . hence , bouwens et al . ( 2005 ) concluded that the actual number of @xmath21 sources in the nicmos parallel fields must be three or fewer . adopting the same selection criterion as in bouwens et al . ( 2005 ) , we compute the number of @xmath104 galaxies detectable in the nicmos udfs predicted by our models . assuming that the three candidates selected by bouwens et al . ( 2005 ) are indeed @xmath105 galaxies , we can set interesting constraints on the models . in particular , for a given chemical feedback strength @xmath43 , we compute the number density of pop iii and pop ii halos at each redshift . we then assume a universal value for the star formation efficiency @xmath106 and escape fraction @xmath77 of ionizing photons and we characterize each stellar population ( pop iii and pop ii / i stars ) by assuming a specific template emission spectrum . in the following , we consider two models , a and b , whose details are given in table 1 . we assume two representative chemical feedback strengths ( @xmath43 = 0.1 for model a and @xmath43 = 1 for model b ) . in both cases , the emission properties of pop ii stars have been computed using a simple stellar population model taken from the galaxev library ( bruzual & charlot 2003 ) . pop ii stars have an initial metallicity of @xmath107 and are assumed to form according to a salpeter imf ( with masses in the range 0.1 @xmath3 - 100 @xmath3 ) . pop iii stars are assumed to be very massive ( with masses larger than 100 @xmath108 ) and their emission properties are computed using the stellar spectra of schaerer ( 2002 ) , including the nebular continuum emission found to be very important for stars with strong ionizing fluxes ( bromm , kudritzki & loeb 2001 ; schaerer 2002 ) , we are not forced to make any assumption about the shape of the imf of pop iii stars as long as these are assumed to be very massive . ] . for each model , we then compute the number of sources per unit solid angle with observed flux in the range @xmath109 and @xmath110 as @xmath111 where @xmath112 is the comoving volume element per unit solid angle and redshift , @xmath113 is the comoving number of objects at redshift @xmath20 with observed flux in the range @xmath114 $ ] , given by @xmath115 here @xmath116 is the halo formation rate , and the factor @xmath117 converts the number density per unit mass in unit flux interval . the average flux @xmath109 from a halo of mass @xmath118 at redshift @xmath20 that has formed at @xmath119 is , @xmath120 where @xmath121 is the template luminosity per unit solar mass for a population of age @xmath122 ( the time elapsed between the redshift @xmath123 and @xmath20 ) , @xmath124 is the luminosity distance , @xmath125 is the instrumental bandwidth , @xmath126 and @xmath127 are the restframe frequencies corresponding to the observed ones . finally , @xmath128 is the intergalactic medium ( igm ) effective optical depth at @xmath129 between redshift @xmath130 and @xmath20 ( see sect . 2.2 of salvaterra & ferrara 2003 ) we compute the expected number of @xmath1 galaxies detectable in the nicmos udfs using eqs . [ eq : counts1]-[eq : counts3 ] and applying the same selection criterion of bouwens et al . ( 2005 ) ( @xmath101 , integrated up to the magnitude limit of the survey , @xmath100 ) . assuming that only the three most luminous sources should be above the detection threshold , we derive the maximum allowed value for the star formation efficiency , @xmath80 , corresponding to a given escape fraction of ionizing photons . the results obtained for models a and b are shown in fig . 4 in terms of comoving star formation density in pop iii and pop ii stars as a function of redshift . model c will be discussed in section [ sec : reio ] . in each panel , vertically ( horizontally ) shaded regions represent the range of star formation histories for pop iii ( pop ii ) stars allowed by the nicmos - udfs counts when a value @xmath131 in assumed ( see table 1 for the corresponding values of @xmath80 ) . for both models and all the explored values of @xmath77 , the three most luminous objects in nicmos udfs are always represented by galaxies hosting pop iii stars , as their strong ly@xmath17 and nebular emission ( schaerer 2002 ) is dominant . as a consequence , @xmath80 is restricted to relatively small values , which increase with increasing @xmath77 . model b , being characterized by the strongest chemical feedback , and therefore by the smallest fraction of halos hosting pop iii stars at redshift @xmath132 ( see also fig . 3 ) , has larger allowed star formation efficiencies than model a. fig . 4 shows that in model a chemical feedback is very inefficient ( @xmath133 ) and pop iii stars dominate the cosmic star formation history down to redshift @xmath134 . conversely , model b shows that when chemical feedback is maximally efficient , the star formation density in pop iii stars is sub - dominant at all but the highest redshifts , but pop iii stars still make a significant contribution to the cosmic star formation history down to redshift 5 . because of the moderate star formation efficiencies implied by the observed nicmos udfs counts , radiative feedback is negligible for the most massive galaxies . to evaluate its impact on the predicted star formation densities shown in fig . 4 , we have assumed a clumping factor @xmath135 . for @xmath136 , the corresponding reionization redshifts vary in the range @xmath137 for model a , and in the range @xmath138 for model b. in model a , radiative feedback effect on the overall pop iii star formation density manifests itself as a progressive decrement of the upper limit ( which corresponds to the largest @xmath77 and @xmath80 ) of the shaded region for @xmath139 . in model b , where the star formation efficiencies implied by the observed counts are larger and , at each redshift , pop iii stars are confined to form only in the smallest halos , radiative feedback is more significant ; around redshift 10 , the lines corresponding to the maximum and minimum pop iii star formation efficiencies cross each other ( the highest the star formation efficiency and escape fraction , the strongest the radiative feedback ) . overall , the observed galaxy counts at redshift @xmath1 favor a cosmic star formation history where the transition from pop iii to pop ii stars is governed by strong chemical feedback , with all pop iii halos efficiently propagating metal enrichment to their descendants along the hierarchy of galaxy mergers . in fact , in less efficient chemical feedback scenarios ( @xmath140 ) , the predicted cosmic star formation density is dominated by pop iii stars down to redshift @xmath53 . stronger constraints on models a and b come from their corresponding reionization histories . figs . 5 and 6 show isocontours of redshifts of reionization as a function of the clumping factor and the escape fraction ( the star formation efficiency is set by the observed counts at @xmath1 ) . we define the redshift of reionization , @xmath141 , as the redshift when the filling factor of ionized regions , given by eq . ( [ eq : reio ] ) is @xmath142 . the shaded regions indicate the parameter space which leads to an optical depth to electron scattering within the range @xmath143 observed by the wmap satellite ( spergel et al . 2003 ; kogut et al . 2003 ) , with the dashed line corresponding to the central value . as it can be inferred from the figures , both models lead to reionization histories which are in poor agreement with observations . indeed , large escape fractions and moderate clumping factors are required to reproduce the observed range of optical depths . independent observations of the escape fraction give , with few exceptions , values in the range @xmath144 ( ciardi & ferrara 2005 and references therein ) whereas numerical simulations ( gnedin & ostriker 1997 ; iliev , scannapieco & shapiro 2005 ) and semi - analytic models ( madau , haardt & rees 1999 ) predict clumping factors in the range @xmath145 . 5 shows that in model a , even assuming an homogeneous igm ( @xmath146 ) , the escape fraction has to be @xmath147 to be marginally consistent with the observations . in model b , the agreement improves but still requires extreme assumptions on either the escape fraction or the clumping factor : when @xmath144 , the predicted optical depth lies within -1@xmath96 error from the observed value only if @xmath148 . our analysis suggests that it is very hard to reconcile the observed number counts at redshift @xmath1 with the reionization history inferred by wmap observations if pop iii stars are assumed to be all very massive , with masses @xmath149 . this result is somehow counter - intuitive because metal - free very massive stars have always believed to play an important role in an early epoch of cosmic reionization ( cen 2003 ; haiman & holder 2003 ; sokasian et al . 2004 , 2004 ; venkatesan , tumlinson & shull 2003 ; wyithe & loeb 2003 ; but see also ciardi , ferrara & white 2003 , and ricotti & ostriker 2004 for alternative interpretations ) . still , due to their strong ly@xmath17 and nebular luminosity , pop iii stars with masses @xmath150 always dominate the predicted galaxy number counts at @xmath1 , even assuming the strongest chemical feedback model . the paucity of candidate objects observed in nicmos udfs implies that star formation in halos with @xmath151 k must have occurred very inefficiently . even if all three objects identified by bouwens et al . ( 2005 ) are truly @xmath1 sources , the contribution of pop iii stars to cosmic reionization is strongly constrained . the constraints are particularly severe when chemical feedback is less efficient ( model a ) , because pop iii stars can form at redshift @xmath1 in relatively large , and therefore luminous , galaxies . even if chemical feedback is assumed to be maximally efficient and pop iii stars are confined to form in the smallest halos , their contribution to cosmic reionization is in agreement with observations only for moderate ( @xmath152 ) clumping factors . unless an additional type of feedback comes into play , the transition between pop iii to pop ii halos , regulated by chemical and radiative feedback , is very smooth . thus , to relax the severe upper limits placed on the star formation efficiencies by nicmos udfs counts , the emission properties of pop iii stars have to be modified . in particular , if pop iii stars are assumed to form according to a salpeter imf with masses in the range @xmath153 , their ly@xmath17 and nebular emission is strongly suppressed . in this limit , ( see model c in table 1 ) , all pop iii halos contribute to chemical feedback , @xmath154 , because each burst of star formation leads to the formation of sn progenitors and nicmos udfs counts at @xmath1 are dominated by pop ii galaxies . as a consequence , the upper limit on the star formation efficiency set by the observed counts is @xmath155 , much higher than those found for models a and b and independent of the escape fraction . the predicted evolution of the star formation density in pop ii and pop iii stars is shown in the lowest panel of fig . similar to model b , pop ii star dominate the star formation history at all but the highest redshifts . indeed , in our approach the transition between pop iii and pop ii dominated star formation epochs is controlled only by the strength of chemical feedback ( @xmath156 ) and it is independent of the assumed star formation efficiencies ( see section [ sec : chem ] ) . however , contrary to model b , the star formation density depends on the escape fraction only indirectly , through the strength of radiative feedback . for this reason , the shaded regions representing the range of allowed star formation densities are not degenerate only after reionization which , for @xmath157 and @xmath158 occurs at redshift @xmath159 ( for pop ii stars , which are preferentially hosted in large halos , this effect is negligible ) . the implications for the history of reionization are summarized in fig . 7 , which shows that model c is in much better agreement with observations than model a and b. the optical depth observed by wmap can be reproduced with @xmath144 even if the clumping factor is as high as @xmath160 . thus , when the observed @xmath1 counts are used to set an upper limit to the allowed star formation efficiency , pop iii stars forming with a conventional imf are more efficient sources of cosmic reionization than very massive stars . note that this conclusion is independent on issues related to chemical and radiative feedback . indeed , both models b and c assume maximally efficient chemical feedback and radiative feedback plays a minor role for model c because the predicted @xmath1 counts are dominated by massive pop ii galaxies , which are largely unaffected by photoevaporation . fig.7 shows that the best - fit value of optical depth implied by wmap observations , @xmath161 , can only be reproduced by model c with @xmath144 if @xmath162 . if larger values of @xmath84 are to be favored or the observed optical depth will be corrected up - ward in future wmap data releases , a revision of model c is to be considered . however , unless the nicmos udfs counts have under - estimated by a large factor the number of @xmath1 sources , our results suggest that the only plausible revision of model c is to modify the primordial imf by increasing the characteristic mass but allowing pop iii stars to form only in the conventional mass range , @xmath163 . in fig . 8 we show the predicted optical depth as a function of the pop iii characteristic stellar mass , @xmath164 . we assume that the characteristic stellar mass enters in the definition of the stellar imf as suggested by larson ( 1998 ) , ( 1 + ) ^-1.35 , which , for @xmath165 leads to the standard salpeter imf . for a given @xmath164 , we compute @xmath166 considering the proper emission rate of ionizing photons ( taken from table 1 of ciardi & ferrara 2005 ) , assuming @xmath167 , three different clumping factors @xmath168 and 30 , and that either all 3 or none of the candidates identified by bouwens et al . ( 2005 ) in nicmos udfs represent @xmath1 galaxies ( upper and lower limits to each shaded region ) . 8 shows that the predicted optical depth depends more on the clumping factor than on the assumed value of @xmath164 . if @xmath135 , the predicted @xmath166 is in agreement with the observed values even assuming @xmath165 , consistent with the results shown in fig . 7 . if @xmath169 , wmap and nicmos udfs data can be reconciled only if all three candidates are confirmed to be @xmath1 galaxies and if the characteristic pop iii stellar mass is @xmath170 . finally , if the observed @xmath166 will be corrected up - ward in future analyses of wmap data , none of the models considered , independently of the assumed @xmath164 , will be able to reconcile the observed reionization history with the counts at @xmath171 , unless @xmath172 . the tentative detections of @xmath1 galaxies in nicmos udfs by bouwens et al . ( 2005 ) provides tight constraints on the relative contributions of pop iii and pop ii stars to the galaxy luminosity function . when coupled to the reionization history of the universe implied by the wmap satellite , this collection of data sets represents already a very demanding benchmark for galaxy evolution models . stimulated by these data , we have followed the assembling history of high - redshift galaxies hosting the first stars in a hierarchical model for structure formation . adopting simple prescriptions for chemical and radiative feedbacks along with available data , we have constrained the evolution of the imf of the first stars . in particular we find that : 1 . two types of feedbacks need to be considered : ( ia ) radiative feedback , which suppresses star formation in h@xmath34-cooling halos and the formation of low - mass galaxies because of photoionization heating following reionization , and ( ib ) chemical feedback due to the progressive enrichment of ( pop ii ) star forming gas by the metals released by the first ( pop iii ) stellar explosions . 2 . if chemical feedback is weak ( @xmath133 ) pop iii stars can form in relatively large halos and dominate the evolution of the cosmic star formation density down to redshift @xmath16 . more realistic scenarios are found assuming @xmath154 , where pop iii stars are confined to form only in the smallest halos at each redshift . 3 . the combined effect of radiative and chemical feedbacks fails to suppress the formation of massive pop iii stars at @xmath139 . if pop iii stars are assumed to form with a top - heavy imf , that is with masses @xmath173 , pop iii galaxy are always found to dominate the predicted @xmath1 nicmos udfs counts . because of their strong ly@xmath17 and nebular luminosities , very tight upper limits on the star formation efficiency in @xmath174 k halos are placed by the paucity of observed counts . the corresponding reionization histories fail to reproduce the observed wmap optical depths unless uncomfortably large values for the escape fraction of ionizing photons and/or too small values for the igm clumping factor are assumed . reionization constraints from wmap and observed @xmath1 counts in nicmos udfs can be reconciled if pop iii stars form according to a larson imf with stellar masses in the conventional range @xmath175 and a characteristic mass @xmath176 if the clumping factor is @xmath177 , or @xmath178 if the clumping factor is as large as 30 . in the latter case , at least 1 candidate identified by bouwens et al . ( 2005 ) has to be a galaxy at @xmath1 . our analysis is based on the assumption that chemical feedback propagates along the hierarchy of galaxy mergers from progenitors to descendants rather than through metal - enriched galaxy outflows ( scannapieco et al . 2003 ) . by adopting this simplified scheme , we are implicitly assuming that ( i ) even if a fraction of metals is ejected out of the host halo , the metals that are kept within the potential well are sufficient to pollute the gas within the host system or within any unpolluted halo that the system will later merge with , to a metallicity level larger than @xmath14 and that ( ii ) metal pollution from neighboring galaxies is negligible . in spite of these simplifying assumptions , the resulting pop iii and pop ii star formation histories are consistent with the predictions of scannapieco et al . in particular , even assuming a maximally efficient chemical feedback , the transition is always predicted to be very smooth , with pop iii stars which continue to form in the smallest halos at each redshift down to @xmath16 . interestingly enough , our conclusion is independent of the primordial imf and/or star formation efficiency . we show that models where all pop iii halos are able to metal pollute their descendants ( @xmath154 ) are favored by observations . the smoothness of the cosmic imf transition is therefore strictly linked to the inhomogeneous character of the metal enrichment process , independently of how it is implemented in the models . even if we have adopted a very approximate description of the thermal evolution of the cosmic gas , we find that radiative feedback following reionization can play an important role in the extinction of pop iii stars . in particular , if reionization occurs at redshifts @xmath179 ( @xmath1 ) the baryon fraction in @xmath180 halos is reduced to 70% ( 99% ) of its universal value ( @xmath60 ) at redshift 10 and to 10% ( 30% ) of this value at redshift 5 . thus , the baryon content in the smallest halos where pop iii stars can form at low redshift is severely limited by radiative feedback . this is strongly related to the epoch of cosmic reionization which , in turn , depends on the degree of inhomogeneities which characterize the cosmic gas . we have parametrized this quantity through a constant clumping factor @xmath84 and we have explored different values of @xmath84 ranging from 1 to 50 . however , it is likely that the degree of clumpiness of the cosmic gas varies with time , rising from @xmath181 at high redshift to a few tens at @xmath1 as structures go increasingly non - linear . recently , iliev et al . ( 2005 ) have studied the effect of igm clumping at redshifts @xmath182 , providing a fit to the evolution of the clumping factor ( see their fig . 7 ) where @xmath183 at redshift 30 and raises to @xmath184 at redshift 10 . adopting this evolution for the clumping factor and @xmath185 we find that models a and b are unable to reproduce the optical depths inferred from wmap data whereas in model c we find that reionization is complete at redshift 13.5 and that the resulting optical depth is @xmath186 in very good agreement with the observed value . thus , our analysis suggests that pop iii stars forming with masses in the same dynamical range as pop ii / i stars , but possibly with higher characteristic stellar mass , appear to be favored by observations . several scenarios that enable the formation of low - mass stars in metal - free environments have been proposed , among which the bi - modal imf ( nakamura & umemura 1999 , 2002 ) , enhanced hd - cooling in relic hii regions ( uehara & inutsuka 2000 ; nakamura & umemura 2002 ) , the influence of a strong uv radiation field due to the vicinity of a very massive star ( omukai & yoshii 2003 ) , fragmentation in shock - compressed shells induced by the first very massive sn explosions ( mackey , bromm & hernquist 2003 ; salvaterra , ferrara & schneider 2004 ) . however , all these processes either require very special conditions , generally related to the existence of a previous generation of very massive stars , or have failed to be realized in realistic numerical simulations . in spite of this , observations seem to indicate that either radiative feedback effects during the accretion phase on protostellar cores and/or a combination of the above mentioned processes can lead to the formation of pop iii stars with masses @xmath187 . according to the results of our analysis , the latter contribution to the pop iii stellar imf must be dominant at least at redshifts @xmath1 . if this scenario is correct , relics of the low - mass pop iii star formation mode should be still present in our galaxy , eventually contributing to the lowermost end of the metallicity distribution function ( salvadori , schneider & ferrara in prep ) . we acknowledge the use of pinocchio v 1.0 package written by pierluigi monaco , tom theuns & giuliano taffoni ( http://www.daut.univ.trieste.it/pinocchio ) . this work has been partially supported by the research and training network `` the physics of the intergalactic medium '' set up by the european community under the contract hprn - ct-2000 - 00126 . 99 abel t. , bryan g. l. , norman m. l. 2002 , science , 295 , 93 barkana , r. & loeb , a. 2001 , phr , 349 , 125 bouwens r. j. et al . 2004 , , apj , 616 , l79 bouwens r. j. , illingworth g. d. , thomson r. i. , & franx m , 2005 , apj , 624 , l5 bromm , v. , ferrara a. , coppi p. s. , larson r. b. 2001 , mnras , 328 , 969 bromm , v. , kudritzki , r. p. & loeb , a. 2001 , apj , 552 , 464 bromm , v. , coppi p. s. , larson r. b. 2002 , apj , 564 , 23 bruzual , g. & charlot , s. 2003 , 344 , 1000 ciardi , b. & ferrara , a. 2005 , space science reviews , 116 , 625 ciardi , b. , ferrara , a. , white , s. d. m. 2003 , mnras , 344 , l7 cen , r. 2003 , apj , 591 , 12 furlanetto , s. & loeb , a. 2005 , apj in press , ( astro - ph/0409736 ) gnedin , n. , ostriker , j. 1997 , apj , 474 , 581 gnedin , n. 2000 , apj , 542 , 535 haiman , z. & holder , g. p. 2003 , apj , 595 , 1 heger , a. & woosley , s. 2002 , apj , 567 , 532 iliev , i. , scannapieco , e. , shapiro , p. 2005 , apj , 624 , 491 kogut , a. et al . 2003 , apjs , 148 , 161 larson , r. b. 1998 , mnras , 301 , 569 mackey , j. , bromm , v. , hernquist , l. 2003 , apj , 586 , 1 madau , p. , haardt , f. , rees , m. j. 1999 , apj , 514 , 648 monaco , p. , theuns , t. & taffoni , g. 2002 , mnras , 331 , 587 nakamura , f. , umemura , m. 1999 , apj , 515 , 239 nakamura , f. , umemura , m. 2002 , apj , 569 , 549 omukai , k. & nishi , r. 1998 , apj , 508 , 141 omukai , k. 2000 , apj , 534 , 809 omukai , k. & palla , f. 2003 , apj , 589 , 677 omukai , k. & yoshii , y. 2003 , apj , 599 , 746 omukai , k. , tsuribe , t. , schneider , r. , ferrara , a. 2005 , apj , 626 , 627 ricotti , m. & ostriker , j. p. , 2004 , mnras , 350 , 539 ripamonti , e. , haardt , f. , ferrara , a. , & colpi , m. , 2002 , mnras , 334 , 401 salvaterra , r. , ferrara a. , schneider r. 2004 , new astronomy , 10 , 11 salvaterra , r. & ferrara a. 2003 , mnras , 339 , 973 scannapieco , e. , schneider , r. & ferrara , a. 2003 , apj , 589 , 1 scannapieco , e. , madau , p. , woosley , s. , heger , a. , ferrara , a. 2005 , apj , in press ( astro - ph/0507182 ) schaerer , d. 2002 , a&a , 382 , 28 schneider r. , ferrara a. , natarajan p. , omukai k. 2002 , apj , 571 , 30 schneider r. , ferrara a. , salvaterra r. , omukai k. , bromm v. 2003 , nature , 422 , 869 sigward , f. , ferrara , a. & scannapieco , e. 2005 , mnras , 358 , 755 sokasian , a. , yoshida , n. , abel , t. , hernquist , l. , springel , v. 2004 , mnras 350 , 47 spergel , d. l. et al . 2003 , apjs , 148 , 175 stahler , s. w. , palla , f. , & salpeter , e. e. 1986 , apj , 302 , 590 tan , j. c. & mckee , c. f. 2004 , apj , 603 , 383 taffoni , g. , monaco , p. & theuns , t. 2002 , mnras , 333 , 623 uehara , h. & inutsuka , s. 2000 , apj , 531 , l91 venkatesan , a. , tumlinson , j. , shull , j. m. 2003 , apj , 584 , 621 wyithe , j. s. b. & loeb , a. 2003 , apj , 588 , l69 yoshii , y. , sabano , y. 1980 , pasj , 32 , 229

numerical modeling of dynamic spacetimes containing black holes ( bhs ) in the framework of einstein s theory has been motivated for a large part of its nearly 50 year long history by the calculation of the gravitational wave ( gw ) signals generated by bh binary systems . these efforts culminated in 2005 in the first evolutions of bh binaries through inspiral , merger and ringdown@xcite . over the following years , numerical relativity ( nr ) simulations have been used in the construction of gw template banks @xcite fundamental for signal searches in gw detectors such as ligo , virgo and geo600 @xcite and have also been studied in the context of gw data analysis @xcite . ninja_@xcite and _ nrar_@xcite collaborations in particular combine the resources of various nr groups with the analytic relativity and gravitational - wave data analysis community . nr studies have furthermore provided valuable insight into a variety of astrophysical scenarios involving bhs as for example the study of gravitational recoil @xcite , jet phenomena @xcite and electromagnetic counterparts for multi - messenger astrophysics @xcite . for reviews of these subjects we refer to refs . . in recent years , however , nr has also emerged as a valuable tool to explore high - energy physics scenarios @xcite . the purpose of this article is to provide an overview of the most recent developments in these areas which , for reasons of clarity , we divide into three categories : ( i ) numerical studies of fundamental properties of bh spacetimes , ( ii ) applications to the modeling of trans - planckian scattering and ( iii ) the gauge gravity duality . this division is not entirely rigorous and we will encounter scenarios of relevance for more than one of these major topics . for example the stability of rotating bhs in higher dimensions is discussed in sec . [ sec : fundamental ] on fundamental properties of bhs but likely also affects the signature in the conjectured formation of mini bhs in trans - planckian scattering events . wherever present , we will emphasize such significance across our somewhat arbitrarily chosen distinction . nr studies in the context of high - energy physics and fundamental properties of classical general relativity frequently involve bhs in higher dimensional spacetimes . in some cases , however , valuable insight can be inferred as well from the more traditional case of spacetimes with `` 3 + 1 '' dimensions ( three spatial and one time - like ) . somewhat in breach of its title , this work is therefore not entirely limited to the numerical investigation of higher - dimensional bhs . as complementary reading to this article , we note review articles including nr applications in high - energy physics by pretorius@xcite , cardoso _ et al._@xcite and yoshino & shibata@xcite . an extended discussion of bhs in higher dimensions is given in . the remainder of this article is organized as follows . we begin with a brief summary of the nr techniques in sec . [ sec : nr ] . the above mentioned main applications of nr to higher - dimensional bhs will be reviewed in secs . [ sec : fundamental ] , [ sec : transplanckian ] and [ sec : adscft ] . in each of these we will provide a brief motivation of the corresponding studies and list more specialized review articles for further reading . we will conclude in sec . [ sec : conclusions ] . unless noted otherwise , we use units such that the speed of light @xmath0 . formulations of the einstein equations suitable for nr can be classified into two main groups . the characteristic bondi - sachs@xcite formalism is based on the characteristic surfaces of the vacuum einstein equations . this formalism leads to a natural hierarchy of the equations which reflects the isolation of the gravitational degrees of freedom ; in the four - dimensional case , for example , one obtains one trivial equation , three complementary equations , four hypersurface equations and two evolution equations for the degrees of freedom . a difficulty arising in the characteristic approach is the potential formation of caustics and the ensuing breakdown of the coordinate system . for the classical general - relativistic two - body problem , the inspiral and merger of two bhs , a robust solution for this problem has not yet been found . nevertheless , characteristic methods have been employed with great success in various codes studying configurations with additional symmetries , most notably in the series of studies by chesler & yaffe@xcite on thermalization of quark - gluon plasmas through the anti - de sitter / conformal field theory ( ads / cft ) correspondence . characteristic evolutions are reviewed in detail by winicour@xcite . the majority of work on the inspiral and coalescence of compact binaries , including the 2005 breakthroughs , however , has been performed in 3 + 1 approaches . here spacetime is decomposed into a one - parameter family of three or , more generally , @xmath1 dimensional , spatial hypersurfaces . the canonical 3 + 1 split has been developed by arnowitt , deser and misner@xcite ( adm ) , reformulated by york@xcite and leads to a system of six second - order evolution equations for the induced spatial metric and four constraint equations , the hamiltonian and momentum constraints . introduction of the extrinsic curvature as an evolution variable converts the evolution equations into a first - order - in - time system . by performing a conformal rescaling of the spatial metric , a decomposition of the extrinsic curvature into trace and a trace - free part and evolving the contracted christoffel symbol as separate fields , baumgarte , shapiro , shibata and nakamura@xcite have developed a strongly hyperbolic formulation@xcite commonly referred to as `` bssn '' which has been used with great success by many numerical groups . the adm / bssn approach to numerical relativity is discussed in great detail in . an alternative cauchy - type formulation has been derived from the einstein equations in harmonic gauge where they take on a manifestly hyperbolic form . the harmonic gauge can be generalized to arbitrary gauge choices by introducing source functions for the coordinate conditions @xcite . more details of this _ generalized harmonic gauge _ ( ghg ) formulation can be found in . the ghg and bssn methods have been the foundation of the numerical relativity breakthroughs of 2005 by pretorius@xcite and the brownsville and goddard groups@xcite . nr applications in higher - dimensional spacetimes have so far been restricted to configurations with symmetry assumptions . this is largely a consequence of the vast increase in computational resources required by each additional dimension . a reduction of higher - dimensional general relativity to an effective 3 + 1 or lower - dimensional computational domain has been achieved in various ways . the most evident approach is to directly impose the symmetry together with the coordinate conditions on the line element . for example , sorkin & oren@xcite describe the @xmath2-dimensional , spherically symmetric spacetime in double - null coordinates @xmath3 by the line element @xmath4 where @xmath5 is the metric on a @xmath6 dimensional unit hyper sphere and @xmath7 is the areal radius . the time evolution is then determined by the einstein equations derived from this metric ; see their eqs . ( 3)-(6 ) . alternatively , a general procedure for a dimensional reduction under the assumption of one or more killing vectors is given by geroch@xcite . by using this procedure , the @xmath2-dimensional einstein equations with @xmath8 symmetries can be converted to an effective 3 + 1 scheme coupled to scalar or vector fields which represent the curvature of the extra - dimensions@xcite . such a dimensional reduction is particularly attractive because it provides a straightforward way to apply the successful numerical techniques developed in 3 + 1 with relatively minor modifications to the study of higher dimensional spacetimes . a different way to achieve the same goal is based on the _ cartoon method_@xcite which employs interpolation of grid functions and/or trading of spatial derivatives to evolve symmetric spacetimes on a reduced cartesian grid . this approach has been modified and extended to higher - dimensional applications by pretorius@xcite and yoshino & shibata@xcite and has proven a rather robust tool to study bh collisions and instabilities . the generation of initial data in higher - dimensional spacetimes relies to a significant extent on similar techniques developed for the 3 + 1 case ; see cooks@xcite review . in particular , it has been shown by yoshino _ et al._@xcite that the existence of analytic solutions of bowen - york@xcite type for the momentum constraints carries over to general relativity in @xmath2 dimensions . solving the hamiltonian constraint for the conformal factor can then be achieved by numerical methods similar to those developed for 3 + 1 dimensions . by using this procedure , initial data for bh binaries with boost have been obtained in refs . . for specific numerical studies , analytic initial data is available in closed form , as for example the @xmath2 dimensional generalization of brill - lindquist@xcite data describing a bh binary at the moment of time symmetry . analytic bh solutions can also be employed to construct approximate solutions to the einstein constraints by superposition of either the solutions themselves or boosted variants thereof . for more details on the numerical methods , the construction of initial data and gauge choices , we refer the reader to the publications reviewed in the following sections . bhs play a central role in the applications of nr to higher - dimensional spacetimes discussed in secs . [ sec : transplanckian ] and [ sec : adscft ] . quite aside from these applications , bhs provide a unique probe for deepening our insight into the gravitational theory and understanding in which regards 3 + 1 dimensions represent a special case . bhs in higher dimensions indeed reveal a much richer phenomenology than their four - dimensional cousins ; cf . emparan & reall@xcite . asymptotically flat , stationary solutions of the einstein equations in vacuum in 3 + 1 dimensions belong to the kerr family and describe single bhs of spherical topology characterized by two parameters , the mass @xmath9 and angular momentum parameter @xmath10 @xcite . the stability of these solutions has been studied extensively ; see for example refs . and dafermos@xcite review . such classical bh solutions are expected to be stable in the region outside the horizon , but admit instabilities in the interior@xcite . penrose s _ cosmic censorship_@xcite furthermore conjectures that collapse of ordinary matter does not lead to the formation of naked singularities ; instead such singularities are hidden from outside view through an event horizon . bhs in higher dimensions are solutions to the @xmath2-dimensional einstein equations ] @xmath11 for the vacuum case @xmath12 . solutions to these equations can be generated straightforwardly from extending the four - dimensional schwarzschild metric with one or more flat dimensions . these _ black strings _ or _ black branes _ are susceptible to the gregory - laflamme@xcite ( gl ) instability related to the vastly different length scales in the horizon geometry . these solutions furthermore suggest a modification of thornes@xcite hoop conjecture ; if an amount @xmath9 of mass - energy is compressed into a volume such that its circumference is below its schwarzschild radius in every direction , a bh forms . the existence of black branes indicates that such a concentration of energy is sufficient in a subset of the spatial directions . the @xmath2-dimensional generalization of the spherically symmetric schwarzschild solution is given by the tangherlini@xcite metric @xmath13 where the constant @xmath14 determines the bh mass @xmath15 . here , @xmath16 denotes the surface area of the @xmath6 dimensional unit hypersphere . note that mass is now of dimension @xmath17 . schwarzschild solutions in @xmath2 dimensions have been found to be stable against linearized gravitational perturbations@xcite . in contrast to their spherically symmetric counterparts , the class of rotating bh solutions exhibits a vastly richer structure in higher dimensions . first , rotation is now possible in more than one plane , giving rise to more than one rotation parameter . second , the radial dependence of the newtonian potential is altered to @xmath18 whereas the centrifugal potential maintains its @xmath19 character . this @xmath2-dependent character of the newtonian potential manifests itself in the absence of stable circular geodesics for @xmath20@xcite . quite remarkably , myers & perry@xcite found an exact solution describing bhs rotating in all possible planes . in @xmath21 and @xmath22 , bhs rotating in one plane with rotation parameter @xmath23 and @xmath24 respectively represent naked singularities . in @xmath25 , in contrast , there exists a horizon for all values of @xmath10 . furthermore , in the limit of infinite rotation rate , this horizon becomes increasingly flattened along the rotation plane@xcite and is therefore expected to be gl unstable . this has been confirmed in linearized studies by dias _ et al._@xcite . in the case of rotations in more than one plane , the existence of horizons in ultra - spinning bh spacetimes requires at least two ( one ) spin parameter to vanish for an odd ( even ) number of dimensions @xmath2 . rotation in higher dimensions also leads to bh solutions of non - spherical topology . _ black rings _ are supported against contraction by the centrifugal potential ; analytic solutions in @xmath22 have been found in thin rings behave locally like boosted black strings@xcite which are gl unstable@xcite while _ fat _ rings are expected to be unstable to variations of their radius@xcite . solutions for _ black saturns _ have been constructed in and demonstrate the existence of stationary vacuum multiple - bhs in higher dimensions . the generalization of the ads - schwarzschild solution@xcite is given by @xmath26 in the limit @xmath27 , this solution reduces to the ads spacetime with curvature radius @xmath28 in _ global _ coordinates ; cf . sec . [ sec : adscft ] . its mass relative to the empty ads solution is given by @xcite @xmath29 . this solution is stable against linearized gravitational perturbations @xcite . the four - dimensional kerr - ads spacetime ( see e. g. eq . ( 2.1 ) in ) has been found by carter@xcite and generalized to higher dimensions in refs . . kerr - ads has been found to be unstable against linear perturbations if the bhs are small@xcite . one of the first and most influential nr results has been obtained in choptuiks@xcite seminal study of the collapse of spherically symmetric massless scalar fields minimally coupled to einstein gravity in four - dimensional , asymptotically flat spacetimes . by evolving various one - parameter families of scalar pulses , he identified critical behaviour as the parameter @xmath30 which characterizes the gravitational interaction strength of the field approaches a critical value @xmath31 . for @xmath32 , the field collapses to a bh and for @xmath33 it disperses to infinity . furthermore , near - critical field configurations exhibit _ universal _ behaviour in the strong field limit : ( i ) bhs which form have a mass @xmath34 with a universal constant @xmath35 . ( ii ) advancing the evolution from a time @xmath36 to @xmath37 , the field profile is recovered up to a `` zoom - in '' by a factor @xmath38 . the numerical study reveals @xmath39 to be a universal constant of about @xmath40 . for subcritical configurations , garfinkle & comer duncan @xcite found a similar scaling of the maximal scalar curvature in the spacetime @xmath41 . continuously self - similar solutions were found by pretorius & choptuik@xcite for scalar fields in 2 + 1 dimensional asymptotically ads spacetimes with a mass scaling characterized by @xmath42 . sorkin & oren@xcite generalized choptuik s result to higher dimensions by evolving scalar fields up to @xmath43 dimensions . their results indicate that @xmath44 reaches a maximum and @xmath39 a minimum around @xmath45 . an extended review of critical collapse phenomena is given by gundlach & martn - garca@xcite . more recently , critical - collapse studies have been extended to asymptotically ads spacetimes in 3 + 1 and higher dimensions . in a remarkable study , bizo & rostworowski found evidence suggesting that the 3 + 1 ads is unstable to bh formation under arbitrarily small perturbations . by evolving spherically symmetric scalar fields they recover choptuik s results for large initial field amplitudes . for configurations below this critical value , however , the ads boundary substantially modifies the outcome . in contrast to asymptotically flat spacetimes , spatial infinity in ads is reached in finite time by massless fields and reflected back onto the origin . as the initial amplitude is reduced below a critical value @xmath46 , the scalar pulse forms a bh upon its second implosion on the origin . further reduction of the amplitude leads to a second critical amplitude @xmath47 and this pattern repeats itself with no indication of a threshold amplitude for bh formation ; cf . their fig . 1 . for each critical amplitude , they furthermore recover choptuik s scaling law with @xmath48 . they interpret this behaviour as a resonant mixing of modes which transfers energy from low to high frequencies . this study has been generalized to higher - dimensional ads spacetimes by jamuna _ et al._@xcite suggesting that ads is unstable to bh formation for generic spacetime dimensions . presumably , this result has eluded a similar study by garfinkle & pando zayas@xcite , because of insufficient length of their numerical simulations for smaller field amplitudes . in a similar investigation using complex scalar fields , et al._@xcite reproduce the instability of ads and the transfer of energy from low to high frequencies . in consequence , the width of initially weak pulses narrows in each reflection cycle and eventually collapses to a bh ; cf . their fig . @xcite have monitored the time of bh formation and the horizon radius and find the amplitude of the scalar field to have a stronger influence on the outcome compared with the width of the pulse . the same type of instability to bh formation has been found by maliborski @xcite for a minkowski spacetime enclosed inside a reflecting wall , indicating that the global structure plays a major role for the effect . perturbative studies support the numerically observed instability of ads @xcite , but the question of the generic stability properties of _ asymptotically _ ads spacetimes , as for example those containing bhs or boson stars , remains under study @xcite . the instability of black strings has been studied numerically in a sequence of papers by lehner & pretorius and collaborators@xcite , see also . it had been known since the 1990s that black strings are subject to the gl instability , but the eventual fate of the string remained unclear . in ref . , lehner & pretorius found evidence supporting indications by earlier work that the string evolves to a sequence of spherical bhs connected by thin string segments which themselves are subject to the gl instability , resulting in a self - similar cascade reaching zero string width in finite asymptotic time . this behaviour shows striking similarity with satellite formation in the flow of low - viscosity fluids . the eventual bifurcation of the horizon resulting from the cascade implies formation of a naked singularity@xcite . because no finetuning is required to trigger the instability , the result constitutes a violation of the cosmic censorship conjecture without `` unnatural '' assumptions about the initial data . in contrast to the higher - dimensional case , nr has as yet not observed any such violation of cosmic censorship in 3 + 1 dimensions . in a recent study@xcite bh collisions in asymptotically de sitter spacetimes have been found to comply with censorship ; bh binaries with a combined mass exceeding the inverse of the hubble constant do not merge for any initial separation provided the initial data do not contain a naked singularity . we have already mentioned the linear instability of rapidly rotating myers - perry bhs . the stability of rotating bhs has been studied numerically by yoshino & shibata@xcite who evolve single holes with a bar - shaped perturbation imposed on the conformal factor in @xmath49 dimensions . in all cases , they observe bhs spinning above a threshold dimensionless spin - parameter @xmath50 to spontaneously emit gravitational waves and settle down into a subcritical configuration . expressed in terms of the ratio of polar to equatorial circumference @xmath51 of the apparent horizon , the onset of instability occurs for nearly identical values for @xmath25 , namely @xmath52 , @xmath53 and @xmath54 , respectively , for @xmath55 , @xmath56 and @xmath57 . in contrast , they obtain a significantly stronger distortion threshold @xmath58 in @xmath22 indicating a somewhat special status of @xmath22 ; we recall in this context the absence of an upper limit on the spin in @xmath25 . bhs spinning below the critical rate , on the other hand , show no signs of instability in any of their simulations . the effect of this instability on the evaporation of bhs conjectured to be produced in trans - planckian scattering ( cf . [ sec : transplanckian ] ) depends on the timescale of the instability and , hence , on the spin parameter of the formed bh ; for details see the discussion in sec . vi d in ref . . even though this review is primarily concerned with time evolutions , we note that numerical techniques are also applied for finding static and stationary solutions to the @xmath2 dimensional einstein equations in vacuum , as for example the construction of four - dimensional rotating bh solutions in a cavity@xcite or asymptotically ads spacetimes in @xmath22 with a conformally schwarzschild boundary @xcite . this topic is reviewed in detail by wiseman@xcite . the standard model of particle physics provides an exceptionally successful description of subatomic particles and their interactions via the electromagnetic , weak and strong forces . in spite of its success , however , there remain important unanswered questions , as for example the unknown nature of dark matter . in the context of this section , the most important open issue is the large discrepancy between the electroweak energy scale @xmath59 and the grand unification scale @xmath60 . this so - called _ hierarchy problem _ manifests itself in the extraordinary weakness of the gravitational interaction relative to the other fundamental forces ; the weak interaction is about 32 orders of magnitude stronger than the gravitational one . at energies comparable to the grand unification energy , on the other hand , gravity is expected to become comparable in strength to the other interactions . a particularly intriguing solution to the hierarchy problem is the _ add _ model @xcite , which proposes large ( relative to the four - dimensional planck scale @xmath61 ) extra dimensions of radius @xmath62 such that the fundamental planck scale @xmath63 may be as low as @xmath64 . consider for illustration the poisson equation in @xmath1 spatial dimensions for a point source @xmath65 , with the @xmath2 dimensional gravitational constant @xmath66 and the source mass @xmath9 . from gauss s law we obtain in spherical coordinates @xmath67 at large distances @xmath68 , the extra dimensions are not accessible to the gravitational field and merely contribute an overall volume factor such that @xmath69 whereas for @xmath70 , the potential is determined by the @xmath1 dimensional poisson equation . ignoring geometrical factors of order @xmath71 , we thus obtain @xmath72 at large distances we thus recover the usual @xmath73 potential with a four - dimensional coupling constant @xmath74 which implies for the energy scales @xmath75 . for example , a fundamental planck mass @xmath63 of the order of the electroweak scale requires @xmath76 in @xmath22 ( ruled out experimentally ) or @xmath77 for @xmath55 . in a different version of the model proposed by randall & sundrum@xcite , a finite length scale is introduced through a warp factor into otherwise infinite extra dimensions . an intriguing consequence of an effective planck mass much below its four - dimensional value @xmath60 is the possibility of bh formation in parton - parton collisions at the large hadron collider ( lhc ) @xcite ; see also the reviews by cavagli @xcite and kanti @xcite . once formed , such mini bhs are expected to evaporate in four stages @xcite : ( i ) a balding phase during which the bh sheds all multipoles except for mass , spin and charge , ( ii ) a spin - down and ( iii ) a schwarzschild phase during which the bh looses first its spin and then its mass via semi - classical hawking radiation and ( iv ) the planck regime as the bh mass approaches the planck mass which is described by an as yet unknown theory of quantum gravity . of particular interest in the context of nr is the fact that the first three of these phases should be well described by classical and semi - classical calculations provided the bh mass exceeds the planck scale by at least a factor of a few @xcite . if we further assume that most of the energy of the collision process resides in the kinetic energy of the particles , such that their internal structure becomes negligible , the dynamics of the collision should be well modeled by two point particles or bhs in @xmath2-dimensional general relativity @xcite . testing this assumption forms one of the main motivations for nr applications and we shall return to this question shortly . bh formation is expected to manifest itself in collision experiments by a special signature in its decay products as for example the jet multiplicity or transverse energy @xcite . for the identification of these signatures , theoretical predictions from monte - carlo event generators such as blackmax @xcite and charybdis @xcite are compared with experimental data . key input parameters for the event generators are the scattering cross section for bh formation and the initial mass and spin distributions of the formed holes . providing this information forms a second main challenge for nr to complement estimates obtained from perturbative calculations on superposed aichelburg - sexl shock - wave backgrounds @xcite . the most recent analysis of data taken from collisions at the lhc with energies up to @xmath78 excludes semi - classical bhs of mass below @xmath79 @xcite and sets exclusion contours in the plane formed by the bh mass and the threshold mass separating the semi - classical from the quantum regime @xcite . it thus appears unlikely that the most optimistic tev - gravity models with @xmath80 are those realized in nature . in summary , the main questions to be addressed by nr simulations are ( i ) the validity of the hoop conjecture in highly dynamical configurations , ( ii ) the scattering threshold for bh formation , ( iii ) the mass and spin parameters of the formed bhs and ( iv ) the impact of the colliding objects structure on the dynamics . high - energy collisions in the framework of general relativity are currently best understood in @xmath21 dimensions , largely because the numerical infrastructure for this setting is most advanced and robust . even though generalization of the results to arbitrary spacetime dimensionality will ultimately be essential , a great deal of insight can be and has already been obtained in the four - dimensional case . we shall review these results before plunging into the latest developments on collisions for arbitrary @xmath2 . the expectation of bh formation in high - energy collisions dominated by gravitational interaction is closely tied to thorne s hoop conjecture @xcite . applied to head - on collisions with boost factor @xmath44 , we expect two objects , each of rest mass @xmath81 and size @xmath82 , to form a bh if the schwarzschild radius @xmath83 associated with @xmath9 satisfies @xmath84 , i. e. @xmath85 . choptuik & pretorius @xcite tested this scenario by numerically simulating head - on collisions of boosted boson stars in @xmath21 dimensions . their initial configurations consist of two boson stars with compactness @xmath86 and the hoop conjecture would imply bh formation for boost factors @xmath87 . their numerical results demonstrate bh formation for @xmath88 , a factor a few below the hoop conjecture and thus strengthening the expectation that kinetic - energy dominated collisions indeed produce bhs . this picture has recently been confirmed in head - on collisions of fluid particles in refs . . the energy radiated in gravitational waves ( gws ) reported by east & pretorius@xcite is @xmath89 for @xmath90 , in good agreement with ultrarelativistic limits @xmath91 and @xmath92 obtained from perturbative calculations @xcite and numerical simulations of bh head - on collisions @xcite respectively . these studies strengthen the conjecture that trans - planckian collisions are well described by high - energy collisions of black holes , general relativity s closest analog of point particles . binary configurations of non - spinning , equal - mass bh binaries are characterized by two parameters , the boost factor @xmath93 and the impact parameter defined as @xmath94 , where @xmath95 , @xmath28 and @xmath96 are the velocity , orbital angular momentum and linear momentum . the head - on case @xmath97 has been studied by sperhake _ @xcite who evolve a sequence of collisions with @xmath98 . the simulations always result in a merger and extrapolation of the radiated energy to @xmath99 leads to the above mentioned @xmath100 of the center - of - mass ( com ) energy , about a factor 2 below penrose s upper limit @xmath101 quoted in . for grazing collisions , the outcome depends on the size of the impact parameter @xmath102 . et al._@xcite identify three regimes : ( i ) for @xmath103 , the bhs scatter off each other and eventually escape to infinity . ( ii ) in a narrow regime @xmath104 the bhs separate after an initial encounter , but have lost enough energy in gravitational radiation such that they form a bound system and eventually merge . as @xmath102 approaches the _ threshold of immediate merger _ @xmath105 , the binary exhibits the _ zoom - whirl _ behaviour identified by pretorius & khurana@xcite with a number of whirls proportional to the logarithmic distance of @xmath102 from @xmath105 . ( iii ) for sufficiently small @xmath106 , the outcome is a prompt merger . by fitting numerical results obtained for binaries with @xmath107 , shibata _ et al . _ @xcite have obtained a functional relation for the scattering threshold @xmath108 the numerical studies further demonstrate that grazing collisions with impact parameters near the range @xmath109 generate enormous amounts of gravitational radiation up to at least @xmath110 of the com energy ; for a detailed analysis of the gw emission see berti _ et al._@xcite . indeed it has been conjectured by pretorius & khurana @xcite that _ all _ kinetic energy may be radiated in the form of gws . this conjecture has recently been investigated numerically in sperhake _ _ @xcite by monitoring the bh horizon properties in collisions of spinning and non - spinning binaries . while most of the kinetic energy is indeed radiated for mild boosts @xmath111 , gw absorption by the holes becomes increasingly efficient for larger collision velocities ; cf . their fig . 1 . extrapolation of the numerical results to the ultrarelativistic limit predicts that approximately equal fractions of the kinetic energy are absorbed by the bhs as are lost in gravitational radiation , and sets a lower bound on the final bh mass of about half of the com energy . this study further demonstrates that the impact of the bh spin on the scattering threshold and radiated energy becomes negligible for @xmath112 lending additional support to the modeling of trans - planckian scattering in terms of high - energy collisions of bhs . the most likely scenarios for bh formation at the lhc are proton - proton collisions . it will be important , therefore , to investigate the impact of electric charge on the dynamics which has been ignored in the above mentioned studies . a first exploration of colliding bhs with electric charge has been presented in zilho _ they extract gravitational as well as electromagnetic radiation generated in head - on collisions of bhs with equal mass and charge starting from rest . as intuitively expected , the electric repulsion slows down the dynamics and leads to a decreasing amount of gw energy @xmath113 as the mass - to - charge ratio @xmath114 increases from @xmath115 towards the extremal case @xmath116 which corresponds to a static , albeit unstable , configuration . the radiated electromagnetic energy @xmath117 , on the other hand , is maximal near @xmath118 but always remains below the gravitational one with the quotient @xmath119 increasing monotonically as a function of @xmath114 reaching @xmath120 as @xmath121 . in contrast to the four - dimensional case , bh collisions in @xmath122 spacetime dimensions appear to be more challenging from the view point of numerical stability and have not yet been explored to the same extent . collisions of two equal - mass , non - spinning bhs starting from rest have been studied by witek _ _ @xcite who use the kodama - ishibashi formalism @xcite for the extraction of gws and find that @xmath123 of the com energy is radiated in gws for @xmath22 , to be compared with @xmath124 in four dimensions @xcite . these results were generalized to the case of unequal mass ratios @xmath125 in ( cf . their fig . 3 ) leading to the extreme - mass - ratio limit @xmath126 . point - particle calculations are in good qualitative and quantitative agreement with these nr results@xcite but additionally make very interesting predictions for bh head - on collisions with _ non - zero _ initial velocities @xcite : @xmath127 typically _ decreases _ with @xmath2 to such an extent that for @xmath43 , head - on collisions starting from rest may radiate more gw energy than their ultrarelativistic counterparts ; cf . fig . 1 in . this counter - intuitive prediction has yet to be tested with nr methods . bh collisions in @xmath22 dimensional cylindrical spacetimes have been simulated in . the cylindrical setup can be viewed in terms of an infinite array of bhs resulting in a breaking effect on the collision dynamics . the most advanced study of bh scattering in @xmath22 has been performed by okawa _ @xcite who find scattering thresholds @xmath128 decreasing in terms of the schwarzschild radius @xmath129 from @xmath130 at @xmath131 to @xmath132 at @xmath133 . above this velocity , numerical instabilities prevent a determination of @xmath128 . in contrast to the four - dimensional case , however , they do not observe any signs of zoom - whirl behaviour , probably due the equal fall - off @xmath134 of the gravitational and centrifugal potential in @xmath22 . by analysing the kretschmann scalar @xmath135 , they further conclude that super - planckian physics may be visible in the scattering of bhs in five spacetime dimensions . a straightforward calculation shows that for a single bh of mass @xmath9 , the kretschmann scalar on the horizon is given by arising from the different choices of @xmath136 in the einstein equations , our eq . ( [ eq : einstein ] ) and eq . ( 1 ) in . ] @xmath137 by calculating @xmath138 in the orbital plane , okawa _ et al . _ observe super - planckian domains where @xmath139 outside the individual holes horizons even for non - merging configurations . regions with curvature radii below the planck scale may thus form in high - energy particle collisions without being cloaked inside an event horizon . the gauge - gravity duality conjectures the mathematical equivalence or _ duality _ between field theories involving gravity in @xmath2 dimensions on the one side and field theories without gravity in @xmath1 dimensions on the other . it is one of the most remarkable results of string theory and , because of maldacena s prototypical example @xcite of the duality between type iib string theory in 5 dimensional ads space times the @xmath140 sphere and the @xmath141 supersymmetric @xmath142 yang - mills ( sym ) cft in 4 dimensions , often referred to as the ads / cft correspondence . following maldacena s original example , a large number of similar dualities have been conjectured ; see for example chapter 13 in ref . . many of these involve type iib string theory or @xmath9 theory compactified on @xmath143 times a sphere or torus . the best understood example to date , however , is given by maldacena s prototype which is also the version underlying present nr studies and , hence , the one we shall focus on in the remainder of this section . the ads / cft correspondence , while not rigorously proven , is inspired and supported by a variety of features displayed by the two theories involved : * holography : black holes can be regarded as thermodynamic systems with entropy @xmath144 , where @xmath145 is the horizon area @xcite . in contrast , the entropy of systems described by local field theories grows in proportion to the volume , suggesting a holographic connection to gravity @xcite . * symmetries : the isometry group of @xmath146 is @xmath147 which also leave @xmath141 sym invariant . fermionic symmetries agree similarly and lead to a supergroup @xmath148 which is identical to the @xmath141 superconformal symmetry ; cf . ref . . * @xmath149 branes : a stack of @xmath150 @xmath149 branes has an open - string description equivalent to @xmath141 sym @xcite . the stack also forms a solution of type iib supergravity , i. e. the low - energy limit of type iib string theory@xcite which in the near - horizon limit of the extremal solution reduces to the @xmath146 metric . this equivalence further relates the coupling constant @xmath151 and number of colors @xmath150 of the gauge theory with the string length and coupling @xmath152 , @xmath153 and the spacetime curvature radius @xmath28 on the gravity side . specifically , @xmath154 * the hawking - bekenstein entropy associated with the @xmath155 bh of the near - extremal @xmath149 brane is equal to the entropy of a non - ideal gas in @xmath141 sym in the strong t hooft coupling limit @xcite @xmath156 ; cf . and sec . 2.3 in . * the ratio of shear viscosity @xmath157 to entropy density @xmath158 is conjectured to have a lower bound given by the value @xmath159 which is universal to all theories with a gravity dual @xcite . this bound appears to be satisfied by all known liquids . * ads / cft based studies ( cf . below ) predict rapid thermalization on time scales @xmath160 for quark - gluon plasma produced at brookhaven s relativistic heavy ion collider ( rhic ) in agreement with observations @xcite . one of the most valuable properties of the correspondence arises from eq . ( [ eq : parameters ] ) . the duality is strongly / weakly coupled and can be applied in both directions : strong coupling of the gauge theory @xmath161 corresponds to @xmath162 , i. e. the regime of validity of general relativity , while the gauge theory is rather well understood for small @xmath163 and may provide information about the strongly coupled limit of string - theory . we need to bear in mind , however , that the correspondence applies to cousins of quantum chromodynamics ( qcd ) instead of qcd itself . nevertheless , the correspondence provides an unprecedented opportunity to theoretically model physical systems in the strong - coupling regime of field theories and , as demonstrated by the above mentioned studies on shear viscosity and thermalization of quark - gluon plasma , is capable of making qualitatively and even quantitatively accurate predictions for physical systems . reviews of the ads / cft correspondence with varying levels of detail are given in . for an overview of potential applications in fluid dynamics , condensed matter physics , super conductors and super fluidity we refer to refs . . anti - de sitter space in @xmath2 dimensions @xmath164 is the maximally symmetric solution of the @xmath2-dimensional vacuum einstein equations with cosmological constant @xmath165 @xmath166 it can be represented as the hyperboloid @xmath167 embedded in a @xmath168-dimensional flat spacetime of signature @xmath169 with metric @xmath170 transforming to coordinates @xmath171 with hyperspherical coordinates : @xmath172 , @xmath173 , @xmath174 , @xmath175 and @xmath176 . ] @xmath177 , we obtain @xmath164 in global coordinates with metric @xmath178 where @xmath179 , @xmath180 . by unwrapping the cylindrical direction , the range of the time coordinate is often extended to @xmath181 . alternatively , poincar coordinates @xmath182 defined by @xmath183,~~ x_i = \frac{lx^i}{z},~~{\rm for}~~i=1\ldots d-2 , \nonumber \\ & & x_{d-1 } = \frac{1}{2z } \left[z^2 - l^2 + \sum_{i=1}^{d-2}(x^i)^2 = t^2 \right],~~ x_d = \frac{lt}{z},\end{aligned}\ ] ] give the ads metric in the form @xmath184\ , , \label{eq : ads_poincare}\ ] ] with @xmath185 , @xmath186 . poincar coordinates cover only half the hyperboloid ( the other half corresponding to @xmath187 ) and this patch is often referred to as the _ poincar wedge_. for its relation to global ads see for example . at the ads boundary @xmath188 or @xmath189 , the ads metric becomes singular . it therefore induces a conformal class of metrics at the boundary , i. e. the induced metric is determined up to a conformal rescaling only . this remaining freedom manifests itself in the boundary topology of the global and poincar metrics . in the limit @xmath188 or @xmath189 , eqs . ( [ eq : ads_globalmetric ] ) and ( [ eq : ads_poincare ] ) become respectively @xmath190 which can be related by a conformal transformation ; cf . ( 8.17 ) in . in consequence , gravity in global or poincar ads is related to field theories on spacetimes of different topology : @xmath191 in the former and @xmath192 in the latter case . there remains the extraction of physical quantities of the field theory from the gravitational side . this _ dictionary _ between bulk and boundary physics is established by the equivalence of the gravitational action of the bulk taken in the limit of the boundary and the effective action of the field theory on the boundary @xcite . this duality involves divergences on both sides , _ infrared _ divergences on the gravity side due to the singular nature of the metric on the boundary and _ ultraviolet _ divergences of the quantum field theory . these anomalies and their elimination through renormalization have been discussed in . nr applications have so far focused on the calculation of the vacuum expectation values of the field theory s energy momentum tensor @xmath193 corresponding to the gravitational sector in the bulk and we shall only consider this case here . for a discussion including matter fields see for example sec . 5 in as well as . through the ads / cft correspondence , the vacuum expectation values @xmath193 of the field theory are given by the quasi - local brown - york @xcite stress - energy tensor and thus directly related to the bulk metric . following de haro _ @xcite , it is convenient to consider the ( asymptotically ads ) bulk metric in fefferman - graham @xcite coordinates @xmath194\ , , \label{eq : feffermangraham}\ ] ] where @xmath195 here , the @xmath196 and @xmath197 are functions of the boundary coordinates @xmath198 , the logarithmic term only appears for even @xmath2 and powers of @xmath7 are exclusively even up to order @xmath1 . as shown in , the vacuum expectation value of the cft momentum tensor for @xmath21 dimensions is then obtained from @xmath199 \right . \nonumber \\ & & \left . - \frac{1}{2 } \gamma_{(2)i}{}^m\gamma_{(2)jm } + \frac{1}{4 } \gamma_{(2)ij } \gamma_{(2 ) } \right\ } \,,\end{aligned}\ ] ] and @xmath200 is determined in terms of @xmath201 . the dynamic freedom of the cft is thus encapsulated in the fourth - order term @xmath202 . note that for @xmath203 , the metric ( [ eq : feffermangraham ] ) asymptotes to the ads metric in poincar coordinates ( [ eq : ads_poincare ] ) . the relation between the gravitational and cft energy - momentum tensor for the case of global ads is discussed in ; see also . many applications of the ads / cft correspondence are concerned with the equilibration of matter in heavy - ion collisions at the rhic or lhc and in particular its rapid thermalization @xcite . while the quark - gluon plasma generated in the collisions is far from equilibrium at early stages , its behavior appears to be well described by hydrodynamics after time scales of the order @xmath204 . this process is in principle governed by qcd but results indicate that many physical aspects can be studied in the framework of @xmath141 sym through the gauge - gravity duality . small perturbations of a static system in thermal equilibrium are known to decay exponentially fast and correspond to quasi - normal modes on the gravity side @xcite and , as we shall discuss , numerical studies in the context of the ads / cft correspondence yield a similar picture for far - from equilibrium configurations . the use of nr methods to study these processes in ads / cft has been pioneered by chesler & yaffe @xcite who evolve an anisotropic source on the ads boundary switched on after a short time using a characteristic approach based on ingoing eddington - finkelstein coordinates on the poincar patch of ads . their numerical scheme is reduced to an effective `` 1 + 1 '' scheme by assuming boost invariance as well as rotational and translation symmetry in the transverse plane . their boundary data generate gravitational waves which propagate into the bulk and lead to formation of a bh . they extract the energy momentum tensor of the cft dual and follow the time evolution of the energy density as well as the transverse and longitudinal pressure components . isotropization of the pressure occurs on time scales inversely proportional to the local temperature at the onset of the hydrodynamic regime which translates into an isotropization time of @xmath205 assuming a temperature of @xmath206 . using the same setup , chesler & teaney @xcite use different definitions of a `` temperature '' based on the energy density and on two - point functions . these two versions start agreeing after a time @xmath204 coincident with the isotropization of transverse and longitudinal pressure . wu & romatschke @xcite employ a similar approach to collide two superposed shockwaves in a boost invariant approximation and find that the late - time behavior of the energy density is given by a hydrodynamic description involving a scale parameter determined by the initial apparent horizon area . a comparison of the fully non - linear numerical results with predictions from the linearized close - limit approximation @xcite was performed by heller _ _ @xcite . instead of sourcing the anisotropy through a boundary term , they prescribe anisotropic data on an initial null hypersurface extending through the bulk . their results confirm the short isotropization times @xmath160 and find the linear approach to reproduce these values within @xmath207 even for large initial anisotropies . a numerical scheme based on the adm formalism of the einstein equations was developed by heller _ @xcite who emphasize that thermalization , when defined as the onset of the hydrodynamical regime , may differ from isotropization . by evolving boost - invariant , transversely homogeneous plasmas with various different initial conditions , they find that a hydrodynamic description may well be applicable when the pressure is still anisotropic . in ref . , chesler & yaffe relax their symmetry assumptions to translation invariance in the transverse direction which effectively constitutes a 2 + 1 numerical scheme . their characteristic formulation works well in this scenario without showing any signs of formation of caustics and enables them to model heavy - ion collisions by colliding two shock waves . specifically , they consider a single shock - wave solution in fefferman - graham coordinates , superpose two of those and transform the result to ingoing eddington - finkelstein coordinates . for stability purposes , they introduce a small energy offset on the cft side which generates an apparent horizon in the gravity dual above the poincar boundary to absorb steep gradients encountered in the metric functions deep in the bulk . by comparing the numerically determined pressure components with hydrodynamical predictions , they confirm the picture of rapid isotropization obtained in scenarios of higher symmetry . translated into values for gold ion collisions at the rhic , they observe isotropization about @xmath208 after their shock waves start overlapping . a numerical code based on cauchy type evolutions of asymptotically global ads spacetimes using the ghg scheme has been developed by bantilan _ _ @xcite . by assuming an @xmath209 symmetry , they reduce their computational domain to 2 + 1 dimensions . initial data is specified in the form of a localized scalar field which promptly collapses to a bh with a highly distorted horizon and settles down to a stationary configuration through quasi - normal ringdown . whereas the lowest ringdown modes agree with linearized predictions , higher angular modes exhibit significant coupling due to non - linear effects . the dual stress - energy tensor of the cft is mapped from the global @xmath210 ads boundary to a minkowski background and found to evolve in agreement with that of a thermalized sym fluid from the start of the simulations . their work furthermore discusses in detail a number of regularization procedures required to obtain a numerically stable framework . a very recent paper by adams _ @xcite presents a first numerical investigation of superfluid turbulence in 2 + 1 dimensions corresponding to a 3 + 1 dimensional asymptotically ads gravity dual . they observe energy injected at long wavelengths to cascade down to short length scales , in contrast to the behavior of normal fluid turbulence in 2 + 1 dimensions . the vast majority of the work reviewed here has been generated during the last four years ; nr in higher dimensions is a very young research field and likely has only scratched the surface of a wealth of possible applications . nevertheless , a number of impressive results have demonstrated its vast potential . discoveries such as the formation of naked singularities in black strings strengthen the belief that higher - dimensional gravity differs from its four - dimensional counterpart in more than a quantitative manner . the same holds for the asymptotic structure of the spacetimes under consideration . the instability of asymptotically anti - de sitter spacetimes to bh formation under arbitrarily small perturbations is in stark contrast to the stability of asymptotically flat or de sitter spacetimes@xcite . one of the remarkable properties of higher - dimensional gravity is its connection with other areas of physics . the holographic principle and the ads / cft correspondence in particular have opened a path towards understanding a variety of high - energy physics phenomena through the study of 4 + 1 gravity . this facilitates direct contact with experiments such as heavy - ion collisions performed at the rhic . at the same time , parton - parton collisions at tev energies at the lhc have provided for the first time a direct experimental test of the existence of extra dimensions . in spite of the great success of nr in recent years , a great deal of work remains to be done . numerical studies in @xmath2 dimensions have as yet been restricted to systems of considerable symmetry . this is largely a consequence of computational requirements which increase enormously with each extra dimension . continuous improvements of computer technology should , however , make possible in the foreseeable future the study of generic systems at least in @xmath22 , opening up a much wider class of problems in the context of the ads / cft correspondence and the stability analysis of bhs . even within the context of symmetric systems , many challenges are awaiting the community . trans - planckian scattering is rather well - understood in @xmath21 . connection with experiment , however , requires extension of these results to higher @xmath2 . here , progress has been obstructed by stability issues of the numerical codes , calling for further investigation of the numerical formulations and gauge conditions . many open questions also remain about the stability of bhs in higher dimensions . here we note in particular that bhs are no longer restricted to spherical topology and even stationary multiple - bh solutions are known . it is unclear how present numerical frameworks are able to handle black rings or saturns , although the relatively straightforward extension of nr studies to black strings may be encouraging . numerical codes developed for the ads / cft correspondence have as yet been tailored to rather specific problems . it would be desirable to have a generic framework effectively representing a numerical ads / cft laboratory . given the enormous range of bh systems in the context of the correspondence , this is likely going to be a tremendous challenge , albeit one of large scientific potential . in summary , numerical relativity in higher dimensions is a highly active and rapidly growing field and it will be very exciting to see its development over the next years . the author thanks e. berti , v. cardoso , r. emparan , p. figueras , l. gualtieri , c. herdeiro , d. mateos , a. nerozzi , h. okawa , h. s. reall , c. f. sopuerta , h. witek and m. zilho for many fruitful discussions . this work was supported by the ramn y cajal programme and grant fis2011 - 304145-c03 - 03 of the ministry of education and science of spain . this work was further supported by the fp7-people-2011-cig grant cbheo-293412 , the fp7-people-2011-irses grant nrhep-295189 , the erc starting grant dybho-236667 , the prace deci-7 grant `` black hole dynamics in metric theories of gravity '' , the nsf grant no . phy-090003 through xsede , grants aect-2012 - 3 - 0011 through res / bsc and icts-234 through cesga . part of this work was performed with the support of the cosmos system , part of the dirac hpc facility funded by stfc and bis , and the baltasar system of ist - centra .

many years ago , superconductivity was discovered in heavy - fermion compounds @xcite . it was suggested @xcite that the superconductivity was due to collective electronic fluctuations and not due to electron - phonon interactions . transport properties in the superconducting state were analyzed @xcite to show that superconductivity was in the d - wave symmetry . it was also suggested that the d - wave symmetry is promoted by antiferromagnetic fluctuations @xcite with long enough correlation lengths . this promotes scattering of fermions near the fermi - surface predominantly through angles around @xmath2 , which is essential for superconducting instability in the `` d - wave '' channel for a suitable fermi - surface @xcite . the idea of long enough afm correlation lengths as essential for this mechanism is supported by the fact that in heavy - fermions , superconductivity occurs generally in the regime near the afm quantum critical point where the correlation lengths are long but the competing afm phase has lower condensation energy . at the same time , random phase approximation on the hubbard model was used to calculate the spin - fluctuation spectra and to suggest that d - wave superconductivity is promoted by such fluctuations @xcite . the properties of the hubbard model have proven controversial in more elaborate calculations ; there are calculations which suggest that the ratio of the transition temperature @xmath0 to the typical electronic kinetic energy parameters @xmath3 is more than @xmath4 @xcite to less than @xmath5 @xcite . since heavy - fermion properties require kondo effect of the f - orbital local moments and their magnetic interactions using the wide - band electrons , a multi - orbital model is obviously required @xcite . the hubbard model was proposed as a sufficient model for the cuprate compounds @xcite . but the discovery in under - doped cuprates @xcite of the predicted time - reversal breaking order parameter @xcite on the basis of a multi - orbital model raises doubts on the validity of the hubbard model for the cuprates . for pnictides , generalization of the hubbard model to multi - orbital situations and inclusions of hund s rule couplings appears essential . we have a more modest goal in this paper than calculating spin - fluctuations from microscopic theory and using it to calculate properties of the superconductor . in recent years inelastic neutron scattering in the heavy fermion compounds cecu@xmath6si@xmath6 @xcite and ceirin@xmath7 @xcite have provided details of the afm fluctuation spectrum in the normal state . the primary aim of this paper is to estimate the superconducting transition temperature using the parameters provided by the experiments in these compounds . to do so , we solve the eliashberg equations for d - wave superconductivity using a phenomenological afm spectral function with which the experimental data is in good accord . the use of the eliashberg equations for quantitative calculations may be open to question because the migdal expansion parameter , which is of @xmath4 for the electron - phonon problem is of @xmath8 for such compounds if one assumes that the scale of the afm fluctuations extends to the order of the electronic bandwidth . however , when the afm correlation length @xmath9 is large compared to the lattice constant @xmath10 or @xmath11 , the scale of the afm fluctuations is reduced correspondingly to @xmath12 . but in the limit of large correlation lengths , new questions arise @xcite which are not important in the electron - phonon problem . the most prominent among them are the role of inelastic scattering in depressing @xmath0 on the one hand @xcite , and the fact that the bcs type coupling constant @xmath13 appears to diverge when the characteristic fluctuation frequency @xmath14 and the bcs prefactor appears to go to @xmath15 . an answer to these questions and various considerations which determine @xmath0 from afm interactions is possible from the numerical solution of the eliashberg equations . we find that it is reasonable to conclude from a comparison of the calculated @xmath0 with experiments that afm fluctuations are responsible for d - wave superconductivity in the heavy fermion compounds . very importantly , with similar parameters we calculate the _ measured _ coefficient of the anomalous @xmath1 contribution to the resistivity in these compounds . a claim to quantitative accuracy on both these quantities can however be made only to factors of @xmath16 . we note here that if one adopts that the dimensionless measure @xmath17 for how high is the electronic fluctuation induced superconducting , the heavy fermions may be said to do very well indeed . for example , in many cases , including the compounds studies here , this ratio is @xmath4 , similar to that of the cuprates . following the proposals that afm fluctuations may also promote superconductivity in the cuprate compounds @xcite , there have been many discussions of the mechanism and many calculations based on the eliashberg equations . a partial list includes the following @xcite . the most complete of these calculations appear to us to be those carried out by monthoux and lonzarich ( ml ) @xcite , both for 2 and 3 dimensional models . we present below calculations for the 2 dimensional square lattice model with a phenomenological spin - fluctuation spectrum , whose results are no different from those of ml for the range of parameters examined that are common . a difference in the calculations is that we vary the parameters in the two dimensional model so that `` nesting '' at the afm wave - vector quantitatively changes . the amount of nesting does have a significant effect on the results . more important is that now that the afm fluctuation spectra is available , we can use the experimental parameters to test the ideas quantitatively . we also discuss how to put limits on the parameters used based on sum - rule for the fluctuation spectra and show that they are inter - related . results for the range of physical parameters that we find relevant for the heavy fermions is not available in the published results of ml . this has bearing also on general conditions to determine the extent to which afm fluctuations give significant @xmath0 for relevant parameters in other compounds . this paper is organized as follows : we present in sec . ( ii ) the models for fermi - surface and for the spin - fluctuations which we have investigated using the linearized eliashberg equations . we discuss there the change of effective coupling constants with the afm correlation length using sum - rules so that the results for numerical solutions of the eliashberg equations presented later are presaged . we present the results of the calculations in sec . ( iii ) and discuss the important conclusions immediately after the description of the models . we also present , in an appendix , the explicit derivation of the coefficient of the @xmath18 resistivity from the measured form of the afm critical fluctuation spectra . this is used in the text to estimate independently the value of a coefficient @xmath13 , which is important for the calculation of @xmath0 . we give the parameters that have been deduced by inelastic neutron scattering for the heavy fermion compounds cecu@xmath6si@xmath6 @xcite and ceirin@xmath7 @xcite and compare the measured @xmath0 with the calculations . we should emphasize that such a comparison is meant to be only illustrative of the physical principles involved ; no detailed quantitative agreement is to be expected , especially given that the electronic structure of these compounds is far more complicated than assumed in the models studied . however , enough details can be provided so that one can conclude that the idea of afm fluctuations near the quantum critical point in these compounds as the source of d - wave superconductivity is well supported . for example using measured properties , different levels of assumed nesting in the band - structure need a coupling constant @xmath13 between 1.5 and 3 to get the measured @xmath0 . in this range of @xmath13 and for the measured afm correlation length @xmath0 is close to being linear in @xmath13 . this range of values is compared with the value of @xmath19 needed to get the measured coefficient of the @xmath18 resistivity , which is relatively insensitive to nesting . one can assess the results from the fact that in the range of @xmath13 deduced , @xmath20 is found to be approximately linear in @xmath13 . in our calculations we will consider two types of fermi - surfaces , a free electron fermi - surface and the others given by the tight binding spectrum in a two dimensional square lattice with nearest neighbor and next nearest neighbor hopping @xmath3 and @xmath21 respectively : @xmath22 the fermi - surface with the tight binding spectrum are shown in fig . [ fig:01 ] for four values of the next nearest hopping @xmath23 and the afm wave - vector . the nesting in the model changes as @xmath21 increases . we will show detailed result for three fermi - surfaces , the free electron fermi - surface , the fermi surface ( fs1 ) with tight binding spectrum with @xmath24 and the fermi surface ( fs2 ) with tight binding spectrum with @xmath25 . of the four fermi - surfaces shown in fig . [ fig:01 ] , the one with @xmath24 has the worst nesting and the one with @xmath25 has the best nesting . [ fig:02 ] shows the circular fermi - surface , fs1 , fs2 and the corresponding afm wave - vectors . we will discuss using the results of ml together with ours , that if properly normalized density of states and fluctuation spectra are used , two dimensional and three dimensional models give similar results for @xmath0 provided one adjusts the ratio of the region of fermi - surface nesting to the total fermi - surface . this is in general is always lower in three than in two dimensions . it is also important to note , as discovered long ago @xcite , @xcite for the case of s - wave superconductors that @xmath20 is a rather gross quantity which depends to a very good approximation on the average density of states near the chemical potential only and not on details such as the number of fermi - surface sheets and shapes . for d - wave superconductors , we we show below , it is important to also include effects of nesting of the fermi - surface near the afm wave - vectors . the dynamical spin - susceptibility for itinerant fermions with afm correlations may be usefully divided into two parts : the normal contribution of non - interacting fermions @xmath26 and the part @xmath27 affected by afm correlations . the two together obey the total magnetic moment sum - rule . the non - interacting susceptibility can play only an insignificant role in promoting superconductivity @xcite and should be ignored . the two contributions to the susceptibility can be distinguished by their momentum dependence . the characteristic momentum dependence of the non - interacting spin fluctuations is on a scale of @xmath28 while that of the afm spin - fluctuations is much shorter . a partial sum - rule on @xmath27 in terms of the ordered moment in the afm phase can be used to relate the integrated fluctuations to the afm correlation length . double counting by using the sum - rule on the total susceptibility for the fluctuations and yet having ( free ) fermions interacting with such spin - fluctuations is incorrect as it over - counts the total degrees of freedom . such considerations have usually been blithely ignored in most of the previous phenomenological work on this problem . the fermions interact with spin - fluctuations with a phenomenological action @xmath29 @xmath30 will be chosen to have dimensions of inverse of energy ( after subsuming a factor of @xmath31 in its definition ) . so @xmath32 is a coupling function of dimension of energy . @xmath32 for heavy fermions is the exchange energy between the conduction electrons and the @xmath33- local moments . its meaning for d - band problems is more ambiguous , and may be best inferred from independent experiments , for example the resistivity above @xmath20 . a suitable phenomenological form for the dynamical spin - fluctuations due to afm correlations , with which experimental results @xcite can be fitted , is @xmath34 . \label{chi}\end{aligned}\ ] ] where @xmath35 is the damping rate of the fluctuations , @xmath36 is the antiferromagnetic vector . the correlation length @xmath9 is related to the deviation from the quantum critical point ( qcp ) by variation in pressure , doping , magnetic field , etc . as well as by temperature . @xmath35 , @xmath37 and @xmath9 may all be determined from experiments . the temperature dependence of @xmath9 has been studied by renormalization group ( rg ) @xcite and by the self - consistent renormalization ( scr ) methods @xcite . @xmath38 near afm qcp in the 3 dimensions and dynamical critical exponent @xmath39 . scr derives using the same dynamical critical exponent that near the magnetic qcp , @xmath40 . in our calculations in two dimensions @xmath9 will be assumed to be of the form @xmath41 where @xmath42 is the asymptotic @xmath43 value of the correlation length . one of our results is that the temperature dependence of @xmath9 is of insignificant consequence in determining @xmath0 . the linearized elisahberg equations give that the kernel for cooper pair coupling in the d - wave channel in a square lattice is proportional to the projection of @xmath44 to @xmath45 . in spin - fluctuations theories , @xmath46 has only a smooth momentum dependence . so , cooper - pair coupling prominently depends depends only on ( i ) the momentum dependence of @xmath47 determined by the correlation length @xmath9 ( ii ) the integrated weight in the momentum dependent part and ( iii ) the energy scale of the momentum dependent fluctuations . the first is qualitatively obvious from the fact that a q - independent spin - fluctuation contributes zero to the cooper channel in the d - wave channel . it is not possible to make quantitative statements on these effects without detailed calculations because the results also depend on the nesting in the band - structure near the afm @xmath37 . we will show that the three ingredients in @xmath47 are not mutually independent . to gain physical insight , the effect of @xmath9 on the integrated spectral weight may be discussed before detailed calculations through the the partial sum - rule on @xmath48 , which determines the effective coupling constant for superconductivity : @xmath49 \right ] . \label{tmsr}\end{aligned}\ ] ] with the assumed lorentzian form , it is necessary to introduce an upper cut - off @xmath50 in the frequencies @xmath51 up to which the fluctuations extend . actually , spin fluctuation are actually quite suppressed for @xmath52 and we can simply use @xmath53 in calculations of eliashberg equations . it is important to take into account that there are four equivalent afm - vector for the two dimensional problem in the paramagnetic regime of the model , however strongly fluctuating it may be . this has been taken into account in the sum - rule by multiplying the measured im@xmath54 by 4 . for @xmath55 , the number of equivalent afm vectors is larger and a correspondingly larger multiplicative factor should be used . in the regime of very long correlation lengths , @xmath56 , i.e. close to the quantum - critical point , the sum rule simply gives @xmath57 @xmath58 may to a first approximation be estimated from the ordered moment @xmath59 in nearby afm phase but more properly from integration of the relevant momentum and frequency range of the measured fluctuations in absolute units using polarized neutrons . [ fig:00 ] shows the @xmath60-dependence of @xmath61 for @xmath62=1.0 , 2.0 and 10.0 . let us now consider the sum - rule in the opposite limit , that the correlation length is small compared to the lattice constant , i.e. the system is very far from the quantum critical point . then @xmath63\ ] ] as already shown by ml and further elaborated below , for a given band - structure , the results of the eliashberg calculations for @xmath64 may be parametrized in terms of a dimensionless `` bare '' coupling constant @xmath13 and a correlation length @xmath9 , @xmath65 @xmath66 may be determined in terms of @xmath58 and therefore ( approximately ) to the ordered moment through the sum - rule . we may define an effective coupling constant @xmath67 to incorporate the effect of the correlation length . using that the sum - rule becomes the total moment sum - rule in the limit of infinite correlation length and the maximum possible ordered moment , i.e. that of the afm insulator ( ignoring the zero - point effects ) , one concludes that in the limit of very large correlation lengths @xmath68 where @xmath33 is the fraction of the maximum possible ordered moment . from fig . ( [ fig:00 ] and from the detailed calculations presented in the next section , one deduces that the limit for @xmath69 is reached for @xmath70 , below which there is an exponential fall off of @xmath64 . for smaller correlation - lengths , fig . ( [ fig:00 ] ) shows that the @xmath67 for @xmath0 decreases with decreasing correlation length . in the work of ml , @xmath13 values from about 5 to about 50 are used in the calculations with varying correlation lengths . actually , one obtains for the considerations of the sum - rules above that for spin-(1/2 ) problems , even the coupling constant @xmath69 is only of @xmath8 , because @xmath71 is of @xmath8 and so is the upper limit on the ratio @xmath62 . an independent estimate of @xmath67 may be obtained from the normal state properties , for example the coefficient of the temperature dependence of the resistivity of non - fermi - liquid form in the quantum critical region . again only @xmath13 of @xmath8 will be found consistent . it is also important to note that these bcs type coupling constants do not carry information on the retardation effects due to the frequency dependence of the interaction ; these as well as the effects of inelastic scattering which are particularly important for anisotropic superconductors are properly treated through the numerical solution of the eliashberg equations . the difference from electron - phonon induced superconductivity where a single parameter @xmath13 need by introduced @xcite should also be noted . it should also be pointed out that in some heavy fermions , the quantum - critical fluctuations do not have the functional form given by the simple rg or scr approximations as above , but displays `` local criticality '' @xcite as suggested for the cuprates @xcite . in this paper , we only consider fluctuations which are well specified by the form given above . the temperature dependence of the resistivity near the quantum - critical points has been derived several times @xcite . here , we rederive it paying special attention to the coefficient in front of the anomalous temperature dependent part . an expression of the resistivity @xmath72 in the antiferromagnetic quantum critical region suitable for heavy fermions may be derived with the following formula derived from the boltzmann equation . @xmath73 where the integration is taken over the fermi - surface , and @xmath74 the fermi velocity . this assumes that the _ actual _ electronic structure near the chemical potential is sufficiently complicated that in the temperature region of interest , vertex corrections which lead to emphasis on large momentum scattering for resistivity are unimportant . in that case the scattering rate which determines the resistivity is the same as the single - particle scattering rate averaged over the fermi - surface . this is true in a multi - sheeted fermi - surface and is suitable for heavy fermions . this is similar to the case of transition metals where the resistivity from electron - phonon scattering at low temperatures is @xmath75 in contrast to the nearly free - electron metals where it is @xmath76 . for weakly anisotropic single band scattering , as in the cuprates , the resistivity for large afm correlation lengths is close to the fermi - liquid temperature dependence although near the hot spots the scattering rate is nearly @xmath77 @xcite . equation ( [ aq1 ] ) can also be expressed as follows . @xmath78 where @xmath79 means the average over the fermi - surface , @xmath80 the renormalized effective mass . here , @xmath81 can be derived from the imaginary part of the self - energy . @xmath82 where the self - energy due to the antiferromagnetic quantum fluctuations is given as follows . @xmath83 the result for the resistivity in the limit @xmath84 is derived in an appendix a. by explicitly calculating the self - energy given by eq . ( [ aq4 ] ) is @xmath85 the superconducting transition temperature is given by the linearized version of the eliashberg equations for the normal self - energy @xmath86 and the anomalous or pairing self - energy @xmath87 . @xmath88 i \omega_n & = - \int_{\rm fs } \frac { { \rm d}^{d } s_{\vec{p } } } { ( 2 \pi)^d v_{\vec{p } } } \pi t \sum_{\omega_m } i\ { \rm sgn}(\omega_m ) g^2 \chi ( \vec{k } - \vec{p},i \omega_n - i\omega_m),\\ \label{leliash2 } w ( \theta_{\vec k } , i \omega_n ) & = - \int_{\rm fs } \frac { { \rm d}^{d } s_{\vec{p } } } { ( 2 \pi)^d v_{\vec{p } } } \pi t \sum_{\omega_m } \frac { w ( \theta_{\vec p } , i \omega_m ) } { \vert \omega_m z ( \theta_{\vec{p } } , i \omega_m ) \vert } g^2 \chi ( \vec{k } - \vec{p},i \omega_n - i\omega_m).\end{aligned}\ ] ] here @xmath89 are the matsubara frequencies ; @xmath32 is a momentum - independent coupling matrix element , which has already been defined , @xmath90 is an angle parameterizing the fermi surface , @xmath91 is the density of states at angle @xmath90 . the @xmath92-integral is over the fermi surface , @xmath93 is the unrenormalized velocity . our principal general results for @xmath0 on the basis of solution of the linearized eliashberg equations in terms of @xmath13 and the parameters in @xmath48 are given in this section . the numerical evaluation is done by first simplifying the eliashberg equations ( [ leliash1 ] ) and ( [ leliash2 ] ) , as far as possible analytically . the final expressions for the numerical evaluation , both for the circular fermi - surface and the tight - binding fermi - surfaces are given in appendix b. figure . [ fig:03a ] show the @xmath94-dependences of @xmath64 for the circular fermi - surface on the bare coupling constant @xmath13 in the large correlation length limit on the right . for small bare coupling @xmath13 , the latter does have the bcs form while for @xmath95 , the dependence is approximately linear . consistent with earlier discussions @xcite , @xmath64 shows a very shallow peak at around @xmath96 . @xmath64 shows a drop - off as the correlation length decreases , while it also shows moderate decreases as the correlation length increases . ml pointed out that this moderate decrease is caused by the rapid diverges of @xmath97 as the correlation length increases . we note also that the quantum - classical crossover correction to the correlation length proportional to the factor @xmath98 in eqs . ( [ corr2 ] ) has a negligible effect on @xmath0 . this will not be considered in any further calculations . the principal message from fig . ( [ fig:03a ] ) is that the infinite correlation length result for @xmath0 is well obeyed up to @xmath99 with a very sharp fall off thereafter which will be seen later to be exponential . for large @xmath100 , no bcs type approximation for @xmath0 is valid . the limit of very large correlation length is equivalent to the effective frequency of fluctuations @xmath14 , as may be seen from eq.([chi ] ) . if we use the mcmillan @xcite type approximation , in which @xmath101 , the inverse of the average squared frequency of fluctuations , we get a divergent coupling . ( [ fig:03a ] ) gives a finite limit to @xmath64 , which depends on the bare coupling constant @xmath13 . one may understand this result from the calculations of allen and dynes @xcite , deduced for s - wave eliashberg equation , that in the limit of a diverging coupling constant @xmath102 , where @xmath103 may be taken approximately to be the square - root of @xmath104 . - next we show in fig . ( [ fig:04a ] ) the @xmath94-dependence of @xmath64 for the four fermi surface shown in fig . ( [ fig:01 ] ) . for @xmath24 , the worst nesting case , @xmath64 is similar to that for the circular fermi - surface . improving the nesting condition increases , @xmath64 . we show in ( [ fig:05a ] ) @xmath20 as a function of @xmath13 for the worst nesting fermi - surface and the best nested fermi - surface of those in fig . ( [ fig:01 ] ) . a increase of @xmath16 in @xmath0 for the similar values of @xmath13 is discerned from the worst to the best nesting conditions . . [ fig:05a ] ml have also presented detailed results for calculations on a 3 d electronic dispersion with the symmetry of a cubic lattice . they remark that other parameters being the same 2 dispersion gives higher @xmath0 than a three dimensional dispersion . based on our results for changes in @xmath0 in the 2 d problem , we conclude that this is because of the much better nesting that is obtainable in model 2 d systems compared to the 3d systems for a given @xmath37 which spans the fermi - surface in some ( usually symmetry ) direction . in fact , we can place our 2 d - results for the very weakly nested fermi - surface over the ml results for the 3d fermi - surface and find for other parameters the same that the systematics of the results for @xmath0 as well as its value is very similar . in this section , we compare the estimates of @xmath0 from the calculations with the experimental result in cecu@xmath6si@xmath6 and ceirin@xmath7 . for convenience , we show the measured intensity @xcite proportional to the dynamic structure factor @xmath105 in fig . ( [ fig : s(q ) ] ) for @xmath106 near @xmath107 . . [ fig : s(q ) ] although the magnetic fluctuation spectrum found through inelastic scattering in cecu@xmath6si@xmath6 is well represented by the form of eq . ( [ chi ] ) , the electronic structure is far more complicated than assumed here or in the 3 d calculations of ml . we have seen that @xmath0 , especially in the limit of large magnetic correlation lengths depends only on gross parameters like @xmath13 and secondarily on the amount of nesting . the comparison can only be very limited and can only give insight into the orders of magnitudes expected and to the physics involved . [ [ cecu_2si_2 ] ] cecu@xmath6si@xmath6 + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + @xcite + to fit the phenomenological susceptibility to these results , the parameters take the following values : + _ afm wave vector _ : @xmath108= ( 0.22 , 0.22 , 1.46 ) + _ ordered moment in the afm phase _ : @xmath109/ce _ correlation length : _ @xmath110 25 ; + _ characteristic energy scales _ : + @xmath111 1.5 [ mev ] ; + _ density of states _ : from the measured uniform @xmath112 paramagnetic susceptibility in the normal state , one deduces the @xmath113 ( 1/7 ) [ mev]@xmath114unit - cell . + _ afm spin - fluctuations parameter _ : @xmath66 : + the experimental results for @xmath115 in ref . ( ) are parametrized in terms of three quantities @xmath116 and @xmath117 . the correlation length @xmath9 in ref . ( ) is the same quantity used by us . for clarity we give here the relation of the other two parameters to the parameters used by us . the conversion from the quantity @xmath118 , which we will call @xmath119 to our @xmath120 is obtained by equating the integral over all @xmath121 of equation s4 in ref . ( ) to the integral of the same physical quantity given in eq . ( [ tmsr ] ) . in the limit of @xmath122 , one gets @xmath123 . the experimental result is @xmath124 . this then gives @xmath125 . + the quantity @xmath117 is related to @xmath126 by @xmath127 . + _ transition temperature _ : @xmath128 0.6[k ] + cecu@xmath6si@xmath6 has a very anisotropic fermi - surface with very little dispersion along the tetragonal axis . the fermi - surface in the plane is very complicated but we assume that just as in s - wave superconductivity @xcite , @xmath0 depends only on the average density of states at the fermi - surface , supplemented by knowledge of nesting of the fermi - surface near @xmath129 . among other things , our results below may be taken to be test of this assumption . [ [ ceirin_5 ] ] ceirin@xmath7 + + + + + + + + + + + + + + + + + + + + + + @xcite + although long - range magnetic order competing with superconductivity in ceirin@xmath7 has not been accessed in this stoichiometric compound , there are strong experimental results indicating that the compound lies in the vicinity to an afm quantum - critical point . the resistivity of this material exhibits a non - fermi liquid behavior similar to that observed in cecoin@xmath7 , which is known to lie in the vicinity of an afm quantum - critical point which has been accessed by doping the compound . moreover , the nuclear spin relaxation rate of ceirin@xmath7 is also similar to that of cecoin@xmath7 . the dynamical susceptibility has been recently deduced by nmr experiments @xcite in agreement with this conclusion . + to fit the susceptibility to these experimental result , the parameters take the following values . + _ afm wave vector _ : @xmath108= ( 0.5 , 0.5 , 0.5 ) + the chosen @xmath108 and ordered moment are taken to be that of the related compound cecoin@xmath7 @xcite + _ ordered moment in the afm phase _ : @xmath130 + _ correlation length _ + @xmath131 at @xmath132[k]@xcite _ characteristic energy scales of afm : _ @xmath133 1.5[mev ] ; + _ transition temperature _ : @xmath0 = 0.4[k ] the experimental results show that both cecu@xmath6si@xmath6 and ceirin@xmath7 lie not far from the asymptotic large correlation length limit and that their @xmath64 are both about 0.03 . for a circular fermi - surface , and using the measure value of @xmath100 in the former , we may refer then simply to fig . ( [ fig:03a ] ) and find that @xmath134 gives the right value of @xmath0 . for the best nested fermi - surface , however , a @xmath135 is sufficient as shown in fig . ( [ fig:05a ] ) . we may now try to estimate @xmath13 to see if these values are reasonable . we do this in two different ways . to utilize the neutron scattering results for this purpose , we need to know @xmath32 besides the directly measured properties listed above . the renormalizations in the heavy fermion problem are such that near the critical point the afm interaction between magnetic moments is of the same order as the heavy fermion bandwidth . then @xmath32 is of the order of the effective fermi - energy , i.e @xmath136 . then we may use the experimental values of @xmath137 and @xmath138 deduced from experiments above in eq.([lambda ] ) to get @xmath139 . the above manner of estimation has forced us to guess the value of @xmath32 . we can estimate the value of @xmath13 much better and independently from the non - fermi - liquid resistivity proportional to @xmath18 observed in the quantum critical regime of cecu@xmath6si@xmath6 , whose coefficient is proportional to @xmath13 . the resistivity @xmath140 in the quantum - critical region for @xmath84 is given by eq.([rho ] ) . using the values of cecu@xmath6si@xmath6 mentioned above , the resistivity is estimated @xmath141 [ @xmath142 from eq . ( [ rho ] ) . the non - fermi liquid resistivity observed in cecu@xmath6si@xmath6 takes the form : @xmath143 @xcite where @xmath144 @xcite . from the comparison of the coefficient of @xmath18-term in the resistivity between theoretical and experimental results , @xmath13 is estimated as @xmath145 . this should be considered an important evidence for the rather obvious idea that fluctuations that determine the normal state scattering also determine @xmath0 , and of the consistency of the present calculations . the extent to which the calculations correctly estimate @xmath0 may be judged from the fact that in the range of @xmath13 from the different estimates for it , @xmath146 . + we comment briefly on an estimation of the condensation energy due to superconductivity and its comparison with the increase in energy of afm fluctuations on entering superconductivity @xcite . the latter has been estimated to be almost a factor of 20 larger than the superconducting condensation energy . the suggestion has been offered that this factor of 20 may be the increase in kinetic energy . in bcs theory for electron - phonon interactions , the absolute magnitude of the change in kinetic and in potential energy are both of the same order as the condensation energy . so , a good reason has to be found for this factor of 20 . we do not have a solution to this enigma . we have presented a solution to the linearized eliashberg equations using a phenomenological spin - fluctuation spectrum and simple fermi - surfaces to highlight the important parameters that determine @xmath0 for d - wave symmetry . careful attention has been paid to the partial sum - rule on the q - dependent part of the spin - fluctuation spectra to estimate the effective coupling constant which depends on parameters such as the total partial spectral weight , the correlation length and the upper frequency cut - off of the q - dependent spin - fluctuations . these parameters are not independent and we show their relationship in the simple model studied . with regard to the electronic structure , a knowledge of the average density of states at the fermi - surface is sufficient for determining @xmath0 in the s - wave channel @xcite . but for d - wave superconductivity through exchange of well correlated spin - fluctuations , this must be supplemented by a knowledge of nesting . the results for the general solutions are employed for two heavy fermion compounds using their measured spin - fluctuation spectra and other quantities such as specific heat and magnetic susceptibility . correct estimates for @xmath0 to factors of @xmath16 are obtained . confidence in these results is bolstered by getting the correct observed temperature dependence of the anomalous @xmath18 resistivity with a coefficient using the same parameters , again correct to factors of @xmath16 . this puts a semi - quantitative backbone to the surmise made long ago that d - wave superconductivity in such heavy fermions is promoted by large amplitude spin - fluctuations with large correlation lengths such as occur near some afm quantum critical points . + _ acknowledgements : _ part of the work by sn was done while visiting university of california , riverside with the aid of the global coe program ( g10 ) from the japan society for the promotion of science . the work of km is partially supported by a grand - in - aid for scientific research on innovative area `` heavy electrons '' ( no . 20102008 ) and a grant - in - aid for specially promoted research ( no . 20001004 ) from the ministry of education , culture , sports , science and technology , japan . the work of cmv is partially supported by nsf under grant dmr-1206298 . + * appendix a : derivation of resistivity near the antiferromagnetic quantum critical point * + an expression for the resistivity under assumptions suitable for heavy fermions with a multi - sheeted fermi - surface and/or sufficient impurity scattering @xcite is given by eq . ( [ rho ] ) in terms of the self - energy function eq . ( [ aq4 ] ) . here , we derive the relation ( [ rho ] ) explicitly . substituting @xmath147 in the spectral representation into eq . ( [ aq4 ] ) and carrying out the @xmath148-summation , one gets @xmath149 taking the analytic continuation of @xmath150 , the imaginary is given as @xmath151 ^ 2 + \left ( \frac { \varepsilon - \xi_{\vec{p } - \vec{q } } } { \gamma_{\rm afm } } \right)^2 } \left ( \tanh \frac{\xi_{\vec{p - q}}}{2k_{\rm b}t } + \coth \frac{\varepsilon - \xi_{\vec{p - q } } } { 2k_{\rm b}t } \right ) . \label{aq6}\end{aligned}\ ] ] we now consider the behavior at around the antiferromagnetic quantum critical point , i.e. , @xmath152 . in a low temperature region where the non - fermi liquid behavior appears , @xmath153 gives the dominant contribution for eq . ( [ aq6 ] ) . moreover , using the following relation , @xmath154 eq . ( [ aq6 ] ) is transformed as @xmath155 next , we consider the @xmath156-integration in eq . ( [ aq8 ] ) . because the denominator in eq . ( [ aq8 ] ) has @xmath157 term , @xmath158 gives the dominant contribution in the @xmath159-integration . therefore , one gets @xmath160 since the integrated function in eq . ( [ aq9 ] ) rapidly decays as @xmath161 increases , we take @xmath162 as @xmath163 and obtain following result by easy calculation . @xmath164 here , we estimate the average over the fermi - surface in eq . ( [ aq11 ] ) assuming that the fermi - surface is spherical . @xmath168 the dominant contribution in eq . ( [ aq12 ] ) comes from `` hot '' line where the relation @xmath169 is satisfied as shown in fig . [ fig:11 ] assuming that the dispersion near the fermi - surface is given by linear dispersion , we obtain @xmath170 , where @xmath171 is the deviation from the `` hot '' line . for one `` hot '' spot , the integration is estimated as @xmath172 changing the integration variable as @xmath173 , eq . ( [ aq13 ] ) is transformed as @xmath174 now , we take the upper limit of the integration as @xmath163 because we consider the low temperature region , and eq . ( [ aq14 ] ) can be calculated as @xmath175 substituting eq . ( [ aq16 ] ) into eq . ( [ aq11 ] ) , we obtain @xmath178 in this calculation , the @xmath36-vector is given by @xmath179 . the @xmath36-vector of the cecu@xmath6si@xmath6 is observed as ( 0.215 , 0.215 , 0.1458 ) giving @xmath180 . therefore , @xmath181 is estimated as @xmath182 , and we obtain the result used for the estimation of @xmath13 : @xmath183 * appendix b : final expressions for evaluation of @xmath0 * + _ circular fermi - surface _ + for a circular fermi - surface , it is possible to do the momentum integrals in the eliashberg equations ( [ leliash1 ] ) and ( [ leliash2 ] ) analytically so that only a diagonalization in discrete frequency space needs to be done numerically . the final expressions used for numerical evaluation for the normal and the anomalous self - energy are : where @xmath185,\\ x & = & \tan^{-1 } \left [ \frac { \vert \vec{q } \vert \sin \theta_{\vec{q } } + \vert \vec{p } \vert \sin \theta_{\vec{p } } } { \vert \vec{q } \vert \cos \theta_{\vec{q } } + \vert \vec{p } \vert \cos \theta_{\vec{p } } } \right ] . \label{f4}\\\end{aligned}\ ] ] _ tight - binding fermi - surfaces _ + with tight binding approximation , only some simplifications in the momentum integrals in the eliashberg equations ( [ leliash1 ] ) and ( [ leliash2 ] ) can be done analytically . the final expressions used in this paper for numerical evaluation are @xmath186 for both circular and tight binding fermi - surfaces , the best numerical strategy to evaluate @xmath0 is to cast eqs . ( [ f3 ] ) and ( [ f6 ] ) in the form of an eigenvalue equation for the eigenvector @xmath187 @xmath188 \left [ \frac{w(i \omega_m ) } { \vert \omega_m \vert } \right ] = 0 . \label{a9}\end{aligned}\ ] ] it should be noted that the matrix of eq . ( [ a9 ] ) is not hermitian because @xmath189 includes the renormalization factor @xmath190 . if the angle dependence of @xmath97 can be neglected , we can define @xmath191 in a form which does not include @xmath97 , and we obtain the eigenvalue equation with a hermitian matrix for the eigenvector @xmath192 . in s - wave superconductor case , such a situation , namely angle - independent self - energy appears . however , in the d - wave case , @xmath193 strongly depends on @xmath90 . on including @xmath97 in the kernel @xmath191 , the latter is no longer symmetric for the frequency exchange , @xmath89 and @xmath194 . at high temperatures the eigenvalues of eq . ( [ a9 ] ) are close to the negative odd integers . as the temperature decreases , the largest eigenvalue increases and crosses zero at transition temperature @xmath195 . + 99 f. steglich , j. aarts , c. d. bredl , w. lieke , d. meschede , w. franz , and h. sch@xmath196fer , phys . lett . * 43 * , 1892 ( 1979 ) . h. r. ott , h. rudigier , z. fisk , and j. l. smith , phys . * 50 * , 1595 ( 1983 ) . g. r. stewart , z. fisk , j. o. willis , and j. l. smith , phys . lett . * 52 * , 679 ( 1984 ) . c. m. varma , in _ proceedings of the nato advanced summer institute on the formation of local moments in metals _ , vancouver , canada , 1983 ( edited by w. buyers ) , plenum press , new york ( 1984 ) ; comments on solid state physics , vol . 11 , 221 ( 1985 ) . s. schmitt - rink , k. miyake and c. m. varma , phys . 57 * , 2575 ( 1986 ) . p. j. hirschfeld , p. w@xmath197lfle , and d. einzel , phys . b * 37 * , 83 ( 1988 ) . k. miyake , s. schmitt - rink , and c. m. varma , phys . b * 34 * , 6554 ( 1986 ) . c. m. varma , rep . prog . phys . * 75 * , 052501 ( 2012 ) . d. j. scalapino , e. loh , jr . and j. e. hirsch , phys . rev . b * 34 * , 8190 ( 1986 ) . t. a. maier , et al . lett , * 95 * , 237001 ( 2005 ) ; d. snchal , et al . lett . , * 94 * , 156404 ( 2005 ) ; m. capone and g. kotliar , phys . b * 74 * , 054513 ( 2006 ) . takeshi aimi , masatoshi imada , j. phys . * 76 * , 113708 ( 2007 ) . c.m . varma and y. yafet , phys . 13 , 2950 ( 1976 ) . p. w. anderson , science , * 235 * , 1196 ( 1987 ) . for a review , see p. bourges and y. sidis , c. r. physique * 20 * , ( 2010 ) . c. m. varma , phys . b * 55 * , 14554 ( 1997 ) ; m. e. simon and c. m. varma , phys . lett . * 89 * , 270003 ( 2002 ) . a. schroeder et al . , phys . 80 ( 1998 ) 5623 c.m . varma et al . , 63 , 1996 ( 1989 ) ; v. aji and c.m . varma , phys . b 79 , 18450 ( 2009 ) . o. stockert , j. arndt , e. faulhaver , c. geibel , h. s. jeevan , s. kirchner , m. loewenhaupt , k. schmalzl , w. schmidt , q. si , and f. steglich , nature phys . * 7 * , 119 ( 2011 ) . o. stockert , j. arndt , a. schneidewind , h. schneider , h.s . jeevan , c. geibel , f. steglich , and m. loewenhaupt , physica b , * 403 * , 973 ( 2008 ) . j. arndt , o. stockert , h. s. jeevan , c. geibel , and f. steglich , phys . lett . * 106 * , 24601 ( 2011 ) . s. kambe , h. sakai , and y. tokunaga , and r. e. walstedt , phys . b * 82 * , 144503 ( 2010 ) . a. millis , s. sachdev and c.m . varma , phys . b * 37 * 4975 ( 1988 ) . see , for example the following and references therein , p. monthoux , a.v . balatsky and d. pines , phys . b * 46 * , 14803 ( 1992 ) ; d.j . scalapino , in _ handbook of high temperature superconductivity _ , edited by j. schrieffer and j. brooks ( springer , new york , 2007 ) . m. r. norman , phys . b * 37 * , 4987 ( 1988 ) ; a. j. millis , phys . rev . b * 45 * , 13047 ( 1992 ) ; p. monthoux , a. v. balatsky , and d. pines , phys . . lett . * 67 * , 3448 ( 1991 ) ; p. monthoux and d. pines , phys . * 69 * , 961 ( 1992 ) ; p. monthoux and d. pines , phys . b * 47 * , 6069 ( 1993 ) ; p. monthoux and d. pines , phys . rev . b * 49 * , 4261 ( 1994 ) ; p. monthoux , j. phys : condens . matter * 15 * , s1973 ( 2003 ) ; p. j. williams and j. p. carbotte , phys . b * 39 * , 2180 ( 1989 ) ; c. t. rieck , d. fay , and l. tewordt , phys . rev . b * 41 * , 7289 ( 1990 ) ; h. chi and j. p. carbotte , phys . rev . b * 49 * , 6143 ( 1994 ) ; t. moriya , y. takahashi , and k. ueda , j. phys . soc . jpn . * 59 * , 2905 ( 1990 ) ; a. abanov , a. chubukov and m.r . norman , phys . b*78 * , 220507(r ) ( 2008 ) . p. monthoux and g. g. lonzarich , phys . b * 59 * , 14598 ( 1999 ) . p. monthoux and g. g. lonzarich , phys . b * 63 * , 054529 ( 2001 ) . j. a. hertz , phys . rev . b * 14 * , 1165 ( 1976 ) . a. j. millis , phys . b * 48 * , 7183 ( 1993 ) . t. moriya and k. ueda , adv . phys . * 49 * , 555 ( 2000 ) ; t. moriya and k. ueda , rep . * 66 * , 1299 ( 2003 ) ; t. moriya and t. takimoto , j. phys . 64 * , 960 ( 1995 ) . t. moriya , _ spin fluctuations in itinerant electron magnetism _ ( springer - verlag , 1985 ) . h. shishido , r. settai , d. aoki , s. ikeda , h. nakawaki , n. nakamura , t. iizuka , y. inada , k. sugiyama , t. takeuchi , k. kindo , t. c. kobayashi , y. haga , h. harima , y. aoki , t. namiki , h. sato , and y. @xmath198nuki , j. phys . jpn . * 64 * , 960 ( 1995 ) . m. yashima , h. mukuda , y. kitaoka , h. shishido , r. settai , and y. @xmath198nuki , phys . b * 79 * , 214528 ( 2009 ) . c. stock , c. broholm , j. hudis , h. j. kang , and c. prtrovic , phys . rev . lett . * 100 * , 087001 ( 2008 ) ; note that the analysis in this reference is pointed out in ref . p. gegenwart , c. langhammer , c. geibel , r. helfrich , m. lang , g. sparn , and f. steglich , phys . lett . * 81 * , 1501 ( 1998 ) . h. schneider , z. kletowski , f. oster , and d. wohlleben , solid state commun . * 48 * , 1093 ( 1983 ) . a. schroeder et al . , phys . 80 ( 1998 ) 5623 c.m . varma et al . , phys . 63 , 1996 ( 1989 ) ; v. aji and c.m . varma , phys . b 79 , 18450 ( 2009 ) . w. l. mcmillan , phys . 167 , 331 ( 1968 ) . p. b. allen and r. c. dynes , phys . b * 12 * , 905 ( 1975 ) . see , for example , h.v . lohneysen , a. rosch , m. vojta and p. woelfle , rev . 79 , 1015 ( 2007 ) . r. hlubina and t. m. rice , phys . b * 51 * , 9253 ( 1995 ) .

anderson localization @xcite of electronic states in a disordered system is a path breaking observation that has extended its realm well beyond the electronic properties of disordered solid materials , and has been found out to be ubiquitous in a wide variety of systems such as the photonic @xcite , phononic @xcite , polaronic @xcite , or plasmonic lattices @xcite to name a few . artificial , tailor made geometries developed using the improved fabrication and lithographic techniques have been instrumental in testing the validity of the basic theory of exciton localization in presence of disorder . examples include the direct observation of localization of matter waves in recent times @xcite . localization is a result of quantum interference that has the strongest manifestation in one dimension where all the single particle states exhibit an exponential decay of amplitude as one moves away from a particular lattice point . for dimensions beyond one , the states retain their characteristic decay in amplitude , with the addition of a possibility of a metal - insulator transition in three dimensions . the results have been substantiated by calculations of the localization length @xcite , density of states @xcite and their multi - fractality @xcite . extensive work has also revealed the intricacies of the single parameter scaling hypothesis @xcite , its validity @xcite or , variance @xcite , and even violation @xcite in low dimensional systems within a tight binding approximation . the studies have subsequently been consolidated by direct experimental measurements of conductance distribution in quasi - one dimensional gold wires @xcite . in the last couple of decades variations in the canonical cases of anderson localization have surfaced , particularly , within a tight binding description . resonant tunnelling of electronic states has been shown to arise even in a one dimensional disordered system as a consequence of certain special kinds of positional correlation @xcite . discrete energy levels corresponding to _ extended eigenfunctions _ @xcite , appear within a typical point spectrum exhibited by the randomly disordered lattices rendering the two - terminal transport completely ballistic at special values of the fermi level . such results are common even to infinite quasicrystalline lattices in one dimension where , the self similar character of the chain really allows an infinite number of unscattered ( extended ) eigenfunctions despite the absence of any translational invariance @xcite . in this context , rather unusual , yet meaningful question could be , is it possible to engineer , in a controlled fashion , an _ absolutely continuous _ portion in the energy spectrum of non - translationally invariant systems with the help of some kind positional correlations ? the existence of real continuous bands are reported recently in quasi one dimensional or two dimensional systems only with diagonal disorder @xcite . we thus have a partial answer to the question , and it remains to be seen whether such bands of extended eigenfunctions can be generated and controlled with off - diagonal disorder are well . we address this issue with respect to quasiperiodically ordered geometries . in this communication we show that , in quasi - one dimensional ladder networks with off - diagonal quasi - periodic order it is possible to engineer absolutely continuous bands of energy eigenvalues . we specifically work with the fibonacci quasiperiodic order , though the observation is by no means , restricted to them . ladder networks within the framework of tight binding model have already been exploited to gain insight into charge transport in dna @xcite , to unravel interesting physics related to polymer structures and ployacene lattice topology @xcite . comparatively speaking , controlled generation of absolutely continuous portions in the energy spectrum of the ladder networks , disordered or quasi - periodic , still remains to be explored . if it works then such an effort may open up new physics , or allow one to explore the possibility of devising novel low dimensional filter networks . the central result of this paper is that , one can _ create _ absolutely continuous subbands in the electronic spectrum of such quasi 1-d systems , by attaching a single quantum dot ( qd ) from one side to a selected set of lattice points in every strand of the ladder . we discuss the results explicitly in terms of fibonacci quasi - periodic two and three arm ladder network models using a tight binding hamiltonian and a real space renormalization group ( rsrg ) method . we also investigate whether a crossover in the fundamental character of the eigenfunctions , going from a typical _ critical _ nature to a bloch - like _ extended _ character can be engineered . as will be shown , this indeed turns out to be quite a possibility in the case of the quasi - periodic ladder network ( qpln ) , particularly when the number of strands ` @xmath0 ' in the transverse direction increases . the results hold even for an off - diagonally disordered ladder where a re - entrant insulator - metal - insulator transition becomes feasible . the system is described by a tight binding hamiltonian , and is depicted in fig . [ lattice1 ] . the on - site potential at every lattice point is kept constant , while the nearest neighbor hopping integral along the arms of the ladder assumes two values dictated by the growth rule of the desired quasi - periodic sequence . the electron can tunnel between the arms of the ladder by hopping from one site to its corresponding site on the other arm through the hopping integral @xmath1 ( see fig . [ lattice1 ] ) . the sites marked by @xmath2 in each arm of the ladder is coupled to a single level qd @xcite attached from one side . the qd couples to every @xmath2-site via a hopping integral @xmath3 . there is no positional correlation , short range or long range in the conventional sense @xcite . attachment of qd s from a side can quench the conductance of a linear chain @xcite , but an array of qd s in a ladder can have reverse effect , as will be seen . we will show in a completely analytical way that , for a particular relationship between the nearest neighbor hopping integrals along the ladder strand and the qd - strand coupling @xmath3 , and for a special value of the qd potential , we get _ absolutely continuous subbands _ in the spectrum . this immediately raises a non - trivial issue that , the nature of the electronic spectrum and the eigenstates over a _ continuous range of eigenvalues _ can become sensitive to the _ numerical values _ of the hamiltonian parameters a fact that is not commonly observed in the conventional anderson localization problem . the full spectrum turns out to be an admixture of conventional _ critical eigenstates _ , characteristics of a fibonacci quasiperiodic chain , and absolutely continuous parts , giving rise to the possibility of a _ phase transition _ , particularly when the system grows in the transverse direction . our results are analytically exact . in section [ sec2 ] we describe the hamiltonian and the basic working principle for the two and three strand quasi - periodic ladders constructed following a fibonacci sequence . along with this , the real space renormalization scheme to obtain the energy spectrum and to judge the character of the single particle states is also presented . in section [ sec3 ] we present a result for the two terminal transmission coefficient of a finite sized qpln , and in section [ sec4 ] we draw our conclusion . in wannier basis the tight binding hamiltonian of a general @xmath0-arm ladder network reads @xcite , @xmath4\ ] ] [ hamilton ] where , @xmath5 is , in general , an @xmath6 on - site potential matrix at the @xmath7-th vertical rung . we shall explicitly discuss the two - strand ladder at first . the cases @xmath8 can be trivially constructed . @xmath9 is the creation ( annihilation ) operator , represented by rows ( columns ) of appropriate dimension , and @xmath10 is the @xmath6 hopping matrix representing hopping along and in between strands of the ladder , @xmath0 being equal to @xmath11 , @xmath12 or any integer . the fibonacci ladder network is grown by attaching identical one dimensional fibonacci chains in the transverse direction . each individual chain grows along the horizontal direction following the binary fibonacci sequence of two letters @xmath13 and @xmath14 ( representing two bonds ) . the growth rule is , @xmath15 and @xmath16 , and the sequence begins with @xmath13 . the on - site potentials at the vertices in each strand are designated by three symbols viz . , @xmath17 , @xmath18 and @xmath19 for sites flanked by two @xmath13-bonds , @xmath13 on left and @xmath14 on right , and @xmath13 on right and @xmath14 on left respectively . the attached qd is shown as @xmath20 " in fig . [ lattice1 ] , and the potential at the side coupled qd is designated by @xmath21 . the nearest neighbor hopping integral @xmath22 or @xmath23 along ladder - arms following the fibonacci sequence . @xmath24 is the inter - arm hopping , and @xmath25 designates the tunnel hopping between the qd and the @xmath2-sites of the ladder network . in our work , we shall assume @xmath26 , and @xmath27 . so , this is purely a _ transfer model _ @xcite in the standard language of one dimensional quasicrystals . we shall analyze the problem using a method described earlier @xcite in which an ` @xmath0'-strand ladder ( @xmath28 in the first phase of our discussion ) is decoupled , making a change of basis , into a system of @xmath0 independent linear chains . the complete spectrum of the quasi @xmath29-d system of @xmath0 strands is then obtained by convolving the spectra of the individual linear chains . in addition , we perform an rsrg analysis , which does not need any decoupling and yields a density of states spectrum that is in conformity with our results obtained using the decoupling scheme . the two terminal transmission spectrum is also computed using the rsrg decimation scheme to highlight the effect of the side coupled qd s . the analysis will be carried over to the three strand ladder , and a generalization to the @xmath0 strand case will be discussed using arguments . at first , we renormalize the potential at the @xmath2-sites by decimating the side coupled qd . this leads to a _ renormalized _ value of the on - site potential at the @xmath2-vertex , given by , @xmath30 . , @xmath31 and @xmath32 are colored differently to distinguish between the nearest neighbor bond - environments . the qd s @xmath20 are coupled to the @xmath2-sites in each case . the coupling between @xmath20 and @xmath2-site is denoted by @xmath3 and the inter - strand coupling is denoted by @xmath1.,width=302 ] the schrdinger equation for the two - arm ladder can now be written equivalently in the form of the difference equations involving @xmath33 matrices : @xmath34 where , @xmath35 and , @xmath36 and , @xmath37 is the @xmath33 unit matrix . written explicitly , the matrix equation reads , @xmath38 \left ( \begin{array}{c } \psi_{i , a } \\ \psi_{i , b } \end{array } \right ) = \left ( \begin{array}{cccc } t_{i , i+1}^a & 0 \\ 0 & t_{i , i+1}^b \end{array } \right ) \left ( \begin{array}{c } \psi_{i+1,a } \\ \psi_{i+1,b } \end{array } \right ) + \left ( \begin{array}{cccc } t_{i , i-1}^a & 0 \\ 0 & t_{i , i-1}^b \end{array } \right ) \left ( \begin{array}{c } \psi_{i-1,a } \\ \psi_{i-1,b } \end{array } \right ) \label{differenceeq}\ ] ] in eq . , the potential @xmath39 takes on values @xmath40 corresponding to a renormalized @xmath2-site in each arm , while it is equal to @xmath41 for both the @xmath31 and @xmath32-sites in each arm . the nearest neighbor hopping integrals @xmath42 are @xmath43 or @xmath23 depending on the local environment of the corresponding site in the arm @xmath44 . a look at the fig . [ lattice1 ] will make it obvious that there are three such equations corresponding to the three different kinds of vertices viz . , @xmath2 , @xmath31 and @xmath32 respectively . the potential matrix @xmath5 corresponding to the renormalized @xmath2-sites is , @xmath45 and that for the @xmath31 and @xmath32 sites are @xmath46 we now diagonalize the @xmath47 matrix using a matrix @xmath48 . the matrix @xmath48 makes a change over to a new basis defined by , @xmath49 . as shown elsewhere @xcite , such a change of basis leads to a decoupling of the set of eq . into two independent equations , one corresponding to the @xmath50-arm alone , and the other , to the arm @xmath51 . written explicitly , the equations read , @xmath52 \phi_{i , a } = t_{l } \phi_{i+1,a } + t_{l } \phi_{i-1,a } \nonumber \\ & \left [ e - ( \epsilon_{0}+\xi ) \right ] \phi_{i , a } = t_{s } \phi_{i+1,a } + t_{l } \phi_{i-1,a } \nonumber \\ & \left [ e - ( \epsilon_{0}+\xi ) \right ] \phi_{i , a } = t_{l } \phi_{i+1,a } + t_{s } \phi_{i-1,a } \label{arm1}\end{aligned}\ ] ] for the sites @xmath2 , @xmath31 and @xmath32 sequentially in arm @xmath50 . similar set of equations for @xmath2 , @xmath31 and @xmath32 sites in arm @xmath51 will be , @xmath53 \phi_{i , b } = t_{l } \phi_{i+1,b } + t_{l } \phi_{i-1,b } \nonumber \\ & \left [ e - ( \epsilon_{0}-\xi ) \right ] \phi_{i , b } = t_{s } \phi_{i+1,b } + t_{l } \phi_{i-1,b } \nonumber \\ & \left [ e - ( \epsilon_{0}-\xi ) \right ] \phi_{i , b } = t_{l } \phi_{i+1,b } + t_{s } \phi_{i-1,b } \label{arm2}\end{aligned}\ ] ] needless to say , each of the sets eq . and eq . separately represents an independent fibonacci chain . the spectrum of each chain bears the usual singular continuous character . the spectrum of the original ladder network can be obtained by convolution of the density of states of the individual decoupled fibonacci chains . however , an analysis of these decoupled sub - systems yields rich spectral insight that would otherwise be difficult to obtain . we are now in a position to throw some light on this issue . first , we need to appreciate that , in the new basis each decoupled fibonacci chain is described by a set of three _ transfer matrices _ , viz . , @xmath54 , @xmath55 and @xmath56 @xcite . for the @xmath50-arm decoupled fibonacci chain , the explicit expressions of these matrices are , @xmath57/t_{l } & -1 \\ 1 & 0 \end{array } \right ) \nonumber \\ \mbox{\boldmath $ m_{\beta , a}$ } & = & \left ( \def\arraystretch{1.5}\begin{array}{cccc } [ e-(\epsilon_{0}+\xi)]/t_{s } & -t_{l}/t_{s } \\ 1 & 0 \end{array } \right ) \nonumber \\ \mbox{\boldmath $ m_{\gamma , a}$ } & = & \left ( \def\arraystretch{1.5}\begin{array}{cccc } [ e-(\epsilon_{0}+\xi)]/t_{l } & -t_{s}/t_{l } \\ 1 & 0 \end{array } \right ) \label{arma}\end{aligned}\ ] ] respectively . a similar set of transfer matrices for the decoupled @xmath51-arm reads , @xmath58/t_{l } & -1 \\ 1 & 0 \end{array } \right ) \nonumber \\ \mbox{\boldmath $ m_{\beta , b}$ } & = & \left ( \def\arraystretch{1.5}\begin{array}{cccc } [ e-(\epsilon_{0}-\xi)]/t_{s } & -t_{l}/t_{s } \\ 1 & 0 \end{array } \right ) \nonumber \\ \mbox{\boldmath $ m_{\gamma , b}$ } & = & \left ( \def\arraystretch{1.5}\begin{array}{cccc } [ e-(\epsilon_{0}-\xi)]/t_{l } & -t_{s}/t_{l } \\ 1 & 0 \end{array } \right ) \label{armb}\end{aligned}\ ] ] @xmath54 and @xmath59 follow an arrangement in the fibonacci sequence @xcite . speaking in terms of the individual decoupled fibonacci chains @xmath50 and @xmath51 , we already know that each such chain offers a singular continuous spectrum . if we set @xmath60 , then each chain describes a separate fibonacci lattice with @xmath61 , and @xmath27 . the total transfer matrix for any @xmath62-th generation of such a chain is related to its predecessors at the @xmath63th and @xmath64th generations by the matrix map @xmath65 , which will exhibit a _ six cycle _ @xcite at @xmath66 for the decoupled @xmath50 and @xmath51 chains respectively . the existence of a six cycle of the matrix map implies that the above energy values are indeed eigenvalues of the infinite decoupled chains , and hence , the infinite ladder network as well . thus , the first fruit of the decoupling scheme is that , with @xmath67 we obtain at least two eigenvalues which definitely belong to the spectrum of the infinite ladder . extending this idea to an @xmath0-strand fibonacci ladder we can easily identify @xmath0 different energy eigenvalues of the full quasi @xmath29-d fibonacci network in the absence of @xmath3 ( i.e. detaching the qd s ) in a completely analytical way . the task becomes non - trivial when one consider the qpln as a whole , without decoupling . now let us look at the more serious effect when we include the qd by making @xmath3 non - zero . let us choose eq . . we set @xmath69 . it can easily be verified that , with this potential , the commutator @xmath70 = 0 $ ] for the decoupled @xmath50-arm , _ irrespective of the energy @xmath71 _ of the electron , if @xmath72 . this implies that , with these conditions the individual @xmath2-sites and the @xmath73 dimers in the decoupled @xmath50-arm can be arranged in any desired pattern , for example , in a perfectly periodic pattern . the spectrum offered by the decoupled @xmath50-arm fibonacci chain will thus , under this choice of the qd potential @xmath21 and the tunnel hopping @xmath3 , will turn out to be _ indistinguishable _ from that of a perfectly periodic sequence of the @xmath2-site ( the ` renormalized ' @xmath2-site , to be precise ) and the @xmath73 doublet . however , this argument needs to go through an acid test which we discuss below . + qd chain , and its renormalized version . ( b ) the @xmath73 binary ordered chain . the ldgf is found at the site on the horizontal line in each case . the qd is renormalized to produce an effective one dimensional periodic chain of @xmath74 sites , as already discussed in the text.,width=264 ] owing to the commutation of @xmath54 and @xmath75 , the decoupled @xmath50-arm can be thought to be composed of two infinite periodic chains , one with @xmath2-sites coupled to the qd s , and the other , a periodic array of @xmath73 clusters . the two chains are shown in fig . [ chains ] . each such infinite sublattice will have its own energy bands . it is important to verify whether , under the special condition imposed on the strand - dot coupling @xmath3 , the periodic @xmath2-lattice and the periodic @xmath73 lattice share the same eigenvalue spectrum . if yes , then we conclusively prove the existence of an absolutely continuous part in the energy spectrum of the original ladder network . to check this , we first of all construct an effective one dimensional chain composed of renormalized @xmath2-atoms , obtained by decimating the side coupled dot . the resulting chain is illustrated in fig . [ chains](a ) with encircled lattice points . each such lattice point has the renormalized on - site potential equal to @xmath76 . each atom in the @xmath73 chain has the potential @xmath77 . we now evaluate the local density of states ( ldos ) at any site of the renormalized ( encircled ) @xmath2-chain ( let s call it the @xmath74 chain ) and any site in the periodic @xmath73 chain ( fig . [ chains](b ) ) . the local densities of states are given , for these two infinite periodic lattices , and for a given set values of @xmath78 , @xmath1 , @xmath43 and @xmath23 , by @xmath79^{-1/2}$ ] and @xmath80^{-1/2}$ ] , where , @xmath81 and , @xmath82 if we now set @xmath83 , then a simple algebra yields that , the difference @xmath84 is given by , @xmath85 ( \lambda^2 + t_{l}^2 - t_{s}^2 ) } { ( e - \epsilon_{0 } - \xi ) ^2 } \label{den}\ ] ] which clearly vanishes for any value of the energy of the electron the moment we tune @xmath72 . thus the ldos s of the two basic constituent periodic chains become identical _ independent of energy _ , resulting in a complete overlap of the spectra of these individual chains under the above resonance condition . as each of the @xmath2 and the @xmath73 sublattices provide a two subband continuum in their energy spectra ( an obvious effect of the binary periodic lattice ) , it is obvious that these two spectral continua will retain their positions in the full spectrum of the fibonacci qpln when the qd potential and the qd - strand tunnel coupling are tuned to the above special values . it should be appreciated that once we fix @xmath86 , the set of eq . corresponding to the decoupled @xmath51 strand , represents a mixed model linear fibonacci chain . its contribution to the full energy spectrum is the usual fragmented , multi - fractal one . as the full spectrum of the two - arm ladder will be a convolution of both , we expect it to have absolutely continuous parts decorated with a spiky envelope , and a mixture of _ extended _ bloch - like states and the _ critical _ eigenfunctions which are characteristic of a quasi - periodic fibonacci sequence of potentials . this is precisely what we observe in fig . [ 2stranddos](a ) . at this point one must be careful to interpret fig . [ 2stranddos](a ) . in the full energy spectrum , critical and extended eigenstates can never coexist . so , even in the full spectrum of the un - decoupled qpln the entire range shown in fig . [ 2stranddos](a ) by grey shade corresponds to extended bloch - like states only . beyond it , every wave function will be critical in nature . the contributions coming from each decoupled strand are only shown separately in this figure . the above arguments are confirmed by an exact real space renormalization group ( rsrg ) calculation of the local density of states , first of each individual decoupled strands and then of the full two - arm fibonacci ladder without decoupling it . the first task is already discussed in literature @xcite , and need not be talked about again in here . we instead , provide the details of the rsrg scheme of the full ladder network . we renormalize the full two - strand ladder by decimating the @xmath31-vertices . this generates _ new _ hopping integrals across the diagonals of the fundamental rectangular plaquettes of renormalized ladder denoted by @xmath87 and @xmath88 in the equations to follow . to implement the decimation we need to assign three different symbols to the inter - arm hopping @xmath1 , viz . , @xmath89 , @xmath90 and @xmath91 connecting the @xmath2 , @xmath31 and @xmath32 vertices of the two strands . of course , we begin with @xmath92 , and @xmath93 . the recursion relations arising out of this decimation procedure are given by , @xmath94 where @xmath95 a small imaginary part is added to the energy @xmath71 , and the ldos at an @xmath2 , @xmath31 or @xmath32 site is obtained by calculating the respective local green s function @xmath96 , where @xmath97 or @xmath32 . @xmath98 represents the fixed point of the respective on - site potential as the hopping integrals flow to zero under the rsrg iterations . the ldos is given by , @xmath99\ ] ] the results are presented in fig . [ 2stranddos ] . in panel ( a ) the spectra of the decoupled @xmath50- and @xmath51-arms are shown separately , while panel ( b ) illustrates the spectrum of the full two - arm ladder as obtained from the rsrg study . the existence of the absolutely continuous portions in the energy spectrum is obvious . the character of the eigenfunctions belonging to the spectrum is checked from an observation of the flow of the the hopping integrals @xmath43 , @xmath23 and the diagonal hopping amplitude under rsrg iterations . for any energy eigenvalue , chosen from any portion of the absolutely continuous sub - bands ( grey shaded portion in fig . [ 2stranddos](a ) ) all the hopping integrals , as obtained from eq . , remain non - zero for an arbitrarily large number of iterations confirming the extended character of the wave functions . beyond the extended regime , the hopping integrals flow to zero under iteration indicating that the eigenfunctions are not extended . in fig . [ 2stranddos](b ) we see the full ldos spectrum from the coupled two - strand qpln obtained from the rsrg method ( using the set of eq . ) . the light blue shade in the spectrum is just to highlight the region of non - zero ldos . the extended eigenstates occupy the precise positions as depicted by the grey shaded part in fig . [ 2stranddos](a ) . the eigenstates form a mixed spectrum of critical and extended states , as one traverses the energy axis from the left . -vertex for a two - strand ladder network . the upper panel shows the ldos spectra of the decoupled individual linear chains , one chain satisfying resonance condition . the lower panel shows the ldos spectrum of the composite ladder network as obtained from the rsrg scheme , under resonance condition . the values of the parameters chosen are @xmath100 , @xmath101 , @xmath102 , @xmath103 , @xmath104 , and @xmath105 , measured in unit of @xmath43.,width=264 ] an additional support in regard of the continuous subbands and the extended eigenstates is also obtained from an observation of the flow pattern of the entire parameter space of the decoupled strands that satisfies the condition @xmath106 and @xmath107 . for such a choice , under successive rsrg step , we always get @xmath108 and @xmath109 for the decoupled @xmath50-strand . this is a typical flow in the parameter space that corresponds to the extended eigenstates in quasiperiodic fibonacci chains @xcite . in the present case it confirms the extended bloch - like character of the wave function for the entire energy regime ( given by the grey color in fig . [ 2stranddos](a ) ) . the motivation behind extending the above analysis to a three - arm fibonacci ladder stems .,width=302 ] from the curiosity to know whether , as one extends the ladder network in the transverse direction , any possibility of a re - entrant transition ( or , crossover ) from a singular continuous to an absolutely continuous energy spectrum becomes apparent . for disordered ladders comprising the same @xmath2 and @xmath73 clusters this would imply the possibility of an _ insulator - metal _ transition that can be triggered by an appropriate choice of the numerical values of the qd potential @xmath21 and the tunnel hopping integral @xmath3 . to check such a possibility we decouple a three - arm fibonacci ladder ( fig . [ lattice2 ] ) . we follow the same prescription for decoupling as outlined in the case of a two - strand ladder . diagonalization of the potential matrix @xmath5 now yields three eigenvalues , viz . , @xmath41 and @xmath110 , and the effective potential at the @xmath2-site in each decoupled arm will be @xmath111 ( for arm @xmath50 , say ) , @xmath112 ( for arm @xmath51 , say ) and @xmath113 ( for arm @xmath114 ) . we now end up having three sets of equations , one for each decoupled strand . for example , @xmath115 \phi_{i , a } = t_{l } \phi_{i+1,a } + t_{l } \phi_{i-1,a } \nonumber \\ & ( e - \epsilon_{0 } ) \phi_{i , a } = t_{s } \phi_{i+1,a } + t_{l } \phi_{i-1,a } \nonumber \\ & ( e - \epsilon_{0 } ) \phi_{i , a } = t_{l } \phi_{i+1,a } + t_{s } \phi_{i-1,a } \label{3decouple1}\end{aligned}\ ] ] corresponding to the @xmath2 , @xmath31 and @xmath32 sites respectively in the decoupled @xmath50-strand . two sets of similar equations are also obtained for the decoupled @xmath51- and @xmath114-strands , with the effective potential at the @xmath2-site being equal to @xmath116 and @xmath117 respectively . the effective potentials at the @xmath31 and @xmath32 sites in the @xmath51 and @xmath114 strand are @xmath110 as there is no side coupling to these sites . if , as before , we set @xmath118 in eq . , then @xmath70=0 $ ] for the decoupled @xmath50-strand . the contribution to the full spectrum coming from this separate chain will be two absolutely continuous sub - bands . with the potential of the qd set at @xmath118 , the two other decoupled strands represent two different fibonacci chains in the mixed model . the spectrum resulting from each of them will remain fragmented cantor type . the overall spectrum , as we now have known , will be a convolution of the three . the absolutely continuous portions will still hold their positions in the main spectrum , and will be flanked by the spiky , fragmented parts characteristics of critical eigenstates of a fibonacci lattice ( fig [ 3stranddos ] ) . the eigenfunctions belonging to the continuous sub - bands will , needless to say , be of extended character . .,width=264 ] one needs to appreciate that each spectrum offered separately by the decoupled @xmath51- and @xmath114-strands will have different band - centres " around which the singular continuous spectra of the individual decoupled fibonacci chain will spread out . with an @xmath0-strand ladder model , there will be @xmath119 such centers from the @xmath119 decoupled strands around which cantor set energy eigenvalues will be distributed , the remaining strand contributing the absolutely continuous sub - bands . the centres will be densely populated around the outer edge of the continuous sub - bands as @xmath120 . thus , with increasing number of strands there is a possibility of observing a smooth critical to extended state crossover in the spectrum of such quasi @xmath29-d aperiodic ladder models . clearly , the argument prevails even if we have an @xmath0-strand off diagonally disordered ladder network . the side coupled qd s with appropriately tuned potential one can then expect a re - entrant _ insulator - metal _ transition . before leaving this discussion , it is pertinent to mention that , the commutaion of transfer matrices leading to a complete , energy independent delocalization of electronic states have been reported recently in certain special examples of quasi - one dimensional lattices @xcite . but , those systems do not open up any possibility of an insulator - metal transition , in the current spirit of the work . at the same time , it is useful to note that , the commutivity of transfer matrices has been used in the past to unravel discrete unscattered eigenstates in an array of quantum wells @xcite or a disordered distribution of pschl - teller potential @xcite . the difference with the present communication and these two latter references lies in the energy - independent character of the matrix - commutation that plays the key role . we now present partial results ( to save space ) for the two terminal transmission coefficient of a finite two strand ladder without decoupling it . no change of basis is made , and the full finite qpln is treated within the rsrg scheme . the basic working formula is already elaborated in the existing literature @xcite . the ladder is clamped between two semi - infinite perfectly ordered leads ( viz . , the source and the drain ) at any two extremities ( fig . [ lattice3 ] ) . the leads are characterized by uniform on - site potential @xmath121 and nearest neighbor hopping integral @xmath122 . the portion of the ladder clamped between the leads is renormalized to bring it down to a _ diatomic molecule _ , and the transmission coefficient is easily obtained @xcite . the procedure is standard and to save space again , we do not provide the details . it should be appreciated that , the precise details of the transmission coefficient definitely depends on which points in the ladder the leads are attached to . only one result is shown here in which the two semi - infinite ordered leads are connected to the extreme points of the upper strand of a @xmath123-th generation fibonacci ladder having @xmath124 bonds in each arm . -th generation two - strand fibonacci ladder network under resonance condition . both the incoming and outgoing leads , viz . , the source and the drain are connected in the upper strand . system parameters chosen are @xmath100 , @xmath101 , @xmath102 , @xmath103 , @xmath104 , and @xmath105 , and the lead parameters are @xmath121 and @xmath122.,width=264 ] one needs to appreciate that , even in the absence of any attached qd , a two strand finite qpln can exhibit high transmission spikes with @xmath125 close to unity . this normally results from the finite size of the system , as well as from the fact that a ladder network consists of loops that brings in a flavor of long range connection , and may well be responsible for quantum interference leading to high transmittivity at selected values of electron - energy . also , the attachment to the leads , and the difference between the hopping integral in the lead and that in the bulk of the system give rise to strong scattering effects at the contacts , affecting the overall transmission of the qpln system . therefore , the results obtained from a finite size analysis should be carefully compared with the predictions which come from the anatomy of an infinite system . nevertheless , the existence of the continuous parts in the energy spectrum of the infinite system is already showing up in the enhanced value of the transmission coefficient at appropriate values of the electron energy as can be seen in fig . [ transport ] . the apparent drop to close to zero value at certain regions even within the continuous zone is due to the scale at which the transmission coefficient is shown here . however , the transmission coefficient exhibits wide fluctuations owing to the quasiperiodic nature of network . the effect of quantum interference is manifestly observed here . the finite value of the transmission coefficient even beyond the absolutely continuous part of the spectrum is due to the finite size of the system , and gets lower and lower , and the spectrum gets more spiky with increasing system size . also , one can anticipate eigenstates with their ` localization length ' much larger compared to the size of the system as considered in the present calculation . we have shown that fibonacci quasi - periodic ladder networks , where quasi - periodic order is introduced in the distribution of bonds , keeping the potential at every vertex same , are capable of producing absolutely continuous portions in their energy spectra . this requires the attachment of quantum dots from one side to a special set of sites in each arm of the ladder . a suitable adjustment of the dot potential , together with the dot - backbone coupling can generate absolutely continuous energy sub - bands , decorated at the flanks by the usual critical states a characteristic of quasi - periodic lattices . apart from opening up the possibility of engineering the transport of excitons in such multi - strand systems , a very basic issue of anderson localization is reviewed . it seems that , the existence of localized eigenstates in similar disordered ladder networks may strongly depend on the numerical values of the parameters in the hamiltonian . the analysis is carried over to other quasi - periodically ordered systems , such as the copper mean chain , the aperiodic thue morse chain and even to the disordered ladder networks with locally and non - locally coupled dots . the conditions for the existence of a continuum in the energy spectrum becomes much more non - trivial in such systems . these issues will be reported in a forthcoming article . is grateful to dst , india for an inspire fellowship . a.c . acknowledges financial support from dst , india through a purse grant , and the abdus salam international centre for theoretical physics , trieste , italy for its hospitality and financial support . 60 p. w. anderson , phys . * 109 * , 1492 ( 1958 ) . b. kramer and a. mackinnon , rep . * 56 * , 1469 ( 1993 ) . e. abrahams , p. w. anderson , d. c. liciardello , and t. v. ramakrishnan , phys . * 42 * , 673 ( 1979 ) . e. yablonovitch , phys . * 58 * , 2059 ( 1987 ) . s. john , phys . lett . * 58 * , 2486 ( 1987 ) . f. r. montero de espinosa , e. jimnez , and m. torres , phys . 80 * , 1208 ( 1998 ) . j. o. vasseur , p. a. deymier , g. frantziskonis , g. hong , b. djafari - rouhani , and l. dobrzynski , j. phys . : condens . matter * 10 * , 6051 ( 1998 ) . i. o. barinov , a. p. alodzhants , and s. m. arakelyan , quantum electron . * 39 * , 685 ( 2009 ) . m. grochol and c. piermarocchi , phys . b * 78 * , 035323 ( 2008 ) . a. tao , p. sinsermsuksaul , and p. yang , nature nanotechnol . * 2 * , 435 ( 2007 ) . a. christ , y. ekinci , h. h. solak , n. a. gippus , s. g. tikhodeev , and o. j. f. martin , phys . b * 76 * , 201405 ( 2007 ) . b. damski _ lett . * 91 * , 080403 ( 2003 ) . et al . _ , nature ( london ) * 453 * , 891 ( 2008 ) . et al . _ , nature ( london ) * 453 * , 895 ( 2008 ) . r. a. rmer and h. schulz - baldes , europhys . lett . * 68 * , 247 ( 2004 ) . a. eilmes , r. a. rmer , and m. schreiber , physica b * 296 * , 46 ( 2001 ) . a. rodrguez , j. phys . a : math . gen . * 39 * , 14303 ( 2006 ) . a. rodriguez , l. j. vasquez , and r. a. rmer , phys . b * 78 * , 195107 ( 2008 ) . a. rodriguez , l. j. vasquez , k. slevin , and r. a. rmer , phys . rev . b * 84 * , 134209 ( 2011 ) . s. d. pinski , w. schirmacher , and r. a. rmer , europhys . lett . * 97 * , 16007 ( 2012 ) . lev i. deych , a. a. lisyanski , and b. l. alshuler , phys . lett . * 84 * , 2678 ( 2000 ) . m. titov and h. schomerus , phys . lett . * 95 * , 126603 ( 2005 ) . j. w. kantelhardt and a. bunde , phys . b * 66 * , 035118 ( 2002 ) . v. e. kravtsov and v. i. yudson , phys . b * 82 * , 195120 ( 2010 ) . p. mohanty and r. a. webb , phys . lett . * 88 * , 146601 ( 2002 ) . d. h. dunlap , h .- wu , and p. phillips , phys . lett . * 65 * , 88 ( 1990 ) . f. a. b. f. de moura and m. lyra , phys . 81 * , 3735 ( 1998 ) . s. sil , s. k. maiti , and a. chakrabarti , phys . b * 78 * , 113103 ( 2008 ) . a. rodriguez , a. chakrabarti , and r. a. rmer , phys . b * 86 * , 085119 ( 2012 ) . a. chakrabarti , s. n. karmakar , and r. k. moitra , phys . b * 50 * , 13276 ( 1994 ) . e. maci , phys . b * 74 * , 245105 ( 2006 ) ; phys . rev . b * 80 * , 125102 ( 2009 ) . m. bravi , r. farchioni , g. grosso , and g. p. parravicini , phys . b * 87 * , 035105 ( 2013 ) . r. farchioni , g. grosso , and g. p. parravicini , eur . j. b * 84 * , 227 ( 2011 ) . b. kubala and j. knig , phys . b * 67 * , 205303 ( 2003 ) . r. farchioni , g. grosso , and g. p. parravicini , phys . b * 85 * , 165115 ( 2012 ) . m. kohmoto , b. sutherland , and c. tang , phys . b * 35 * , 1020 ( 1987 ) . j. a. ashraff and r. b. stinchcombe , phys . rev . b * 37 * , 5723 ( 1988 ) . a. chakrabarti , s. n. karmakar , and r. k. moitra , phys . b * 39 * , 9730 ( 1989 ) . a. chakrabarti , s. n. karmakar , and r. k. moitra , phys . lett . a * 168 * , 301 ( 1992 ) . b. pal , s. k. maiti , and a. chakrabarti , europhys . lett . * 102 * , 17004 ( 2013 ) . i. gmez , f. dminguez - adame , and e. diez , physica b * 324 * , 235 ( 2002 ) . a. rodrguez and j. m. cerver , phys . b * 74 * , 104201 ( 2006 ) . a. d. stone , j. d. joannopoulos , and d. j. chadi , phys . b * 24 * , 5583 ( 1981 ) . a. chakrabarti , phys . b * 74 * , 205315 ( 2006 ) . b. pal and a. chakrabarti , eur . j. b * 85 * , 307 ( 2012 ) .

the electron transport via resonant tunnelling in mesoscopic systems has been the subject of extensive theoretical research due to recent development in fabrication of small electronic devices and their potential applications . some interest has been focused on the transport properties of a quantum dot ( qd ) under the influence of external time - dependent fields . new effects have been observed and theoretically described , e.g. , photon - assisted tunnelling through small quantum dots , @xcite , photon - electron pumps @xcite and others . in most theoretical investigations a qd placed between two leads was considered ( e.g. , refs . @xcite ) and the current flowing through a qd under periodic modulation of the qd electronic structure or periodic ( nonperiodic ) modulation of the tunnelling barriers and electron energy levels of both ( left and right ) leads was calculated . one of the important problems of the mesoscopic physics is the interference of the charge carries . this interference appears when two ( or more ) transmission channels for electron tunnelling exist . such possibility exists , e.g. in the electron transport through a qd embedded in a ring in the aharonov - bohm geometry and much theoretical interest has been paid to description of the phase coherence in this and related systems , e.g. refs . @xcite . another experimental situation in which the interference may occur can be realized in the scanning tunnelling microscope ( stm ) . the recent experimental and theoretical studies with a low - temperature stm of a single atom deposited on a metallic surface showed the asymmetric fano resonances in the tunnelling spectra , e.g. refs . in the stm measurements tip probes the transmission of electrons either through the adsorbed atom or directly from the surface . the transport of electrons through both channels leads to an asymmetric shape of the conductance curves which is typical behavior for the fano resonance resulting from constructive and destructive interference processes . the quantum interference can be also observed in the mesoscopic system with multiple energy levels @xcite . a model which incorporates a weak direct nonresonant transmission through a qd as well as the resonant tunneling channel was also considered in ref . @xcite in the context of the large value of the transmission phase found in the experiment for the kondo regime of a qd @xcite . a number of works has been devoted to the problem of the electron transport in the multiterminal qd systems and here we mention only a few of them . in ref . @xcite the conductance of the n - lead system was considered showing that the kondo resonance at equilibrium is split into n-1 peaks . in ref . @xcite an explicit form for the transmission coefficient in the electron transport through a qd connected with three leads was found . the electron transport and shot noise in a multi - terminal coupled qd system in which each lead was disturbed by classical microwave fields were studied in ref . multiterminal qd systems or magnetic junctions were also intensively investigated in context of the spin - dependent transport , e.g. refs . a three - terminal qd system was studied in ref . @xcite to measure of the nonequilibrium qd density of states ( splitting of the kondo resonance peak ) . the cross correlations of the currents and the differential conductance of the qd coupled with three leads described by the infinite - u anderson hamiltonian were considered in ref . the general formulation of the time - dependent spin - polarized transport in a system consisting of the resonant tunnelling structure coupled with several magnetic terminals was considered by zhao _ et al_. @xcite and as an application of this formalism the electron transport in a system with two terminals under an ac external field was investigated . in all papers mentioned above and in the studies relating to the electron transport through a qd with the additional ( bridge ) nonresonant transmission channels , the time - dependent external fields were not applied and the considered systems were driven out of the equilibrium only by means of a dc voltage bias ( see , however , ref . @xcite ) in this paper , we address the issue of a qd coupled with three leads with additional , nonresonant coupling between leads driven out of equilibrium by means of a dc voltage bias and time - dependent external fields . the qd is connected with three metal leads and one of these leads , say the left ( l ) lead , is coupled with the two remaining leads , say the first ( @xmath0 ) and the second ( @xmath1 ) right leads . the possible stm experimental setup corresponding to our model system is presented in fig . 1 . in literature , different theoretical approaches have been developed to treat the time - dependent electron transport in the mesoscopic systems . the most popular nonequilibrium green s function method depends on the two time arguments and for time - dependent problems it is a rather difficult task to calculate them without any approximations ( e.g. beyond the wide - band limit ) . therefore , in our treatment of the time - dependent problems we use the evolution operator which , as a rule , essentially depends on one time argument ( e.g. refs . such an approach is especially well suited for the problem with time - dependent coupling between the qd and leads . the outline of the paper is as follows . in sec . ii we start with the model and method for the derivation of the qd charge and currents . in sec . iii we present the results for the time - averaged currents and their derivatives with respect to the qd energy level position ( or equivalently , with respect to the gate voltage ) obtained for different time - dependence of the parameters characterizing the considered system . we consider also the transient current characteristics in the case of the rectangular - pulse modulations imposed on the qd - lead barriers . the last section presents the conclusions and in the appendix we give the short derivation of the evolution operator matrix elements needed in the qd charge and current calculations . we consider a qd coupled through the tunneling barriers @xmath2 @xmath3 to three metal leads . one of these leads , say the left lead ( l ) is coupled additionally with the remaining two leads , say the first and second right leads ( @xmath0 , @xmath1 ) by the tunneling barriers @xmath4 . in the following we will denote the wave vectors associated with the left lead by the letter @xmath5 and the wave vectors corresponding to the first and second right leads by the letters @xmath6 and @xmath7 , respectively . the chemical potentials @xmath8 of the three metal leads may not be equal , and their difference is not necessarily small . we write the hamiltonian of the considered system in the form @xmath9 , where @xmath10 @xmath11 the operators @xmath12 , @xmath13 are the annihilation ( creation ) operators of the electrons in the corresponding leads and the dot , respectively . for simplicity the dot is characterised only by a single level @xmath14 and the intra - dot electron - electron coulomb interaction is neglected . the neglect of coulomb interaction is quite reasonable in some systems and , as we are going to concentrate on the investigations of the influence of the third lead ( in comaprison with the qd - two leads system ) and the additional tunneling channels between the leads on the time - dependent transport , then in the first step ignoring the coulomb interaction should be justified . we consider our mesoscopic system in the presence of external microwave ( mw ) fields which are applied to the dot and three leads . in most theoretical treatments of photon - assisted electron tunneling it is assumed that the driving field equals the applied external field . however , the situation is more complicated and the internal potential can be different from the applied potential @xcite . one of the consequences will be , e.g. the asymmetry between the corresponding left and right sidebands @xcite . the main feature of the time - dependent transport remains , however , unchanged and in our treatment as usual we assume that in the adiabatic approximation the energy levels of the leads and qd are driven with the frequency @xmath15 and the amplitudes @xmath16 ( @xmath17 ) , @xmath18 and read @xmath19 , @xmath20 , respectively . the time - evolution of the qd charge and the current flowing in the system can be described in terms of the time - evolution operator @xmath21 given by the equation of motion ( in the interaction representation ) : @xmath22 with @xmath23 and @xmath24 where @xmath25 denotes the time ordering and the units such that @xmath26 have been chosen . here we have assumed that the interactions between the qd and leads , as well as between the left and right leads are switched on in the distant past @xmath27 . the qd charge is calculated according to the formula ( cf . @xcite ) : @xmath28 here @xmath29 and @xmath30 denote the matrix elements of the evolution operator calculated within the basis functions containing the single - particle functions @xmath31 , @xmath32 , @xmath33 and @xmath34 corresponding to three leads and qd , respectively . @xmath35 and @xmath36 represent the initial filling of the corresponding single - particle states . the tunneling current flowing , e.g. from the left lead , @xmath37 , can be obtained using the time derivative of the total number of electrons in the left lead , @xmath38 ( in the following we put @xmath39 ) , where @xmath40 in the following only the matrix elements of the evolution operator present in eqs . ( 4 ) and ( 5 ) are required and they can be obtained solving the corresponding sets of coupled differential equations constructed according to eq . ( 3 ) with @xmath41 written as follows : @xmath42 where @xmath43 and @xmath44 correspond to @xmath31 , @xmath45 , @xmath33 or @xmath34 , respectively . as the example , the matrix element @xmath29 required for the calculation of the first term of the qd charge , eq . 4 , can be obtained solving the following set of coupled differential equations @xmath46 the total number of coupled equations in this case is equal to @xmath47 , @xmath48 being the number of discrete wave vectors @xmath5 , @xmath6 and @xmath7 taken to perform the corresponding summation over the wave vectors . usually , the number @xmath48 equal to 100 - 200 is sufficient to achieve the desired accuracy of the calculations . we have solved numerically this and other similar sets of the coupled differential equations needed in calculations of all matrix elements of the evolution operator present in eqs . we have used this method for the special case of time - dependent couplings of the qd with leads and the couplings of the left lead with two right leads . the set of eqs . 7 - 9 in the case of vanishing over - dot couplings between the left and right leads is greatly simplified and gives , e.g. for @xmath29 : @xmath49 where @xmath50 and @xmath51 is the fourier transform of the @xmath52-th lead density of states and @xmath53 . similar simplifications occur in the calculations of other matrix elements of @xmath21 required in the formulas for @xmath54 and @xmath55 . however , for the nonvanishing couplings @xmath56 and @xmath57 ( over - dot bridge between the left and right leads ) one has to solve the starting set of eqs . much more analytical calculations can be done in the case of constant values of the tunneling matrix elements present in our model . in this case the general equation satisfied by @xmath29 is derived in the appendix and under the wide band limit ( wbl ) approximation , e.g. @xcite this equation takes the simple form @xmath58 here @xmath59 , @xmath60 , @xmath61 being the bandwidth of the lead energy band ( @xmath62 ) and @xmath63 . in the appendix we give the derivations of all functions needed for calculation of the qd charge and currents . we assumed the simplified assumption that all tunneling matrix elements are independent of the wave vectors . the interactions between the qd and leads are assumed to be equal between themselves and denoted by @xmath64 and the interactions between the left and two rights leads ( i.e. @xmath65 and @xmath66 ) corresponding to the over - dot tunneling channels are also equal one with another and denoted by @xmath67 . it is easy to show that the first term of the general formula for the qd charge , eq . 4 , together with the solution of eq . ( 12 ) , @xmath68 , tends to zero for @xmath69 as @xmath70 . the next terms of the qd charge formula can be calculated using the functions @xmath71 , @xmath72 and @xmath73 , eqs . 43,47 , being the solutions of the corresponding differential equations , eqs . finally , the time - averaged qd charge is given by @xmath74 where @xmath75 @xmath76 where @xmath77 denotes the time - averaging and @xmath78 denotes the fermi function of the @xmath52-th ( @xmath79 lead . noticing , that @xmath80 in order to calculate the current @xmath37 the functions @xmath81 , @xmath82 , @xmath83 and @xmath84 are required and they are given in the appendix in eqs . 37 , 44 , 49 . after lengthly but straightforward calculations we obtain for the time averaged current leaving the left lead the following formula : @xmath85 \,,\end{aligned}\ ] ] where @xmath86 @xmath87 and @xmath88 denotes the bessel function . the corresponding formula for the time - averaged current @xmath89 leaving @xmath90-lead , @xmath91 , can not be written in such symmetrical form as in eq . 18 , because the @xmath90-lead is coupled with @xmath92-lead only . for @xmath93 we have : @xmath94 \,,\end{aligned}\ ] ] where @xmath95 , @xmath96 , @xmath97 and @xmath98 for @xmath99 . note , that the integrals present in the formula for the qd charge and currents , eqs . 17,18,21 , can be easily performed analytically and final algebraic expressions can be obtained . especially simple and transparent form can be given for the conductance @xmath100 , @xmath101 . for example , @xmath102 reads as : @xmath103 where @xmath104^{-1 } \,,\end{aligned}\ ] ] @xmath105 analyzing eq . 22 one can find the origin of the asymmetric line shapes in differential conductance resulting from the interference of resonant and nonresonant tunneling paths . for the case of @xmath106 we observe the lorentzian resonances localized at @xmath107 . the amplitudes of these resonances are determined by the @xmath108-th order bessel functions calculated for the argument @xmath109 . for the case of nonvanishing @xmath110 , the resulting curve is a superposition of the lorentzian - like resonances and asymmetric parts weighed by the factors @xmath111 and @xmath112 , respectively . the lorentzian - like resonance is centered at @xmath113 with the peak width at half maximum ( fwhm ) equal to @xmath114 and the maximum value equal to @xmath115 . the asymmetric part of the differential conductance corresponding to the @xmath116th sideband is also centered in the same point with the distance between its maximum and minimum equal to @xmath117 and the absolute values of these extrema are equal to @xmath118 . for comparison , in the case of the qd coupled with two leads the corresponding lorentzian - like part of the differential conductance corresponding to the @xmath116th sideband is centered at @xmath119 , with fwhm equal to @xmath120 and the maximum value equal to @xmath121 . knowing the explicit expressions for the currents one can check the following relations between different elements of the conductance matrix @xmath122 , e.g. @xcite . current conservation implies @xmath123 , @xmath124 . on the other hand , @xmath125 only for @xmath126 . for other relations between the amplitudes @xmath18 and @xmath127 , @xmath128 , @xmath129 we have : @xmath130 for @xmath131 , and @xmath132 for @xmath133 . we consider the qd coupled with three , say the left , the first right and second right metal leads with the additional over - dot ( bridge ) couplings between the left and right leads . we present the results for the time - averaged currents and derivatives of the average current with respect to the qd energy level in the presence of external microwave fields which are applied to the dot and three leads , respectively . the time - dependent currents are also calculated in the case when the periodic rectangular - pulse external fields are applied to each qd - lead barrier . we have taken for @xmath110 the values comparable with @xmath134 and estimated @xmath134 ( assuming its @xmath5-independence , @xmath135 ) using the relation @xmath136 , where @xmath137 is the @xmath138-lead bandwidth and @xmath139 ( @xmath140 , @xmath141 was assumed ) . in our calculations we assumed @xmath142 , all energies are given in @xmath143 units , time in @xmath144 units , the current , its derivatives and frequency are given in @xmath145 and @xmath146 units , respectively . firstly , we consider the time - averaged currents in the presence of the time - varying ( harmonic case ) external fields . here we give the explicit formula for the averaged current , @xmath147 , performing the corresponding integrals in the general formula given in eq . ( 18 ) , and compare it with the current @xmath148 flowing in the system of the qd coupled with two leads only : @xmath149 \nonumber\\ & + & { \gamma \over \pi}{x(1 - 2x^2 ) \over ( 1 + 2x^2)^3 } \nonumber\\ & \times & \left . \sum_k \left [ j_k^2 \left ( { \delta_d-\delta_l \over \omega } \right ) \ln(g_l)- j_k^2 \left ( { \delta_d-\delta_i \over \omega } \right ) \ln(g_i ) \right ] \right\},\end{aligned}\ ] ] where @xmath150 , @xmath151 and @xmath152 , whereas for @xmath148 we have : @xmath153 \nonumber\\ & + & { \gamma \over \pi}{x(1-x^2 ) \over ( 1+x^2)^3 } \nonumber\\ & \times & \sum_k \left [ j_k^2 \left ( { \delta_d-\delta_l \over \omega } \right ) \ln(g_l^{(2)})- j_k^2 \left ( { \delta_d-\delta_i \over \omega } \right ) \ln(g_r^{(2 ) } ) \right],\end{aligned}\ ] ] where @xmath154 and @xmath155 , @xmath156 . note , that @xmath147 consists of the two terms and each term is similar in its structure to the current flowing in the qd - two leads ( qd-2l ) system , @xmath148 . however , due to the interference of the charge carriers propagating along the different ways the arguments of the _ arctangens _ and _ logarytmic _ functions are different and the individual terms in @xmath147 and @xmath148 are weighted by different @xmath157-dependent factors . against @xmath14 in the system of a qd coupled with three ( solid line ) or two ( broken line ) leads . the microwave field is applied to the qd with @xmath158 and @xmath159 , @xmath160 , @xmath161 , @xmath162 , @xmath163 . ] in figure [ fig2 ] we compare the averaged values of the current flowing from the left lead in the systems in which the qd is coupled with three or two leads , the solid and broken curves , respectively . the external microwave field is applied only to the qd and dc bias between the left and right leads is small in comparison with @xmath15 , @xmath18 and @xmath143 . the coupling @xmath110 is assumed to be zero . in such a case the sidebands on the current curve are clearly visible . the width of the corresponding peaks is smaller for the case of the qd coupled with two leads . analyzing the expressions for @xmath164 and @xmath148 one can obtain ( for @xmath165 and @xmath161 ) the subsequent peaks in @xmath147 ( as functions of @xmath14 ) in the form @xmath166 with the fwhm equal @xmath167 . for comparison , in the case of the qd coupled with two leads the corresponding peaks are described by the functions @xmath168 with the fwhm equal @xmath169 . against @xmath14 in the system of a qd coupled with three ( thick lines ) or two ( thin lines ) leads . the solid , broken and dotted curves correspond to @xmath170 and @xmath171 , respectively . @xmath172 , @xmath173 , @xmath174 , @xmath159 . ] next , in fig . [ fig3 ] we analyze the influence of the over - dot coupling between the left and right leads . again , as in fig . [ fig2 ] , we consider an external microwave field applied only to the qd with @xmath172 and @xmath175 . the thin ( thick ) lines correspond to the qd coupled with two ( three ) leads . we show @xmath147 for three values of the inter - leads coupling strength represented by the parameter @xmath176 . for @xmath177 we have well defined sideband peaks as previously shown in fig . next , we show the results for @xmath178 and @xmath179 . note , that the expression for @xmath147 ( and @xmath148 ) consists of three terms . the first term depends on the difference @xmath180 and does not depend on the qd energy level @xmath14 . the second term corresponds to the loretzian - type contribution to a given sideband ( it disappears for @xmath181 ) and the last term corresponds to an asymmetric contribution ( it disappears for @xmath182 ) , respectively . this last term influences the sideband shape only for nonvanishing over - dot coupling between leads and is the most prominent sign of the interference effects . for @xmath183 this term disappears and the resulting sidebands have a form of symmetric dips due to the negative value of the coefficient @xmath184 in eq . 27 . in this case the nonresonant tunneling channels modify the dips center position and its fwhm in comparison with the position and fwhm of sidebands presented for @xmath177 . note , that for a qd coupled with two leads , the corresponding sidebands ( the thin dotted line ) are not fully symmetric curves as in this case the last term of eq . 28 disappears for @xmath185 and not for @xmath186 ( c.f . @xcite ) . for @xmath181 the corresponding sidebands are fully asymmetric curves as in this case the second term in eq . 27 ( which introduces asymmetry ) disappears . again , for a qd coupled with two leads the corresponding sidebands ( the thin broken curve ) are described by not fully asymmetric curves as the second term of eq . 28 , which gives a symmetric contribution to sidebands , disappears for @xmath187 and not for @xmath188 . , @xmath189 and @xmath190 ( solid , broken and dotted lines , respectively ) against @xmath14 in the system of a qd coupled with three leads . the upper ( lower ) panel corresponds to @xmath177 ( @xmath191 ) . @xmath172 and other parameters as in fig . the values of @xmath192 for @xmath177 ( upper panel ) have been multiplied by a factor of 20 for illustrating purposes . ] in the next fig . [ fig4 ] we show the currents flowing in the qd - three leads ( qd-3l ) system for @xmath177 and @xmath191 . for vanishing values of the over - dot coupling the current @xmath147 is characterized by a sequence of the symmetric peaks , but the current @xmath190 is a superposition of the asymmetric structures placed in the points where the symmetric sidebands occur on @xmath147 curve . analyzing the current @xmath190 according to eq . 29 we have for the parameters in fig . [ fig4 ] : @xmath193,\end{aligned}\ ] ] where @xmath194 . one can see that each sideband is the sum of the peak ( two first terms in eq . 29 ) and the dip ( the last two terms in eq . 29 ) resulting in the asymmetric structure shown in fig . [ fig4 ] . and @xmath15 . the upper ( middle ) panel shows @xmath147 ( @xmath190 ) for @xmath177 and the lower panel gives @xmath147 for @xmath191 . @xmath172 , @xmath173 , @xmath195 . ] to learn more about the influence of the third electrode and additional over - dot coupling between leads we present in fig . [ fig5 ] the currents @xmath147 for @xmath177 and @xmath191 and @xmath190 for @xmath177 . for vanishing coupling between leads ( @xmath177 ) the current @xmath147 exhibits a well known sideband structure for @xmath196 and for small frequencies , @xmath197 , the two broad maxima at @xmath198 are present . at the same time , the current @xmath190 exhibits the asymmetric structures centered on the @xmath199 plane at the points where photon sidebands occur on the @xmath147 curves . these asymmetric structures on the @xmath190 curve exist also at @xmath197 . on the other hand , similar structure of the @xmath164 ( the lower panel of fig . [ fig5 ] ) is obtained for @xmath191 , i.e. we observe a number of asymmetric resonances separated by the photon energy for @xmath200 . notice the similarity of both pictures , i.e. @xmath147 calculated for @xmath191 and @xmath190 for @xmath177 , respectively . note , however , the different scale for these currents . against @xmath14 for different values of the over - dot coupling strength between the left and right leads . the broken , dotted , thin solid and thick solid lines correspond to @xmath201 and @xmath171 , respectively . @xmath202 and @xmath203 , @xmath204 , @xmath205 , @xmath206 , @xmath159 . ] in fig . [ fig6 ] we present @xmath147 for different over - dot coupling assuming a strong asymmetry of the applied microwave field ( ac potential is applied only to the @xmath0-lead in the qd-3l system ) . the additional @xmath1-lead is characterized by the chemical potential @xmath207 . for better demonstration of the influence of the ac potential on @xmath147 we moved down each curve by the constant value @xmath202 , for @xmath177 we observe a shoulder on the left side of the main peak and a small negative current for the positive values of @xmath14 . this picture is similar to the known results ( for @xmath177 ) obtained experimentally and theoretically in the qd-2l systems , e.g. refs . @xcite . with the increasing over - dot coupling @xmath110 the height of the main resonant peak decreases and disappears at all for @xmath191 but the enhancement of the current for negative values of @xmath14 ( for @xmath208 ) is more and more expressive . at the same time , for all values of @xmath157 we observe a negative current for the small positive @xmath14 with a greater absolute value for stronger coupling between leads . for greater values of @xmath157 the shape of the @xmath147 curve is changed dramatically and for @xmath186 , @xmath147 becomes nearly reversed in comparison with that calculated for @xmath177 . against @xmath14 for the case of the microwave field applied only to the qd with @xmath209 ( thin line ) or @xmath158 ( thick line ) - upper panel . the lower panel corresponds to the case of the microwave field applied to the qd and @xmath1 lead with @xmath210 and @xmath211 or @xmath212 ( thin or thick lines , respectively ) . the broken lines show the results when the microwave field applied to the qd and @xmath1 lead are out of phase ( with the phase difference of @xmath213 ) , @xmath205 , @xmath206 , @xmath159 . ] in fig . [ fig7 ] we analyze the influence of the third lead ( here named as @xmath1 ) on the current @xmath147 when the external microwave field is applied to this lead and to the qd with @xmath210 . for comparison , we add in the upper panel the results for @xmath147 obtained for the case when this additional third lead is not irradiated by the microwave field . in this case , as before , see e.g. fig . [ fig2 ] , we observe typical sidebands on the current curves ( the difference between the lead chemical potentials is small in comparison with the amplitude @xmath18 ) . however , after including the third lead irradiated by the external microwave field the dependence of the current @xmath147 on the gate voltage ( or equivalently on the qd energy level position ) is quite different - compare the thin or thick solid lines of the upper and lower panels . for smaller values of @xmath214 and @xmath129 , the averaged current @xmath215 is very similar to the corresponding current @xmath216 obtained by applying the external microwave field only to one lead ( see the broken line in fig . these curves are , however , related between themselves by a relation @xmath217 . now we can observe a small negative current at small negative values of @xmath14 and some enhancement of the current on the right side of the main peak . similarly , the significant differences between the corresponding currents are observed also for greater values of the amplitudes @xmath214 and @xmath129 ( compare the thick solid lines in the upper and lower panels of fig . [ fig7 ] ) . note , that very similar behavior of the current @xmath147 as the function of the gate voltage is observed if we compare the case when the microwave field is applied only to one lead and the case when the microwave field is applied simultaneously to the qd and @xmath218lead but with the phase difference of @xmath213 - compare the thin broken curve in fig . [ fig7 ] with the broken curve in fig . [ fig6 ] . against @xmath14 in the system of a qd coupled with three ( thick lines ) or two ( thin lines ) leads . the upper ( lower ) panel corresponds to @xmath177 ( @xmath191 ) . in a qd - three leads system @xmath219 , @xmath220 and @xmath221 and @xmath222 ( solid , broken and dotted curves , respectively ) . in a qd coupled with two leads @xmath219 , @xmath223 and @xmath222 ( solid , broken and dotted curves , respectively ) . the values of solid curves have been multiplied by a factor of 10 for the illustrating purposes . @xmath159 , @xmath172 , @xmath173 . ] next , we consider the qd-2l system for different values of the chemical potential of the right lead ( @xmath221 and 10 ) and show the results of including the next lead ( @xmath224lead ) to the system with @xmath220 . in fig . [ fig8 ] we present the results obtained for @xmath147 in the case when the microwave field is applied only to the qd . for the case @xmath174 and @xmath177 the sidebands are very clearly visible and the corresponding peaks are lower and broader for the qd-3l system as we discussed before ( the upper panel ) . for the non - zero over - dot coupling between the left and right leads the sideband peaks get asymmetric forms and for @xmath191 they become fully asymmetric ( the lower panel ) . the current for the qd-2l system is also composed of a number of asymmetric components although now these forms are fully asymmetric for @xmath225 as we know from the earlier discussion . for greater values of @xmath226 , the corresponding current flowing in the qd-2l system achieves ( for vanishing , as well as for nonzero coupling between leads ) greater negative values and its dependence on the qd energy level position is well - marked in comparison to the results characterizing the qd-2l system . against @xmath14 for the qd-2l systems : @xmath227 - thin solid lines and @xmath219 , @xmath228 - thick solid lines and the qd-3l system : @xmath229 , @xmath230 - broken lines . the upper ( lower ) part corresponds to @xmath177 ( @xmath191 ) and @xmath159 , @xmath231 , @xmath173 . the dotted lines correspond to the sum of the currents flowing in two qd-2l systems . ] in order to emphasize the influence of the additional lead on the currents we show in fig . [ fig9 ] the current @xmath164 for qd-2l and qd-3l systems calculated for the parameters for which the corresponding curves are relatively simple . we assumed the small amplitude of the qd energy level oscillations , @xmath231 , and @xmath159 , for which ( for @xmath177 ) only the central peak corresponding to elastic tunneling is visible on the @xmath147 curves . we show @xmath164 obtained for two different qd-2l systems which can be viewed as components of more complicated qd-3l system . we observe that due to the interference effects , the current @xmath147 flowing in the qd-3l system is not simply a sum of currents ( dotted lines ) flowing in the corresponding qd-2l systems . the difference between this sum and the current corresponding to the qd-3l system is relatively large and exists independently of the coupling between leads . against @xmath14 and @xmath15 for @xmath177 . we compare the results obtained for the qd coupled with two leads - the upper ( middle ) panel - for @xmath232 , @xmath233 , @xmath234 , @xmath235 , @xmath236 ( @xmath232 , @xmath237 , @xmath234 , @xmath235 , @xmath238 ) with the results obtained for the qd coupled with three leads - the lower panel - @xmath234 , @xmath239 , @xmath240 , @xmath241 , @xmath232 , @xmath242 , @xmath243 , @xmath159 . ] in order to present more information about the differences between the electron transport in the qd-3l and qd-2l systems we display in fig . [ fig10 ] the derivative @xmath244 as a function of @xmath14 and @xmath15 . the lowest panel corresponds to the qd-3l system and two other panels correspond to the qd-2l systems . these qd-2l systems are characterized by such parameters that combined together give us the considered qd-3l system . one can see , that the considered characteristics of the electron transport in the qd-3l system are not simply the algebraic sum of the corresponding curves of both qd-2l systems . in all three cases shown in fig [ fig10 ] , the position of the corresponding minima and maxima ( along the @xmath245axis ) can be identified with the values of the leads chemical potentials . however , the corresponding structures are less clear in the case of the qd-3l system in comparison with those for the qd-2l models . , @xmath0 and @xmath1 . the @xmath92-lead is coupled with the qd only - the left panels and with the qd and two other leads , @xmath246 - the right panels . the couplings between the qd and @xmath247 leads are changed periodically . the upper , middle and lower panels correspond to @xmath248 and -3 , respectively . @xmath249 . the thin , thick and broken curves correspond to @xmath250 and @xmath251 currents , respectively . ] in the next step of our investigations of the electron transport in the qd-3l systems we consider the time - dependent currents flowing in response to the time - dependent barriers between the qd and leads or in response to suddenly removed ( or inserted ) connection of the qd with one of the leads . as a first we assume a rectangular - pulse modulation applied to the qd-@xmath0 lead and qd-@xmath1 lead barriers . we assume that these modulations are with a phase difference of @xmath213 . in the first ( second ) half - cycle , @xmath252 ( @xmath253 ) and the qd is coupled only to the @xmath1 lead ( @xmath0 lead ) . in addition , the qd is coupled to the next , say @xmath254lead , with a constant value @xmath255 . in the following we consider the time - dependent currents @xmath256 , @xmath257 and @xmath258 for the three specific conditions : @xmath259 , @xmath260 and @xmath261 . in addition , we assume @xmath262 , @xmath263 and take for the period of the considered barrier modulation @xmath264 . in a such case we integrate numerically the corresponding set of the differential equations for the matrix elements of the evolution operator and in the next step calculate the currents according to the formula @xmath265 , where @xmath266 or @xmath1 and @xmath267 is given in eq . 5 ( or similar to it ) . we checked that the qd charge hardly depends on the additional @xmath110 couplings . although the qd charge is almost insensitive to the additional over - dot couplings the currents demonstrate such dependence . especially visible are the differences for the case when the chemical potential @xmath268 of the third electrode @xmath92 lies between the chemical potentials of two other leads , see fig . [ fig11 ] b , e . for other values of @xmath268 , the influence of the over - dot tunneling channels for the parameters considered here is smaller . note , that after abrupt changing of the coupling between the qd and @xmath0 , or @xmath1 leads the currents @xmath269 and @xmath251 rapidly change too , and after a short time reach the steady values . the qd coupled with three leads could be considered as the three - state system . we observe that for some values of the lead chemical potentials the currents change their values periodically e.g. from zero to the positive value ( see @xmath251 in fig . [ fig11]b ) , from zero to the negative value ( see @xmath270 in fig . [ fig11]b ) or from the negative to the positive value ( see @xmath215 in fig . [ fig11]b).the additional couplings between leads modify the values of the currents but the qualitative picture remains the same . note , that in the first moment after abrupt changing of the coupling between the qd and @xmath0 or @xmath1 lead we have @xmath271 as in this case @xmath272 ( not shown here ) . after some delay time the qd charge stabilizes around its equilibrium value , the currents tend to constant values and their sum is equal to zero . , @xmath0 and @xmath1 leads in the system shown in fig . the thin solid , thick solid and broken curves correspond to different couplings of the @xmath92 lead to the other elements of the system : ( @xmath273 , @xmath274 ) , ( @xmath273 , @xmath275 , @xmath276 ) and ( @xmath277 ) , respectively . ] in the last step we consider the response of the currents to the abrupt inclusion into the qd-2l system of the third electrode ( in our case , @xmath92-lead ) . the results are presented in fig . [ fig12 ] together with the schematic view of the subsequent tunnel connections between the qd and the three leads . we show the time - dependent currents @xmath37 , @xmath278 and @xmath258 corresponding to the three different ways of inclusion of the @xmath92-lead . we assumed the chemical potentials @xmath279 , @xmath260 and @xmath280 . consider the current @xmath251 flowing from the @xmath1-lead characterized by the highest chemical potential @xmath226 . before adding to the system of the @xmath92-lead the current @xmath251 has a constant value and flows from @xmath1-lead through the qd energy level to @xmath0-lead . when the @xmath92-lead is included into the system ( the tunneling coupling @xmath281 changes abruptly at @xmath282 from zero to nonzero value ) the current @xmath251 is almost unchanged - its value decreases slightly without any transients at short times after the time @xmath282 . next , we consider the case when the @xmath92-lead is abruptly connected simultaneously with the qd and @xmath0-lead . now the current @xmath251 decreases significantly during the short time after the moment of inclusion of @xmath92-lead and settles to its constant value . note , that @xmath251 decreases despite the additional charge transfer channel between @xmath1 and @xmath92 leads ( through the qd energy level ) . the destructive interference appears in this case as we have two transmission channels for tunneling electrons between @xmath1 and @xmath0 leads . however , the constructive interference is visible if we consider the next case when the @xmath92-lead is abruptly coupled with the qd-2l system assuming nonzero values of @xmath281 , @xmath283 and @xmath284 . now , we have one additional charge transfer channel ( from @xmath1-lead ) in comparison with the former case . the current @xmath251 rapidly increases with some fluctuations and after the time @xmath285 decreases to the constant value greater than in qd-2l system . similar analysis can be made considering the currents @xmath286 and @xmath270 although the transient current changes are more visible now at short times after the abrupt inclusion of additional electron tunneling channels . the above discussion concerns the specific values of the lead chemical potentials and the position of the qd energy level . nevertheless , the similar qualitative conclusions can be made also for other values characterizing the considered system . we have studied the time - dependent tunneling transport through the qd coupled with three metal leads using the evolution operator technique . the time - dependent qd charge and currents were determined in terms of the appropriate evolution operator matrix elements . applying the wide band limit to the integrodifferential equations satisfied by the evolution operator matrix elements we were able to give the analytical expressions for the time - averaged currents and differential conductance . we considered the external harmonic microwave fields applied to different parts of the considered system , as well as , the rectangular - pulse modulation imposed on different qd - leads barriers . in addition , we have studied also the time dependence of the currents due to abrupt inclusion into the qd - two leads system of the third electrode . we have considered also the effect of the additional couplings between leads ( we coupled one of the leads with the other two leads ) on the conductance and current flowing in the system . our main results can be summarized as follows : + - for the vanishing nonresonant tunneling path , @xmath106 , and for the parameters for which the photon - assisted sidebands are clearly visible on the @xmath147 curve the subsequent sideband peaks have the lorentz - type form with the fwhm equal @xmath167 in comparison with @xmath169 for the qd - two leads system . for the increasing value of @xmath110 the form of the sidebands transforms from the lorentz - type to the fully asymmetric form for @xmath287 . for greater @xmath157 the form of the sidebands changes and gains again the lorentz - type shape for @xmath183 . for the qd coupled with two leads the corresponding values of @xmath157 are equal to @xmath288 and @xmath289 , respectively . + - for the vanishing @xmath110 the differential conductance curve , e.g. @xmath290 , possesses the sidebands of the lorentz - type localized at @xmath291 . for @xmath292 these sidebands are described by the superposition of two parts , the lorentz - type and the asymmetric one centered at @xmath293 weighed by the factors @xmath294 and @xmath295 , respectively . for the qd - two leads system the corresponding @xmath116th sidebands are centered at @xmath296 and their symmetric and asymmetric parts are weighted by the factors @xmath297 and @xmath298 , respectively . + - the symmetry properties of the sidebands corresponding to the current flowing from the given lead depend on the position of the chemical potential of this lead in comparison with the chemical potentials of the other two leads . taking , for example @xmath106 and @xmath226 localized in the middle between @xmath268 and @xmath299 one can observe on the @xmath164 curve the sidebands of nearly regular ( lorentz - type ) forms while the sidebands on the @xmath190 curve have asymmetric structures . however , in the presence of the nonresonant tunneling path the sidebands on the @xmath164 curve change their form and for @xmath181 they have a fully asymmetric shape . on the other hand , the sidebands on the @xmath190 curve have for @xmath181 a nearly lorentz - like shape . + - especially large interference effects can be observed if we compare the current flowing in the qd-3l system with the sum of currents flowing in the two qd-2l systems which can be viewed as the components of the considered more complicated qd-3l system . the difference between them is relatively large independently of the over - dot coupling between leads ( see fig . [ fig9 ] ) + - in the case of strong asymmetry of the applied external field ( @xmath300 ) we observe for @xmath177 a shoulder on the left side of the main resonant peak on the @xmath147 curve vs. the gate voltage . with the increasing over - dot coupling between @xmath92 and @xmath301leads the main resonant peak disappears and transforms in a dip for strong coupling @xmath110 . at the same time a shoulder of this curve increases with the increasing @xmath110 . + - let us consider the time dependence of currents flowing in response to the time - dependent barriers between the qd and two leads ( we assume a constant coupling of the qd with the third lead ) . for the assumed rectangular - pulse modulation applied to the qd-@xmath0 lead and qd-@xmath1 lead barriers one can consider the qd-3l system as a three - state one . for example , the currents @xmath302 , @xmath278 and @xmath258 change their values periodically between zero and positive , positive and zero and negative and positive values , respectively ( see fig . the additional couplings between the leads introduce only small quantitative changes . + * acknowledgements : * the work of one of us ( rt ) has been partially supported by the kbn grant no . pbz - min-008/p03/2003 . in this section we present the derivations of the general equations satisfied by the required functions @xmath29 , @xmath303 , @xmath304 and @xmath305 , @xmath306 , @xmath307 needed for the calculations of the qd charge @xmath54 and the currents flowing in the considered system . in the next step , using the wbl approximation we simplify these equations and give the analytical solutions for them . let us begin from the derivation of the integro - differential equation satisfied by @xmath29 . writing down the formal solution of eq . ( 11 ) @xmath308 and inserting them to the formal solution of eq . ( 10 ) @xmath309 one can obtain after straightforward calculations the function @xmath81 expressed in the terms of @xmath29 , only @xmath310 where @xmath311 we remember , that the wave vectors @xmath312 and @xmath7 correspond to the left lead and the first and second right leads , respectively . inserting into eq . 9 the expressions for the functions @xmath313 , @xmath314 and @xmath315 , eqs . ( 17 , 18 ) , we can write the integro - differential equation for @xmath29 in the form @xmath316 u_{dd}(t_1,t_0 ) \nonumber\\ & + & \sum^\infty_{j=1}(-i)^{2j-1 } \int\limits^t_{t_0}\int\limits^{t_1}_{t_0 } \ldots \int\limits^{t_{2j-1}}_{t_0 } dt_1 \ldots dt_{2j } \sum_{\vec k_1,\vec k_2 , \ldots,\vec k_j } [ r_{d\vec k_1}(t , t_1 ) r_{\vec k_1\vec k_2 } ( t_2,t_3 ) \ldots\nonumber\\ & & r_{\vec k_{j-1}\vec k_j}(t_{2j-2},t_{2j-1 } ) \tilde v_{\vec k_jd}(t_{2j } ) \nonumber\\ & + & \tilde v_{d\vec k_1}(t ) r_{\vec k_1\vec k_2}(t_1,t_2)\ldots r_{\vec k_jd } ( t_{2j-1},t_{2j } ) ] u_{dd}(t_{2j},t_0 ) \nonumber\\ & + & \sum^\infty_{j=1}(-i)^{2j } \int\limits^t_{t_0}\int\limits^{t_1}_{t_0 } \ldots \int\limits^{t_{2j}}_{t_0 } dt_1\ldots dt_{2j+1 } \sum_{\vec k_1\vec k_2 \ldots \vec k_j } [ r_{d\vec k_1}(t , t_1 ) r_{\vec k_1\vec k_2 } ( t_2,t_3 ) \ldots \nonumber\\ & & r_{\vec k_{j-1}\vec k_j } ( t_{2j-2},t_{2j-1 } ) \ , r_{\vec k_jd}(t_{2j},t_{2j+1 } ) \\ & + & \sum_{\vec k_{j + 1 } } \tilde v_{d\vec k_1}(t ) \ , r_{\vec k_1\vec k_2}(t_1,t_2 ) \ldots r_{\vec k_j,\vec k_{j+1}}(t_{2j-1},t_{2j } ) \tilde v_{\vec k_{j+1}d } ( t_{2j+1 } ) ] u_{dd}(t_{2j+1},t_0 ) \ , . \nonumber % 10 ( a.5)\end{aligned}\ ] ] this rather untractable general equation can be greatly simplified using the wbl approximation . assuming @xmath317 , @xmath318 , the multidimensional time integrations and summations over the wave vectors can be performed giving in result @xmath319 with the solution @xmath320 here @xmath321 and @xmath60 , @xmath61 being the bandwidth of the lead energy band ( @xmath62 ) . the function @xmath81 given in eq . ( 19 ) can be reduced within the wbl approximation to the form @xmath322 where @xmath323 in order to calculate @xmath71 we write down accordingly with eq . ( 3 ) the corresponding set of coupled differential equations for the functions @xmath324 , @xmath82 , @xmath325 and @xmath326 . the subsequent steps of the calculations are similar to those performed in the derivation of eq . ( 34 ) . inserting the formal solutions for @xmath82 , @xmath327 @xmath328 \ , , \nonumber\\ \\ % 13 ( a.11 ) & & u_{\vec p\vec k}(t , t_0 ) = - i\int\limits^t_{t_0 } dt_1 \left[\tilde v_{\vec pd}(t_1 ) u_{d\vec k } ( t_1,t_0 ) + \sum_{\vec k_1 } \tilde v_{\vec pk_1}(t_1 ) u_{\vec k_1\vec k } ( t_1,t_0)\right ] \ , , % 14 ( a.12)\end{aligned}\ ] ] into the differential equation satisfied by @xmath71 one obtains the derivative @xmath329 expressed in terms of @xmath71 and @xmath82 . on the other hand , on the basis of eqs . ( 33 , 34 ) the function @xmath82 can be represented in the form containing only @xmath330 . finally , one obtains @xmath331 \nonumber\\ & + & ( -i)^3 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0 } dt_1 dt_2 \left[\sum_{\vec k_1,\vec q_1 } \tilde v_{d\vec k_1}(t ) \tilde v_{\vec k_1\vec q_1}(t_1 ) \tilde v_{\vec q_1\vec k } ( t_2 ) + \sum_{\vec k_1,\vec r_1 } \ldots \right ] \nonumber \\ & + & ( -i)^4 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0 } \int\limits^{t_2}_{t_0 } dt_1 dt_2 dt_3 \left[\sum_{\vec q_1,\vec k_1,\vec q_2 } \tilde v_{d\vec q_1}(t ) \tilde v_{\vec q_1\vec k_1}(t_1 ) \tilde v_{\vec k_1\vec q_2 } ( t_2 ) \tilde v_{\vec q_2\vec k}(t_3 ) \right . + \nonumber\\ & & \hspace{3cm}\left . \sum_{\vec r_1,\vec k_1\vec q_2 } \ldots + \sum_{\vec q_1\vec k_1 \vec r_2 } \ldots + \sum_{\vec r_1\vec k_1\vec r_2 } \ldots \right ] + \ldots \nonumber\\ & + & ( -i)^2 \int\limits^t_{t_0 } dt_1 \left[\sum_{\vec k_1 } \tilde v_{\vec d\vec k_1}(t ) \tilde v_{\vec k_1d } ( t_1 ) + \sum_{\vec q_1 } \tilde v_{d\vec q_1}(t ) \tilde v_{\vec q_1d}(t_1)\right . \nonumber\\ & & \left . \hspace{3 cm } + \sum_{\vec r_1}\tilde v_{d\vec r_1}(t ) \tilde v_{\vec r_1d}(t_1)\right ] u_{d\vec k}(t_1,t_0 ) \nonumber \\ & + & ( -i)^3 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0 } dt_1 dt_2 \left[\sum_{\vec k_1,\vec q_1 } \tilde v_{d\vec k_1}(t ) \tilde v_{\vec k_1\vec q_1}(t_1 ) \tilde v_{\vec q_1d}(t_2 ) \right.\nonumber\\ & & \hspace{3cm}\left . + \sum_{\vec k_1,\vec r_1 } \ldots + \sum_{\vec q_1\vec k_1 } \ldots + \sum_{\vec r_1\vec k_1 } \ldots \right ] u_{d\vec k}(t_2,t_0 ) + \ldots % 15 ( a.13)\end{aligned}\ ] ] in the wbl approximation this equation reduces to the form @xmath332 with the solution @xmath333 where @xmath334 . taking into account eqs . ( 33 , 34 ) the function @xmath82 can be written in terms of @xmath335 @xmath336 in order to calculate @xmath337 we first write down the coupled set of equations ( on the basis of eq . ( 3 ) ) satisfied by the functions @xmath72 , @xmath338 , @xmath339 and @xmath83 . inserting the formal solutions for the functions @xmath338 , @xmath339 and @xmath83 into the differential equation for the function @xmath72 one obtains the derivative @xmath340 expressed in terms of @xmath341 , @xmath83 , @xmath338 and @xmath339 . inserting again into this equation the formal solutions for the functions @xmath338 , @xmath339 and @xmath83 , and repeating this process again and again , one obtains : @xmath342\nonumber\\ & + & ( -i)^4 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0 } \int\limits^{t_2}_{t_0 } dt_1 dt_2 dt_3 \left[\sum_{\vec k_1\vec q_1\vec k_2 } \tilde v_{d\vec k_1}(t ) \tilde v_{\vec k_1\vec q_1}(t_1 ) \tilde v_{\vec q_1\vec k_2}(t_2 ) \tilde v_{\vec k_2\vec q } ( t_3 ) \right . \nonumber \\ & & \left . \hspace{3 cm } + \sum_{\vec k_1\vec r_1\vec k_2}\ldots + \ldots \right ] + \ldots \nonumber\\ % % % % % % % % % % % % % & + & ( -i)^2 \int\limits^t_{t_0 } dt_1 \sum_{\vec p=\vec k_1,\vec q_1,\vec r_1 } \tilde v_{d\vec p}(t ) \tilde v_{\vec p d}(t_1 ) u_{d\vec q } ( t_1,t_0 ) \nonumber \\ & + & ( -i)^3 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0 } dt_1 dt_2 \left[\sum_{\vec q_1,\vec k_1 } \tilde v_{d\vec q_1}(t ) \tilde v_{\vec q_1\vec k_1}(t_1 ) \tilde v_{\vec k_1d}(t_2 ) \right.\nonumber\\ & + & \left . \sum_{\vec r_1\vec k_1 } \ldots + \sum_{\vec k_1\vec q_1 } \ldots + \sum_{\vec k_1\vec r_1 } \ldots \right ] u_{d\vec q}(t_2,t_0 ) + \ldots % 18 ( a.19)\end{aligned}\ ] ] in the wbl approximation this equation is reduced to the following form : @xmath343 where @xmath344 and has the solution as follows : @xmath345 the function @xmath73 is identical with @xmath72 . to calculate , e.g. the current @xmath37 , we still need the functions @xmath346 and @xmath347 . the function @xmath83 can be obtained solving the set of the coupled differential equations for the functions @xmath72 , @xmath348 , @xmath339 and @xmath83 . writing down the formal solution for @xmath83 and inserting into it , in the first step , the formal solutions for @xmath349 and @xmath339 and , in the second step , the formal solutions for @xmath72 and @xmath83 , and so on , one obtains @xmath350 u_{d\vec q}(t_1,t_0 ) \nonumber\\ & + & ( -i)^3 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0}\int\limits^{t_2}_{t_0 } dt_1 dt_2 dt_3 \left[\sum_{\vec q_1,\vec k_1 } \tilde v_{\vec kq_1}(t_1 ) \tilde v_{\vec q_1\vec k_1}(t_2 ) \tilde v_{\vec k_1d}(t_3 ) + \sum_{\vec r_1\vec k_1 } \ldots \right ] u_{d\vec q } ( t_3,t_0 ) \nonumber \\ & + & \ldots + ( -i ) \int\limits^t_{t_0 } dt_2 \tilde v_{\vec k\vec q } ( t_1,t_0 ) + ( -i)^3 \int\limits^t_{t_0 } \int\limits^{t_1}_{t_0 } \int\limits^{t_2}_{t_0 } dt_1 dt_2 d_3\nonumber \\ & & \hspace{3 cm } \left[\sum_{\vec q_1,\vec k_1}\tilde v_{\vec kq_1}(t_1 ) \tilde v_{\vec q_1\vec k_1}(t_2 ) \tilde v_{\vec k_1d}(t_3 ) + \sum_{\vec r_1\vec k_1 } \ldots\right ] + \ldots % 26 ( a.24)\end{aligned}\ ] ] under the wbl approximation this equation reduces to the form : @xmath351 where @xmath341 is given in eq . ( the function @xmath84 has the identical form as @xmath352 . m. tsukada and n. shima , in : `` dynamical processes and ordering on solid surfaces '' , eds . a. yoshimori and m. tsukada ( springer , berlin 1995 ) , p. 34 ; s. tsuneyuki , n. shima , m. tsukada , surf . * 186 * 26 ( 1987 )

kempe s universality theorem @xcite is one of the great theorems of theoretical mechanism science ( `` beautiful '' @xcite , `` surprising '' @xcite , `` incredible theoretical significance '' @xcite , `` shocking '' @xcite ) . it states that any bounded portion of a planar algebraic curve can be traced out by one joint of a planar linkage with revolute joints . discovered only shortly after the invention of the first straight line linkages , kempe s theorem must have been a true surprise to his contemporary kinematicians . throughout the 20th centry , it was considered a milestone result . kempe s constructive proof can be used to actually compute a linkage that draws a planar algebraic curve . however , it was clear from the beginning that this construction is of no practical relevance . it requires an excessive number of links and joints , even for curves of low degree . nowadays , an asymptotic bound of @xmath3 for the number of links necessary to draw an algebraic curve of degree @xmath0 in an ambient space of dimension @xmath4 is known @xcite . nonetheless , drawing an ellipse with a kempe linkage already requires hundreds of links @xcite . a wealth of more practical examples for algebraic curve generation can be found in the monograph @xcite . however , the constructions there are rather specific to certain classes of curves and , more importantly , use mechanical constraints different from rigid links and revolute joints . kempe s theorem talks about algebraic curves and it is natural to ask for simplifications in case of rational curves . this was done recently in @xcite where the authors constructed scissor - like linkages to draw rational planar curves . their construction is based on the factorization of certain polynomials over a non - commutative ring that describe the motion of one of the links . the upper bounds on the number of links and joints for curves of degree @xmath0 reduce dramatically to @xmath5 and @xmath6 , respectively . in this article , we extend the ideas of @xcite to rational space curves . our aim is to provide a construction that works for all rational space curves and , at the same time , to reduce the number of links and joints as far as possible . for this purpose , we introduce several new ideas that also improve the planar case . we use rational motions of minimal degree in the dual quaternion model of rigid body displacements for the link that draws the given space curve . this acknowledges the importance of circularity . if the rational space curve is entirely circular , the motion degree is particularly low @xcite and factorization of the motion polynomial is straightforward , without the need for prior degree elevation . in the non - circular case , a degree elevation is necessary but in contrast to @xcite , we only preserve one relevant trajectory , not the complete rational motion . this allows to keep the degree lower and saves links and joints . if the curve is of degree @xmath0 and circularity @xmath1 , the bounds for links and joints are @xmath7 and @xmath8 , respectively . another advantage of our spatial approach concerns a certain defect in kempe s original construction that later was even considered a flaw ( * ? ? ? * section 3.2 ) . kempe used parallelogram and anti - parallelogram linkages as basic building blocks of his linkages . it is well known that the configuration curves of these linkages consists of two irreducible algebraic components that intersect at flat folded positions . that is , the linkage may switch between parallelogram and anti - parallelogram mode , thus entering unwanted components of the configuration curve . the effect is that kempe s linkages draw more than the originally intended curve . at the cost of introducing additional links and joints , this defect can be overcome by the `` bracing constructions '' of @xcite . the approach of @xcite uses anti - parallelograms and is therefore subject to the same defect and its resolution . the basic building blocks of our approach are bennett linkages whose configuration curve has only one irreducible component . thus , no additional links and joints are needed to prevent the linkage from switching modes . we also capture this in the phrase `` the configuration curve is free of spurious components '' . this article uses several results from other , recently published , papers . whenever we use such a result we give a concise summary and references . moreover , we provide algorithmic descriptions so that a reader of this paper will be able to construct linkages for drawing an arbitrary rational space curve . we continue this article with a formulation of the main theorem and an overview of its proof in . in we provide a concise introduction to dual quaternions and motion polynomials . the proof of our main theorem is done in . it is subdivided into several steps : construction of a minimal motion to the given rational curve , factorization of this motion , and subsequent linkage construction by means of `` bennett flips '' . in we present several examples to illustrate important points of our construction . our main result in this paper is a statement about linkages and bounded rational curves in three - space . a rational curve is a curve admitting a parametric equation of the shape @xmath9 with polynomials @xmath10 $ ] and @xmath11 . it is no restriction to assume that this parametric equation is _ reduced , _ that is , @xmath12 because we may always divide by a common factor . the _ circularity _ of a reduced parametric equation is @xmath13 . it counts the number of intersection points with the absolute conic of euclidean geometry , is a positive integer and invariant with respect to rational re - parameterizations and similarity transformations . finally , the rational curve is called _ bounded , _ if @xmath14 has no real zeros and @xmath15 . note that in this sense bounded segments of unbounded rational curves are bounded rational curves . a linkage in our context consists of a set of lines in space , called the _ joints _ , and links ( rigid bodies which could have different shapes ) that connect two or more axes . whenever two links are connected by a joint , their relative position is constrained to a rotation about their common joint . more precisely , the relative position is determined by the rotation angle about this joint with respect to a given reference configuration . the set of all tuples of possible rotation angles is called the linkage s _ configuration space . _ if it is of dimension one , we call it a _ configuration curve _ and say the linkage has _ one degree of freedom . _ if this is the case , we designate one link as fixed and another as moving and call them _ fixed _ and _ moving frame , _ respectively . we view the relative displacements of the moving frame with respect to the fixed frame as a curve in @xmath16 . the _ orbit _ or _ trajectory _ of a point attached to the moving frame is the locus of all positions in space when the point is subject to all possible displacements in this curve . a linkage is called _ spherical , _ if all axes are concurrent and _ planar _ if they are parallel . the trajectories of spherical linkages are spherical curves and the trajectories of planar linkage are curves in parallel planes . for a more formal definition of ( at least planar ) linkages we refer to @xcite . let us illustrate some of these concept at hand of a bennett linkage @xcite which will play a crucial role in our linkage construction in . a bennett linkage is a spatial four - bar linkage with one degree of freedom . its four axes @xmath17 , @xmath18 , @xmath19 , @xmath20 are the perpendiculars to two incoming edges in the vertices of a spatial parallelogram , that is , a spatial quadrilateral with equal opposite edge lengths . displays an abstract representation and a 3d model of a bennett linkage . the joint axes @xmath17 and @xmath20 are in the fixed frame while @xmath18 and @xmath19 are in the moving frame . two things about bennett linkages are important to us : * bennett linkage constitute the only type of spatial four - bar linkages with one degree of freedom and * their configuration curve consists of only one component . img / bennett ( 14,45)@xmath17 ( 16,19)@xmath18 ( 38,3)@xmath19 ( 49,36)@xmath20 ( 58,45)@xmath17 ( 61,19)@xmath18 ( 85,3)@xmath19 ( 94,36)@xmath20 here is our main result : [ th:1 ] for every bounded rational curve of degree @xmath0 and circularity @xmath1 in three space , there exists a spatial revolute linkage with one degree of freedom with at most @xmath7 links and @xmath8 joints such that the trajectory of one point attached to the moving frame is precisely the given rational curve . before we embark on a proof of this theorem ( which will consume large parts of this paper ) , we would like to make a few remarks : * the theorem talks about _ bounded _ rational curves . this is necessary because all trajectories of a linkage with only revolute joints are bounded . it is , however , admissible that the rational curve parameterizes a bounded portion of an unbounded curve , for example a line segment . * boundedness also implies that @xmath0 is even and @xmath1 is an integer so that our bounds on the number of links and joints are integers as well . * the linkage of is not unique . in fact , uncountably many linkages exist . even planar or spherical curves may be drawn by spatial linkages . * however , our approach is capable of producing planar linkages for planar curves ( ) . thus , improves the bound of @xmath21 links and @xmath22 joints given by @xcite even for non - circular ( @xmath23 ) planar rational curves . * our approach can also produce spherical linkages for spherical curves . for them , the number of links and joints reduces to @xmath24 and @xmath25 , respectively ( ) . * the linkages can be constructed in such a way , that their configuration curve is free of spurious components . this is however not the case if the linkage is required to be planar or spherical . our proof of is constructive and can be translated into algorithms , mostly based on polynomial algebra over the dual quaternions . our presentation pays attention to these algorithmic approach and gives ample information for actual implementation . the proof consists of several steps . we begin by constructing a rational motion such that one point has the given rational curve as trajectory . in order to keep small the number of links and joints , it is advantageous to require a minimal motion degree in the dual quaternion model of rigid body displacements . in @xcite , we proved that these properties determine a unique rational motion which may be computed in rather straightforward manner by . this rational motion is parameterized by a certain polynomial with dual quaternion coefficients which we call a `` motion polynomial '' . general motion polynomials can be written , in several ways , as products of linear motion polynomials @xcite . these factorizations correspond to the decomposition of the rational motion into products of rotations . the axes are determined by the linear motion polynomials and the rotation angles are linked by a common parameter . on these open chains of revolute joints we base our linkage construction . unfortunately , there are motion polynomials that do not allow factorizations . even worse , the motion polynomials of minimal degree with prescribed trajectory typically fall into this category . therefore , we have to artificially raise their degree in such a way that factorizations exist . this can be done by multiplying them with real polynomials as in @xcite . this does not change the underlying motion and may be advantageous in certain applied situations . however , the bounds in are only obtained by a refinement of this procedure . we right - multiplying the motion polynomial with a certain quaternion polynomial . this changes the motion but not the trajectory in question . a curious side - effect of this approach is that it may turn a planar or spherical motion into a spatial motion . having constructed a factorizable motion polynomial to the prescribed trajectory , we have to construct a linkage . in general , it is possible to combine the open chains obtained from different factorizations to form a linkage with one degree of freedom @xcite . we do , however , not pursue this approach because , it seems difficult to prove that the resulting linkage has _ always _ ( not just in general ) only a single degree of freedom . moreover , spurious motion components do exist @xcite . instead , we adapt the scissor - linkage construction of @xcite to the spatial case , replacing the anti - parallelograms of planar linkages by spatial bennett linkages . this automatically guarantees that the configuration space is of dimension one and has no spurious components . this section provides an introduction to dual quaternions and their relation to space kinematics . in particular , we introduce a homomorphism between a certain subgroup of dual quaternions into @xmath16 and the important concept of motion polynomials . denote by @xmath26 the non - commutative ring of quaternions . an element @xmath27 can be written as @xmath28 . the quaternion units @xmath29 , @xmath30 , and @xmath31 satisfy the multiplication rules @xmath32 by @xmath33 we denote the ring @xmath34/\langle { \varepsilon}^2 \rangle$ ] . its elements are called _ dual numbers . _ the scalar extension @xmath35 of @xmath26 by @xmath33 gives the ring of dual quaternions . a dual quaternion @xmath36 may be written as @xmath37 with quaternions @xmath38 and @xmath39 , the _ primal _ and _ dual part _ of @xmath36 , respectively . the conjugate dual quaternion is @xmath40 and quaternions in @xmath26 are conjugated by multiplying the coefficients of @xmath29 , @xmath30 , and @xmath31 with @xmath41 . the dual quaternions with non - zero primal part are invertible . the inverse of @xmath36 is @xmath42 . here , @xmath43 denotes the _ dual quaternion norm . _ it is a dual number and , provided @xmath44 , the inverse of number @xmath45 is @xmath46 . denote by @xmath47 the multiplicative subgroup of dual quaternions with real , nonzero norm . it acts on @xmath48 according to @xmath49 this equation defines a homomorphism from @xmath47 to @xmath16 . it is surjective and the kernel is the real multiplicative group @xmath50 . now we make this group homomorphism parametric . denote by @xmath51 $ ] the skew ring of polynomials over @xmath52 with indeterminate @xmath53 . we define multiplication in this ring by the convention that @xmath53 commutes with all coefficients . this is a natural convention because @xmath53 will later act as a real motion parameter and @xmath54 is in the center of @xmath52 . some notions that have already been defined for @xmath52 can be transferred to @xmath51 $ ] . for @xmath55 $ ] the _ conjugate polynomial _ @xmath56 is obtained by conjugating all coefficients of @xmath57 . if @xmath58 with @xmath59 $ ] , then @xmath60 and @xmath61 are called _ primal _ and _ dual part , _ respectively . the norm polynomial is @xmath62 . its coefficients are dual numbers . if @xmath63 , the value of @xmath57 at @xmath64 is defined as @xmath65 . with these definitions , evaluation of polynomials at a fixed value @xmath64 is not a ring homomorphism . still , the dual quaternion zeros of polynomials over @xmath52 have a meaning in our algorithms . the polynomial @xmath55 $ ] is called a _ motion polynomial , _ if @xmath66 \setminus \{0\}$ ] and if its leading coefficient @xmath67 is invertible . motion polynomials are a central concept in this article . their introduction is motivated by the fact that for every @xmath68 , the value @xmath69 is an element of @xmath47 so that @xmath57 acts on @xmath70 . varying @xmath71 , we get a one - parametric set of rigid body displacements , that is , a motion . by virtue of , the orbit of any point is ( part of ) a rational curve , whence the motion itself is called _ rational . _ in order to get the complete rational curve as trajectory , we should extend the parameter range from @xmath54 to @xmath72 . with the definition @xmath73 the trajectories are complete rational curve . of particular importance to us are linear motion polynomials . the linear polynomial @xmath74 is a motion polynomial if @xmath75 is real . this is the case if both @xmath76 and @xmath77 are real . it is well - known that the motion parameterized by @xmath57 is either a rotation about a fixed axis or , if the primal part of @xmath36 is real , a translation in fixed direction . in either case , the parameter value @xmath78 corresponds to the identity transformation , that is , zero rotation angle or translation distance . in this article , we often assume that a given motion polynomial has no nontrivial real factor . we call those motion polynomials _ reduced . _ from a kinematic viewpoint , this is no restriction as @xmath57 and @xmath79 with @xmath80 \setminus \{0\}$ ] parameterize the same motion . note , however , that multiplication with a real polynomial is a useful technique to ensure existence of a factorization ( see @xcite and ) . in this section we prove . for each step in our constructive proof we provide a theoretical justification , often with references to existing literature , and an algorithmic description . given is a bounded rational curve by a reduced rational parametric equation @xmath81 . we also encode it as polynomial @xmath82 $ ] . the parametric equation @xmath83 and the polynomial @xmath84 are of the same degree @xmath0 . the first step is the construction of a rational motion such the trajectory of one point equals the parametric curve @xmath84 . an obvious and simple choice is the curvilinear translation along @xmath84 , given by the motion polynomial @xmath85 . however , in order to keep low the number of links and joints , we try to find a motion polynomial @xmath57 of minimal degree with trajectory @xmath84 . the main result in this context states uniqueness of this motion and characterizes curves for which the trivial curvilinear translation is not optimal @xcite . in order to fully appreciate it , we need a definition : [ def : trajectory - degree ] the _ trajectory degree _ of a reduced rational motion @xmath86 $ ] is the maximal degree of its trajectories . the _ quaternion degree _ is the degree of @xmath57 as polynomial in @xmath51 $ ] . the _ spherical degree defect _ of @xmath57 is the degree of the real polynomial factor of maximal degree of the primal part @xmath60 . some authors refer to the trajectory degree of a reduced rational motion as just the `` degree '' . however , we have to distinguish between this trajectory degree and the degree of @xmath57 as polynomial in @xmath51 $ ] . the latter was called the motion s `` quaternion degree '' in @xcite and we follow this convention . the spherical degree defect accounts for a difference in the respective trajectory degrees of @xmath57 and its spherical motion component @xmath60 . if the motion polynomial is monic ( or at least its leading coefficient is invertible ) , the spherical degree defect can be computed as degree of @xmath87 where @xmath88 denotes the monic real polynomial factor of maximal degree . [ th:2 ] the rational motion of minimal quaternion degree with a prescribed rational trajectory is _ unique . _ if the trajectory is of degree @xmath0 and circularity @xmath1 , this minimal motion is of degree @xmath89 and has a spherical degree defect of @xmath90 . lets call the motion of the trajectory s _ minimal motion . _ tells us that in general ( if the trajectory is of circularity zero ) , the curvilinear translation is minimal . but for curves with positive circularity , we can do better . this has effects on our linkage construction : curves of high circularity lead to minimal motions of low degree and low spherical degree defect and lend themselves well to a realization by a linkage with few links and joints . the constructive proof of in @xcite can be turned into a short algorithm to actually compute minimal motions . we make a few technical assumptions : * the reduced rational curve @xmath91 satisfies @xmath92 . this can always be accomplished by a suitable translation of the coordinate frame . * we want to find a motion polynomial @xmath58 of minimal degree @xmath93 such that @xmath94 . in view of this means that @xmath84 is the trajectory of the affine origin in @xmath70 . * we assume that @xmath57 is monic which is consistent with the assumption @xmath95 and entails that @xmath96 is the identity . this can always be accomplished by a suitable rotation of the coordinate frame about its origin . an important ingredient in the computation of minimal motions is right division of quaternion polynomials . given @xmath97 $ ] , there exist unique polynomials @xmath98 $ ] , called ( right ) quotient and ( right ) remainder , with @xmath99 and @xmath100 . in case of monic @xmath101 , they can be computed by . we denote the right quotient by @xmath102 and the right remainder by @xmath103 . the latter is used in ( euclidean algorithm ) for computing the left gcd of two quaternion polynomials @xmath97 $ ] in case of monic @xmath101 . the left gcd is the unique monic polynomial @xmath104 of maximal degree such that there exist polynomials @xmath98 $ ] with @xmath105 and @xmath106 . the function @xmath107 in line [ line : lcoeff ] of returns the leading coefficient of a polynomial so that @xmath108 is monic . two polynomials @xmath97 $ ] , @xmath101 is monic . polynomials @xmath98 $ ] such that @xmath100 and @xmath99 . @xmath109 , @xmath110 @xmath111 , @xmath112 @xmath113 leading coefficient of @xmath114 @xmath115 [ line : gct]@xmath116 @xmath117 @xmath118 two polynomials @xmath119 , @xmath120 $ ] , @xmath101 is monic . monic polynomial @xmath121 $ ] such that there exist polynomial @xmath98 $ ] with @xmath105 and @xmath106 . @xmath122 @xmath101 [ line : lcoeff]@xmath123 the computation of the minimal degree rational motion with the prescribed trajectory @xmath84 is illustrated in . it computes a monic rational motion polynomial @xmath57 such that the trajectory of the affine origin is parameterized by @xmath84 . the correctness of has been proved in @xcite . the function @xmath124 in line [ line : quo ] denotes the quotient of polynomial division for real polynomials . it may also be computed by . reduced rational parameteric equation @xmath125 with @xmath95 . monic motion polynomial @xmath57 of minimal degree such that @xmath126 ( trajectory of affine origin is parameterized by @xmath84 ) . @xmath127 [ line : quo]@xmath128 quotient of polynomial division in @xmath129 $ ] @xmath130 @xmath131 , @xmath132 , @xmath133 if the rational curve is planar or spherical , then the minimal motion is also planar or spherical , respectively . if a planar curve with a non - planar minimal motion existed , we could reflect the motion in the curve s plane and obtain a contraction to uniqueness . for spherical curves , this follows from folgerung 9 and the proof of satz 6 of @xcite . let s look at some examples of minimal motions : [ ex:1 ] consider the parameteric curve @xmath134 with @xmath135 . for @xmath136 , it is an ellipse and the minimal motion polynomial is @xmath137 . it parameterizes the curvilinear translation along the ellipse . if @xmath138 , the curve degenerates to a straight line segment and the minimal motion is the translation @xmath139 ( back and forth along this segment ) . if @xmath140 , the parametric curve is a circle . its circularity is one and the minimal motion polynomial @xmath141 is linear , as predicted by . it parameterizes the rotation around the circle axis . [ ex:2 ] as a second example , consider viviani s curve @xmath142 , given by @xmath143 it lies on the sphere of radius @xmath144 with center @xmath145 and our parameterization satisfies @xmath95 . the minimal motion is @xmath146 . it is a spherical motion because it fixes the sphere center @xmath145 . we call it _ viviani motion . _ having constructed a monic motion polynomial @xmath57 with trajectory @xmath84 , we have converted our trajectory generation problem to a motion generation problem : we are looking for a linkage with one degree of freedom such that one link follows the motion parameterized by @xmath57 . this we do by decomposing the motion into the product of rotations about certain axes . the rotation angles are linked by the common motion parameter @xmath53 . the basic tool for this is the factorization theory for motion polynomials as introduced in @xcite and its extension to non - generic cases in @xcite . [ def : factorization ] a bounded motion polynomial @xmath57 is said to _ admit a factorization _ if there exist bounded linear motion polynomials @xmath147 , , @xmath148 such that @xmath149 we already said that each linear factor @xmath150 in parameterizes a rotation with fixed axes or translation in fixed direction . because @xmath150 is bounded , the later can not occur here . hence , the product parameterizes the composition of such rotations . note that a factorization of the shape need not exist and if it exists , it need not be unique . call a motion polynomial `` generic '' if its spherical degree defect ( ) is zero . a generic motion polynomial can always be factored ( * ? ? ? * theorem 1 ) but the factorization is in general not unique . elliptic and circular translation and the viviani motion serve as examples : [ ex:3 ] the elliptic translation of does not admit a factorization if @xmath151 . if @xmath140 , it admits infinitely many factorizations @xmath152 with @xmath153 , @xmath154 . the viviani motion in admits only the factorization @xmath155 by considering primal parts only ( the dual parts are merely there because the sphere center is not the affine origin ) , we see that it is the composition of rotations with equal angular speed about the second and the third coordinate axis . an algorithm for computing factorizations of generic motion polynomials has been presented in @xcite . it is displayed in and . the non - uniqueness of the factorization comes from the undetermined order of the quadratic factors in line [ line : quadfac ] of . in line [ line : lquo ] we compute the quotient of polynomial left division . this can be done by a variant of but with @xmath156 in line [ line : gct ] replaced by @xmath157 . monic , bounded motion polynomial @xmath158 $ ] , quadratic factor @xmath159 of @xmath160 . bounded linear motion polynomial @xmath161 such that @xmath162 . @xmath163 @xmath164 with @xmath165 . [ li : zero]@xmath166 unique zero of @xmath114 @xmath167 ( @xmath168 is invertible ) @xmath36 . @xmath169 $ ] , a generic , monic motion polynomial . a list @xmath170 $ ] of bounded linear motion polynomials such that @xmath171 . @xmath172 $ ] initialize empty list [ line : quadfac]@xmath173 $ ] each @xmath174 $ ] , @xmath175 is a + quadratic , irreducible factor of @xmath176 $ ] . @xmath177 $ ] . remove @xmath178 from list @xmath119 . [ line : czero]@xmath179 @xmath180 + l$ ] add @xmath181 to start of list . [ line : lquo]@xmath182 quotient of left division , variant of . @xmath183 . [ ex:4 ] can not be used to factor the elliptic or circular translation of examples [ ex:1 ] and [ ex:3 ] because its primal part @xmath184 is real . it fails in line [ li : zero ] of where the dual quaternion @xmath168 is not invertible . we can , however , use to obtain the factorization of the viviani motion . because of @xmath185 the quadratic factors of @xmath160 are all equal ( @xmath186 ) and the output of is , indeed , unique . our linkage construction requires existence of at least one factorization of the motion polynomial @xmath57 . thus , we have to find a way `` factorize '' non - generic motion polynomials as well . one possibility to do this has been presented in @xcite . there , we showed that for every bounded motion polynomial @xmath58 of degree @xmath4 a polynomial @xmath80 $ ] of degree @xmath187 exists such that @xmath79 admits a factorization . this statement for planar motion polynomials this has already been proved in @xcite . although not reduced , the motion polynomial @xmath79 parameterizes the same motion as @xmath57 . thus , at the cost of possibly doubling the degree of @xmath57 , we can find a factorized representation to base our linkage construction upon . in fact , the worst case @xmath188 occurs for curvilinear translations , that is , for generic ( non - circular ) trajectories . multiplication with @xmath80 $ ] changes the algebraic properties of the motion polynomial but not the motion itself . however , for our purpose it is sufficient to preserve just one particular trajectory . therefore , the following refinement is conceivable . if @xmath189 $ ] is a quaternion polynomial , the orbit of the affine origin with respect to @xmath190 is the same as the orbit with respect to @xmath57 . indeed , if @xmath58 , this orbit is parameterized by @xmath191 but , because of @xmath192 $ ] , also by @xmath193 in order to retain the orbit of a point different from the affine origin , we can similarly right - multiply with a motion polynomial that fixes that point . besides the extension to three - dimensions , a major contribution in this article over @xcite is to base our linkage construction on factorizations of @xmath190 instead of @xmath79 , thus improving the bounds on the number of links and joints , even in the planar case . but before doing so we have to answer three questions : * given a motion polynomial @xmath57 , does there always exist a quaternion polynomial @xmath194 such that @xmath190 admits a factorization ? * assuming a positive answer to the previous question : how can we determine @xmath189 $ ] ? * finally , how can we compute the factorization of @xmath190 . note that will not do unless it can already be used to factorize @xmath57 ( in which case we simply have @xmath195 ) . the first question will be positively answered in , below . its proof answers the second and third question ( and ) . our solution is based on ( part of ) the factorization algorithm of @xcite for non - generic motion polynomials . we are going to present just its relevant details . the factorization algorithm of @xcite incrementally builds a polynomial @xmath80 $ ] and populates lists @xmath196 $ ] , @xmath197 $ ] of bounded linear motion polynomials such that @xmath198 . it is displayed in but we omit an else - branch that we do not need hereafter . @xmath58 , a monic , bounded motion polynomial ; @xmath80 $ ] , lists @xmath199 , @xmath200 of linear , bounded motion polynomials . initially , @xmath201 , @xmath202 $ ] @xmath203 $ ] . @xmath80 $ ] , @xmath196 $ ] , @xmath204 $ ] such that @xmath198 . [ line : gfactor ] @xmath114 , @xmath199 , @xmath205 use @xmath206 of . @xmath207 [ line : ifgcd ] [ line : p1 ] @xmath208 quadratic real divisor of @xmath209 @xmath210 , @xmath211 [ line : prpl ] @xmath212 , @xmath213 @xmath214 @xmath215 $ ] @xmath216 @xmath217 @xmath218 + f_r$ ] @xmath219 [ line : endif ] [ line : else ] irrelevant for this paper . the important part is in lines [ line : p1][line : prpl ] . note that boundedness implies that @xmath209 in line [ line : p1 ] is strictly positive . hence , its real quadratic divisor @xmath220 is irreducible over @xmath54 . we compute rotation quaternions @xmath221 , @xmath222 that are zeros of @xmath220 ( see , below ) and polynomials @xmath223 , @xmath224 such that @xmath225 depending on the respective degrees of @xmath223 and @xmath224 , we either augment the list of left factors or the list of right factors and recursively call @xmath226 with new arguments . note , however , that the polynomial factor @xmath114 remains unchanged in these parts of . ( it changes in the `` else''-branch in line [ line : else ] which is irrelevant for this paper : once the `` if''-condition in line [ line : ifgcd ] is satisfied , the recursive never enters that `` else''-branch , see below . ) [ th:3 ] given a bounded monic motion polynomial @xmath86 $ ] there exists a polynomial @xmath189 $ ] of degree @xmath227 such that @xmath190 admits a factorization . the bound on the degree of @xmath194 given by is strictly lower than the bound on the degree of the real factor produced by ( the complete version of ) unless @xmath228 . this is the generic case , we trivially have @xmath229 and factorization is passed on to . our proof of uses the following auxiliary result : [ lem:1 ] the motion polynomial @xmath58 with @xmath230 $ ] admits a factorization if @xmath231 divides @xmath232 . the lemma s claim follows from inspection of . in lines [ line : p1][line : endif ] we compute quaternions @xmath221 , @xmath222 and quaternion polynomials @xmath223 , @xmath233 , @xmath224 , @xmath234 such that @xmath235 either @xmath236 or @xmath237 are used in the next iteration of . both cases are similar , so lets focus on the `` left '' case . the quadratic factor @xmath220 of @xmath231 and @xmath232 can be factored as @xmath238 . thus , if we write @xmath231 as @xmath239 , then @xmath240 and @xmath241 . because of @xmath242 @xmath243 still divides @xmath244 . iterating this procedure , we reduce the degree of the primal part s maximal real polynomial factor to zero , whence the factorization is completed by a call to @xmath245 in line [ line : gfactor ] of . for the proof of some facts about quaternion zeros of polynomials are useful . to begin with , a polynomial @xmath246 $ ] of degree @xmath4 and without non - trivial real factors can have at most @xmath4 quaternion zeros . if @xmath60 has real factors , the situation is slightly more complicated . however , we will only encounter the simpler situation that @xmath231 is a _ strictly positive _ real polynomial , i.e. , it is irreducible over @xmath54 . then the quaternion zeros of @xmath231 are precisely the quaternion zeros of the irreducible , real , quadratic factors of @xmath231 and these are well known : [ lem:2 ] if the real polynomial @xmath247 is irreducible over @xmath54 ( @xmath248 ) , the set of its quaternion zeros is @xmath249 one consequence of is existence of a unique zero whose vector part ( the projection onto @xmath250 ) equals a positive multiple of an arbitrary non - zero vector @xmath251 . we call this a zero in _ direction of @xmath252 . _ in view of , we have to determine @xmath189 $ ] such that @xmath253 divides @xmath254 . set @xmath255 , @xmath256 and @xmath257 . there exist polynomials @xmath258 $ ] such that @xmath259 and @xmath260 . because @xmath57 is bounded , @xmath101 is strictly positive and the same is true for @xmath261 . we select @xmath194 subject to the constraints @xmath262 and @xmath263 . the second is a equivalent to @xmath264 . this is always possible : write @xmath265 with quadratic , irreducible polynomials @xmath266 and initiate an iteration procedure ( ) . allows us to pick a quaternion zero @xmath267 for a quadratic factor @xmath178 such that @xmath268 is not among the finitely many zeros of @xmath269 . after / before each iteration , we have @xmath270 . after @xmath271 iterations , we have @xmath272 and the claim follows . @xmath58 , a monic , bounded motion polynomial . @xmath189 $ ] such that @xmath190 admits a factorization . @xmath273 , @xmath274 , @xmath275 , @xmath276 @xmath277 $ ] where each @xmath178 is an irreducible quadratic factor of @xmath261 @xmath278 , @xmath279 , @xmath280 [ line : pickhi]@xmath281 quaternion zero of @xmath178 , @xmath282 [ line : pickh]@xmath283 and @xmath284 @xmath194 at several points in and we are free to pick one among infinitely many zeros of irreducible quadratic polynomials . this freedom is quite advantageous for applications of our algorithm in order to fulfull engineering needs . [ ex:5 ] we continue and illustrate algorithms [ alg : afactor ] and [ alg : ffactor ] at hand of the elliptic translation @xmath137 with @xmath285 . by , there exists a linear polynomial @xmath189 $ ] such that @xmath190 admits a factorization . using the notation of , we have @xmath286 and @xmath287 . hence , the only restriction on @xmath288 is @xmath289 . one possible factor is @xmath290 . we have @xmath291 where @xmath292 our theory guarantees that @xmath190 admits a factorization . we use for computing a factorization . because @xmath293 , we enter the `` if''-branch in lines [ line : ifgcd][line : endif ] where we have @xmath294 thus , we have found the representation @xmath295 where @xmath296 upon the next iteration , defers factorization of @xmath236 to the function @xmath206 of . we obtain the lists @xmath297,\\ f_r = \bigl [ t + \frac{2b((a^2-b^2){\mathbf{i}}+ 2ab{\mathbf{k } } ) + { \varepsilon}{\mathbf{j}}(a^2+b^2)^2}{2b(a^2+b^2 ) } , t - \frac{2b{\mathbf{i}}+ ( a^2-b^2){\varepsilon}{\mathbf{j}}}{2b } \bigr ] \end{gathered}\ ] ] which immediately yields the factorization @xmath298 however , there are other choices for @xmath194 . considering the planarity of @xmath57 it is natural to pick @xmath299 . running , we obtain the lists @xmath300 , \quad f_r = [ t - { \mathbf{k}}- \tfrac{1}{2}{\varepsilon}{\mathbf{j}}(a - b ) , t - { \mathbf{k}}+ \tfrac{1}{2}{\varepsilon}{\mathbf{j}}(a + b)].\ ] ] and the rather simple factorization @xmath301 in contrast to the factorization , this is a _ planar motion polynomial . _ one can show that all trajectories of the motion are ellipses ( or line segments ) in non - parallel plane . this is a characterizing property of the _ darboux - motion _ whose factorizations have already been discussed in @xcite . in we showed how to factor a bounded motion polynomial , possibly after right multiplication with a quaternion polynomial , into the product of bounded linear motion polynomials . this section is a little intermezzo before constructing linkages from these factorizations in . it introduces a technique we call `` bennett flip '' . it will be an essential component of our linkage construction . in general , a motion polynomial admits several factorizations , each giving rise to an open chain of revolute joints that is capable of performing the motion parameterized by @xmath57 . the distal joints of different factorizations can be attached to a common link and the resulting multi - looped linkage can still perform the motion @xmath57 . adding sufficiently many factorizations one can reasonably expect to reduce the dimension of the linkages configuration space to one . this idea has already been successfully applied for constructing linkages for engineering applications @xcite but it is not really suitable for proving the general statement of because it may fail to produce linkages with only one degree of freedom . consider , as a warning , the viviani motion . it admits only one factorization and hence gives raise to just one single chain with two revolute joints and two degrees of freedom . it is , however , possible to use above idea in the special case of quadratic motion polynomials . they are sufficiently simple to allow a complete discussion of all unwanted cases . once this is available we can combine degree two motion polynomials and their linkages to generate mechanisms that satisfy the criteria of . [ def : bflip ] the _ bennett flip _ is the map @xmath302 where @xmath303 and @xmath304 . in order to understand the idea behind , set @xmath305 and consider the polynomials @xmath306 , @xmath307 where @xmath308 and @xmath309 . from @xmath310 we see that @xmath311 is the zero of @xmath312 and @xmath313 is the zero of @xmath314 . in other words , @xmath313 and @xmath311 are obtained by applying to @xmath306 and yield the second factorization @xmath315 . this interpretation accounts for the name `` bennett flip '' : given a rotation quaternion @xmath36 , denote its axes by @xmath316 . in general the axes @xmath317 , @xmath318 , @xmath319 , and @xmath320 form a bennett linkage . exceptional cases exist and will be described in detail . [ prop : bflip ] consider two rotation quaternions @xmath321 , @xmath322 with @xmath323 and let @xmath324 . then the following hold : 1 . also @xmath313 and @xmath311 are rotation quaternions . the restriction of the bennett flip to pairs of rotation quaternions is an involution , that is , @xmath325 . the restriction of the bennett flip to pairs of rotation quaternions is birational . we have @xmath326 and @xmath327 where @xmath328 is the _ minimal polynomial _ of @xmath64 . we have @xmath329 , @xmath330 and @xmath331 . items 1 and 2 follow from the interpretation of the bennett flip as application of to @xmath306 . item 2 implies item 3 . item 4 follows again from because @xmath332 in line [ line : czero ] of that algorithm . in order to see the last item , we may multiply the equation @xmath333 with @xmath334 from the left and with @xmath335 from the right . with @xmath336 we obtain @xmath337 . because @xmath338 commutes with the other factors , the first equation of item 5 follows . the other equations follow by iterating this argument . the statements of on the restriction of @xmath339 to pairs of rotation quaternions are also true for @xmath339 itself . here , we only need the weaker formulation that allows a simpler proof . with @xmath305 we have @xmath333 . this ensures that the four - bar linkage with axes @xmath317 , @xmath318 , @xmath319 , and @xmath320 moves with at least one degree of freedom . moreover , it is elementary to see that it moves with at most one degree of freedom , if these axes are all different . if this is not the case , we have @xmath340 and @xmath341 ( this comprises the case where all axes coincide ) . the following lemma provides a sufficient criterion to exclude this . [ lem:3 ] given are rotation quaternions @xmath321 , @xmath342 , @xmath313 , @xmath311 with @xmath343 . * we have @xmath344 if and only if @xmath345 and @xmath346 are linearly dependent . in this case , @xmath347 . * provided @xmath348 , we have @xmath340 and @xmath341 if and only if @xmath349 . the dual quaternion @xmath350 has zero scalar part . hence , there exist real numbers @xmath351 such that @xmath352 and the plcker line coordinates of @xmath317 are @xmath353 $ ] . this implies the first claim . now we turn to the second claim . by it is obvious that @xmath349 implies equality of axes . conversely , we have to show @xmath354 and @xmath355 for @xmath356 under the assumption of @xmath340 and @xmath341 . this equality of axes implies linear dependence of their plcker vectors . hence , there exist real numbers @xmath357 , @xmath358 , @xmath359 , @xmath360 such that @xmath361 and @xmath362 . by comparing coefficients on both sides of @xmath333 we obtain @xmath363 the first equation minus its conjugate is @xmath364 . because @xmath365 and @xmath366 are linearly independent , this implies @xmath367 . but then , we are left with @xmath368 and @xmath369 follows . thus , @xmath370 and @xmath371 . but then .4 implies @xmath372 . [ rem : flip - mobility ] summarizing the result of , we can say that the four - bar linkage with axes @xmath317 , @xmath318 , @xmath319 , and @xmath320 has exactly one degree of freedom if @xmath350 and @xmath346 are linearly independent and @xmath373 . under the assumptions of the previous remark , a four - bar linkage obtained by a bennett flip is of one of the following types : * a bennett linkage , * a planar anti - parallelogram linkage , or * a spherical linkage where the connections of opposite joint pairs are of equal length . the spatial case is clear since bennett linkages are the only movable spatial four - bar linkages . in particular , the orthogonal distances and angles of opposite axes pairs are equal . the statement on the planar case has been proved in @xcite . the spherical situation is similar but a distinction between parallelogram and anti - parallelogram linkages is neither possible nor necessary . the reason is that a joint may equally well be realized by a circular arc or its supplementary arc . equal distance of joint pairs follows from projection of the spatial case on the primal part , that is , by considering only the spherical motion component . at any rate , the configuration curve of planar or spherical linkages consists of two irreducible components and only one of them is relevant to us . [ rem : circular - flip ] it is worth mentioning that techniques similar to bennett flips are conceivable . for the circular translation @xmath374 one can find infinitely many different factorizations and two of them may be combined , similar to the bennett flip , to form a parallelogram linkage . this was used to construct new overconstrained 6r linkages in @xcite . now we come to the actual construction of a linkage last step in our proof of . in the preceding sections we have shown how to 1 . compute a motion polynomial @xmath58 of minimal degree @xmath93 ( where @xmath0 is the curve s degree and @xmath1 is its circularity ) such that the trajectory of the affine origin equals the given trajectory and 2 . determine a quaternion polynomial @xmath189 $ ] of degree @xmath375 such that @xmath190 admits the factorization @xmath376 with rotation quaternions @xmath377 and @xmath378 . this factorization gives rise to an open chain of revolute axes that can generate the motion parameterized by @xmath190 in the following way . * for @xmath379 , the quaternion @xmath380 describes a rotation about the axis @xmath381 whose plcker coordinates @xmath382 $ ] can be computed from @xmath383 . * the lines @xmath384 determine the configuration of the linkage at parameter time @xmath78 ( zero rotation angle ) . the configuration at parameter time @xmath385 is obtained from this configuration by successively subjecting @xmath386 to the rotation @xmath387 for @xmath388 . this open revolute chain has @xmath4 degrees of freedom . it is our aim to constrain its motion in such a way that a link attached to the last joint ( with axis @xmath389 ) performs a motion that can be parameterized by @xmath190 . the basic technique for doing this is the bennett flip of . we can constrain the motion of this chain to the motion parameterized by @xmath190 by 1 . picking a `` suitable '' motion polynomial @xmath390 and 2 . recursively defining @xmath391 this gives us rotation quaternions @xmath392 , @xmath393 , and @xmath394 that can be assembled to a linkage whose link graph is depicted in . the vertices of the linkgraph correspond to links and the edges correspond to joints . two vertices ( links ) are connected by an edge ( joint ) if relative motion of the two links is constrained by the corresponding revolute joint . the recursion builds the linkage from the bottom row of edges ( labeled @xmath321 , , @xmath395 ) and the first vertical edge ( labeled @xmath396 ) but it is clear that we may equally well start from any other vertical edge @xmath397 , @xmath379 . also shows the joint hypergraph ( not a graph because the joint triples @xmath398 and @xmath399 belong to the same links for @xmath400 ) . the structure of this hypergraph suggests the name `` scissor linkage '' for the resulting linkage type . the axes of @xmath321 and @xmath396 are attached to the base frame , while @xmath395 and @xmath401 are attached to the moving frame . their link performs the motion parameterized by the given motion polynomial . in general , each loop @xmath380 , @xmath397 , @xmath402 , @xmath403 forms a bennett linkage . in this case the linkage has just one degree of freedom and its configuration curve has a single component . in other words , it satisfies all requirements of . it is , however , possible , that planar or spherical four - bar linkages occur . this may be desirable , acceptable , or not acceptable , depending on circumstances . in order to prove we must , however , avoid coinciding axes , that is , we have to show that one can pick @xmath396 in such a way that the conditions of are never fulfilled . + [ lem:4 ] given rotation quaternions @xmath404 there exists a rotation quaternion @xmath396 such that the quaternions @xmath405 and @xmath393 obtained from the recursion are well - defined and the four - bar linkage with axes @xmath406 , @xmath381 , @xmath407 , @xmath408 has precisely one degree of freedom for any @xmath409 . by , we have to avoid linear dependency of @xmath410 and @xmath411 as well as equality @xmath412 of minimal polynomials . by we have @xmath413 . thus , in order to ensure different minimal polynomials , we just have to avoid the set @xmath414 when picking @xmath396 . this is the union of at most @xmath4 algebraic sets of positive codimension . in order to avoid linearly dependent vector parts , we recursively define maps @xmath415 for @xmath379 . they are well - defined birational maps by and satisfy @xmath416 . the vectors @xmath417 and @xmath411 are linearly dependent if the projection of @xmath397 onto the vector part is in the span of the projection of @xmath380 onto its vector part . this defines a vector subspace @xmath418 of dimension three that @xmath397 should avoid . this is the case if and only if @xmath396 avoid the union @xmath419 , that is , another finite union of varieties of positive codimension . this is certainly possible . this finishes our proof of . we just want to explain how to bound the number of links and joints . the minimal degree motion polynomial with a given rational curve of degree @xmath0 and circularity @xmath1 as trajectory is of degree @xmath93 and spherical degree defect @xmath90 ( ) . by , there exist a polynomial @xmath189 $ ] of degree not larger than @xmath420 such that @xmath190 admits a factorization . the product @xmath190 is of degree @xmath421 and this is the same number @xmath4 as in . the numbers of links and joints are the numbers of vertices and edges , respectively , in the linkgraph , that is , @xmath422 and @xmath423 , respectively . we conclude this section with a discussion of planar and spherical four - bar linkages as components of the scissor linkage . the summary is that they can be avoided if wanted but can also be enforced if the given rational curve is planar or spherical . [ cor:1 ] under the assumptions of we may pick @xmath396 in such a way that none of the quadruples @xmath424 is planar or spherical . in particular , our construction yields linkages such that the configuration space is free of spurious components . the quadruple will be spherical or planar , respectively , if and only if the axes of @xmath403 and @xmath380 intersect or are parallel . similar as in the proof of , this defines subvariaties of positive codimension that @xmath403 should avoid for @xmath425 . mapping back this subvariaties via the maps @xmath426 adds finitely many further components of positive codimension to the set that @xmath396 should avoid . shows that we can avoid planar or spherical quadrilaterals in the scissor linkage if we wish . this ensures that the configuration space of the resulting linkage is free from spurious components . this is an important difference to planar versions of kempe s universality theorem where `` bracing constructions '' for anti - parallelograms are necessary in order to suitable constrain the configuration space . of course , this then yields spatial linkages even for planar or spherical motions which may not always be desirable . the next corollary states that it is also possible to generate planar or spherical linkages if the input data is suitable . [ cor:2 ] if , under all assumptions of , the axes of @xmath392 are incident with the same point or are parallel to the same direction , we may pick @xmath396 in such a way that all axes of @xmath405 and @xmath393 are incident with the this point or parallel to this direction , respectively . for the planar case , this is ( * ? ? ? * lemma 6.5 ) . as to the spherical case , assume , without loss of generality that all dual parts are zero . the proof of shows that we can pick @xmath427 which then implies @xmath428 and the claim follows . finally , we mention two ways to improve our bounds on the number of links and joints . the first is just a hint that will often work for spatial linkages without a formal proof . the second is a substantial improvement of kempe s universality theorem for rational spherical curves . if two successive four - bars @xmath381 , @xmath407 , @xmath408 , @xmath406 and @xmath429 , @xmath430 , @xmath431 , @xmath407 are in a scissor linkage are both bennett linkages , it is , in general , possible to eliminate their common joint @xmath397 without increasing the mobility . the two bennett linkages are replaced by a closed - loop linkage with six revolute joints and one degree of freedom . ( it is of type `` waldron s double bennett hybrid '' @xcite . ) we refrain from any attempts to formalizing this idea because it is a little tricky to guarantee that neither the number of irreducible components nor dimension of the configuration space increases in non - generic situations . [ cor : spherical ] a spherical rational curve of degree @xmath0 appears as trajectory of linkage with at most @xmath24 links and @xmath25 joints . the statement follows from the bounds of and the observation that spherical rational curves are of maximal circularity @xmath432 . in this section we present several examples of linkages to given rational curves . they are meant to illustrate type and properties of the linkages we obtain , the degrees of freedom we have in our construction , and how to use them in the design of linkages . some of them continue previous examples . [ ex : pascal ] the limacon of pascal , parameterized by @xmath433 where @xmath434 has the minimal motion @xmath435 because of @xmath185 , this is the only factorization . we select @xmath436 and , using the recursion , obtain @xmath437 the resulting linkage for @xmath438 is depicted in , left and bottom . in this case , the curve is a cardioid . the joints are labeled by their corresponding rotation quaternions . the special geometry allows to realize the links connecting @xmath321 , @xmath342 , @xmath439 and @xmath313 , @xmath311 , @xmath439 , respectively , by straight line segments . the linkage has two flat positions where all joints are collinear . there , either of the two anti - parallelograms ( @xmath380 , @xmath397 , @xmath402 , @xmath403 for @xmath356 ) can switch to parallelogram mode . thus , the motion has four components and only one of them is relevant for drawing the cardioid . using @xmath440 , we obtain a different , less symmetric , linkage to draw the same curve ( , right ) . here , the triangles @xmath321 , @xmath439 , @xmath342 and @xmath313 , @xmath439 , @xmath311 act as links , that is , they are rigid throughout the motion . the motion of the link attached to @xmath342 and @xmath441 is the same for both linkages . ' '' '' ' '' '' we continue the discussion of viviani s curve . however , we not use but the simpler motion polynomial @xmath442 which is obtained from by a translation . based on this minimal motion we construct a linkage . we initialize the construction with @xmath443 and , using the recursion , compute @xmath444 it is already apparent that this linkage has a flat folded position in the plane spanned by @xmath30 and @xmath31 . one of its configurations is depicted in , left . we may as well construct a different linkage by starting with @xmath445 whence we get @xmath446 the corresponding linkage is shown in , right . here , the joint triples @xmath447 and @xmath448 are not collinear . note that maybe more natural choices like @xmath449 or @xmath450 would violate the condition @xmath451 . img / viviani3b ( 50,91 ) ( 17,92 ) ( 59,44 ) ( 28,21 ) ( 80,24 ) ( 40,59 ) ( 96,33 ) img / viviani4b ( 47,86 ) ( 76,89 ) ( 8,66 ) ( 0,55 ) ( 90,29 ) ( 81,70 ) ( 86,9 ) in our next example , we return to the elliptic translation of and start with the planar factorization . we assume @xmath452 , choose @xmath453 and compute @xmath454 the centers of the rotation quaternions in describe the linkage in the configuration at @xmath78 and we can verify that all joints lie on the first coordinate axis . the linkage is quite similar to the linkage of but requires eight links and ten joints and it is similar to the example of @xcite which was constructed in a similar manner . the linkage has two flat positions at which any of the three anti - parallelograms @xmath403 , @xmath402 , @xmath397 , @xmath380 may switch to parallelogram mode . thus , the configuration curve has six components and only one is relevant for drawing the ellipse . we may also construct a spatial linkage based on the factorization and the rotation quaternion @xmath455 . we refrain from displaying the values of @xmath439 , @xmath441 , @xmath456 , @xmath313 , @xmath311 , and @xmath457 ( which would still be possible , even without specifying @xmath168 and @xmath458 ) and rather discuss the resulting spatial linkage ( ) . it shares the linkgraph with the previous example and can be thought of as a scissor linkage made of bennett linkages . displays a schematic representation where the four links and joints of one bennett linkage are displayed in the same color ( blue , gray , and red , respectively ; note that the joints @xmath439 and @xmath441 belong to two bennett linkages ) . the axis triples @xmath459 , @xmath460 , @xmath461 , and @xmath462 are rigidly connected and one point attached to @xmath463 and @xmath456 draws the ellipse . in contrast to all other examples so far , the configuration curve of this linkage is irreducible . one can show that not only one but all trajectories are ellipses ( or line segments ) but in non - parallel plane . thus , the linkage generates a so - called darboux motion @xcite . it is possible to remove @xmath439 or @xmath441 without increasing the dimension of the configuration curve but this may come at the cost of introducing spurious components of the configuration curve . img / elliptic - translation - b ( 5,33 ) ( 25,33 ) ( 15,34 ) ( 1,22 ) ( 49,10 ) ( 62,17 ) ( 79,1 ) ( 89,2 ) ( 95,10 ) ( 93,20 ) finally , we illustrate how to draw a bounded portion of an unbounded rational curve . using with @xmath464 is prevented because then the choice @xmath453 is invalid . but using the curve @xmath465 , the minimal motion @xmath466 , the factor , @xmath299 and the rotation quaternion @xmath467 produces the linkage shown in . this work was supported by the austrian science fund ( fwf ) : p 26607 ( algebraic methods in kinematics : motion factorisation and bond theory ) . j. e. baker , _ on the motion geometry of the bennett linkage _ , proceedings of the 8th international conference on engineering computer graphics and descriptive geometry ( austin , texas , usa ) , 1998 , pp . 433437 . alexander kobel , _ automated generation of kempe linkages for algebraic curves in a dynamic geometry system _ , bachelor s thesis , university of saarbrcken , 2008 , available at https://people.mpi-inf.mpg.de/~akobel/publications/kobel08-kempe-linkages.pdf .

starting from a paper by dziuk and elliott @xcite , much insight into the stability and convergence properties of finite elements on evolving surfaces has been obtained by studying a linear parabolic equation on a given moving closed surface @xmath0 . the strong formulation of this model problem is to find a solution @xmath1 ( for @xmath2 and @xmath3 ) with given initial data @xmath4 to the linear partial differential equation @xmath5where @xmath6 denotes the material time derivative , @xmath7 is the laplace beltrami operator on the surface , and @xmath8 is the tangential divergence of the _ given _ velocity @xmath9 of the surface . we refer to @xcite for an excellent review article ( up to 2012 ) on the numerical analysis of this and related problems . optimal - order @xmath10 error bounds for piecewise linear finite elements are shown in @xcite and maximum - norm error bounds in @xcite . stability and convergence of full discretizations obtained by combining the evolving surface finite element method ( esfem ) with various time discretizations are shown in @xcite . convergence of semi- and full discretizations using high - order evolving surface finite elements is studied in @xcite . arbitrary euler lagrangian ( ale ) variants of the esfem method for this equation are studied in @xcite . convergence properties of the esfem and of full discretizations for quasilinear parabolic equations on prescribed moving surfaces are studied in @xcite . beyond the above model problem , there is considerable interest in cases where the velocity of the evolving surface is _ not given explicitly _ , but depends on the solution @xmath11 of the parabolic equation ; see , e.g. , @xcite for physical and biological models where such situations arise . contrary to the case of surfaces with prescribed motion , there exists so far no numerical analysis for solution - driven surfaces in @xmath12 , to the best of our knowledge . for the case of evolving _ curves _ in @xmath13 , there are recent preprints by pozzi & stinner @xcite and barrett , deckelnick & styles @xcite , who couple the curve - shortening flow with diffusion on the curve and study the convergence of finite element discretizations without and with a tangential part in the discrete velocity , respectively . the analogous problem for two- or higher - dimensional surfaces would be to couple mean curvature flow with diffusion on the surface . studying the convergence of finite elements for these coupled problems , however , remains illusive as long as the convergence of esfem for mean curvature flow of closed surfaces is not understood . this has remained an open problem since dziuk s formulation of such a numerical method for mean curvature flow in his 1990 paper @xcite . in this paper we consider different velocity laws for coupling the surface motion with the diffusion on the surface . conceivably the simplest velocity law would be to prescribe the normal velocity at any surface point as a function of the solution value and possibly its tangential gradient at this point : @xmath14 for @xmath15 , where @xmath16 denotes the outer normal vector and @xmath17 is a given smooth scalar - valued function . this does , however , not appear to lead to a well - posed problem , and in fact we found no mention of this seemingly obvious choice in the literature . here we study instead a _ regularized velocity law _ : @xmath18 with a fixed regularization parameter @xmath19 . this elliptic regularization will turn out to permit us to give a complete stability and convergence analysis of the esfem semidiscretization , for finite elements of polynomial degree at least two . the case of linear finite elements is left open in the theory of this paper , but will be considered in our numerical experiments . the stability and convergence results can be extended to full discretizations with linearly implicit backward difference time - stepping , as we plan to show in later work . our approach also applies to the esfem discretization of coupling a _ regularized mean curvature flow _ and diffusion on the surface : @xmath20 where @xmath21 denotes mean curvature on the surface @xmath22 . the error analysis is further extended to a _ dynamic velocity law _ @xmath23 a physically more relevant dynamic velocity law would be based on momentum and mass balance , such as incompressible navier stokes motion of the surface coupled to diffusion on the surface , where the velocity @xmath9 is determined together with the density @xmath24 and the pressure @xmath25 by the equations @xmath26 we expect that our analysis extends to such a system , but this is beyond the scope of this paper . surface evolutions under navier stokes equations and under willmore flow have recently been considered in @xcite . the paper is organized as follows . in section [ section : problem ] we describe the considered problems and give the weak formulation . we recall the basics of the evolving surface finite element method and describe the semidiscrete problem . its matrix - vector formulation is useful not only for the implementation , but will play a key role in the stability analysis of this paper . in section [ section : main result ] we present the main result of the paper , which gives convergence estimates for the esfem semidiscretization with finite elements of polynomial degree at least @xmath27 . we further outline the main ideas and the organization of the proof . in section [ section : preliminaries ] we present auxiliary results that are used to relate different surfaces to one another . they are the key technical results used later on in the stability analysis . section [ section : uncoupled problem ] contains the stability analysis for the regularized velocity law with a prescribed driving term . in section [ section : coupled problem ] this is extended to the stability analysis for coupling surface pdes and surface motion . in section [ section : geometric estimates ] we briefly recall some geometric estimates used for estimating the consistency errors , which are the defects obtained on inserting the interpolated exact solution into the scheme . section [ section : defect bounds ] deals with the defect estimates . section [ section : proof completed ] proves the main result by combining the results of the previous sections . in section [ section : extensions ] we give extensions to other velocity laws : the regularized mean curvature flow and the dynamic velocity law addressed above . section [ section : numerics ] presents numerical experiments that are complementary to our theoretical results in that they show the numerical behaviour of piecewise linear finite elements on some examples . we use the notational convention to denote vectors in @xmath12 by italic letters , but to denote finite element nodal vectors in @xmath28 and @xmath29 by boldface lowercase letters and finite element mass and stiffness matrices by boldface capitals . all boldface symbols in this paper will thus be related to the matrix - vector formulation of the esfem . we consider the evolving two - dimensional closed surface @xmath30 as the image @xmath31 of a sufficiently regular vector - valued function @xmath32\to { \mathbb{r}}^3 $ ] , where @xmath33 is the smooth closed initial surface , and @xmath34 . in view of the subsequent numerical discretization , it is convenient to think of @xmath35 as the position at time @xmath36 of a moving particle with label @xmath25 , and of @xmath37 as a collection of such particles . to indicate the dependence of the surface on @xmath38 , we will write @xmath39 when the time @xmath36 is clear from the context . the _ velocity _ @xmath40 at a point @xmath41 equals @xmath42 note that for a known velocity field @xmath9 , the position @xmath35 at time @xmath36 of the particle with label @xmath25 is obtained by solving the ordinary differential equation from @xmath43 to @xmath36 for a fixed @xmath25 . for a function @xmath1 ( @xmath2 , @xmath44 ) we denote the _ material derivative _ as @xmath45 at @xmath46 and @xmath44 , we denote by @xmath47 the outer normal , by @xmath48 the tangential gradient of @xmath11 , by @xmath49 the laplace beltrami operator applied to @xmath11 , and by @xmath50 the tangential divergence of @xmath9 ; see , e.g. , @xcite for these notions . as outlined in the introduction , we consider a parabolic equation on an evolving surface that moves according to an elliptically regularized velocity law : @xmath51 here , @xmath52 and @xmath53 are given continuously differentiable functions , and @xmath19 is a fixed parameter . this system is considered together with the collection of ordinary differential equations for every label @xmath25 . initial values are specified for @xmath11 and @xmath38 . on applying the leibniz formula as in @xcite , the weak formulation reads as follows : we seek to determine the weak solution @xmath54 and @xmath55 such that for all test functions @xmath56 with @xmath57 and @xmath58 , @xmath59 \int_{{\gamma}(x ) } v \cdot \psi + \alpha \int_{{\gamma}(x ) } \!\ ! \nabla_{\gamma(x ) } v \cdot \nabla_{\gamma(x ) } \psi = & \ \int_{{\gamma}(x ) } g(u,\nabla_{\gamma(x ) } u ) \,\nu_{{\gamma}(x ) } \cdot \psi , \end{aligned } \ ] ] alongside with the ordinary differential equations for the positions @xmath38 determining the surface @xmath60 . we describe the surface finite element discretization of our problem , following @xcite and @xcite . we use simplicial elements and continuous piecewise polynomial basis functions of degree @xmath61 , as defined in ( * ? ? ? * section 2.5 ) . we triangulate the given smooth surface @xmath62 by an admissible family of triangulations @xmath63 of decreasing maximal element diameter @xmath64 ; see @xcite for the notion of an admissible triangulation , which includes quasi - uniformity and shape regularity . for a momentarily fixed @xmath64 , we denote by @xmath65 the vector in @xmath29 that collects all @xmath66 nodes of the triangulation . by piecewise polynomial interpolation of degree @xmath61 , the nodal vector defines an approximate surface @xmath67 that interpolates @xmath62 in the nodes @xmath68 . we will evolve the @xmath69th node in time , denoted @xmath70 with @xmath71 , and collect the nodes at time @xmath36 in a vector @xmath72 provided that @xmath70 is sufficiently close to the exact position @xmath73 ( with @xmath74 ) on the exact surface @xmath75 , the nodal vector @xmath76 still corresponds to an admissible triangulation . in the following discussion we omit the omnipresent argument @xmath36 and just write @xmath77 for @xmath76 when the dependence on @xmath36 is not important . by piecewise polynomial interpolation on the plane reference triangle that corresponds to every curved triangle of the triangulation , the nodal vector @xmath77 defines a closed surface denoted by @xmath78 . we can then define finite element _ basis functions _ @xmath79:\gamma_h({{\mathbf x}})\to{\mathbb{r } } , \qquad j=1,\dotsc , n,\ ] ] which have the property that on every triangle their pullback to the reference triangle is polynomial of degree @xmath61 , and which satisfy @xmath79(x_k ) = \delta_{jk } \quad \text { for all } j , k = 1 , \dotsc , n .\ ] ] these functions span the finite element space on @xmath78 , @xmath80 , \phi_2[{{\mathbf x } } ] , \dotsc , \phi_n[{{\mathbf x } } ] \big\ } .\ ] ] for a finite element function @xmath81 the tangential gradient @xmath82 is defined piecewise . we set @xmath83(p_h ) , \qquad p_h \in \gamma_h^0,\ ] ] which has the properties that @xmath84 for @xmath85 , that @xmath86 for all @xmath87 , and @xmath88 the _ discrete velocity _ @xmath89 at a point @xmath90 is given by @xmath91 a key property of the basis functions is the _ transport property _ @xcite : @xmath92(x_h(p_h , t ) ) \bigr ) = 0,\ ] ] which by integration from @xmath43 to @xmath36 yields @xmath93(x_h(p_h , t ) ) = \phi_j[{{\mathbf x}}(0)](p_h).\ ] ] this implies that the discrete velocity is simply @xmath94(x ) \quad\hbox { for } x \in \gamma_h\bigl({{\mathbf x}}(t)\bigr ) , \qquad \hbox{with } \ v_j(t)=\dot x_j(t),\ ] ] where the dot denotes the time derivative @xmath95 . the _ discrete material derivative _ of a finite element function @xmath96(x ) , \qquad x \in \gamma_h\bigl({{\mathbf x}}(t)\bigr),\ ] ] is defined as @xmath97 by the transport property of the basis functions , this is just @xmath98(x ) , \qquad x \in \gamma_h\bigl({{\mathbf x}}(t)\bigr).\ ] ] the finite element spatial semidiscretization of the problem reads as follows : find the unknown nodal vector @xmath99 and the unknown finite element functions @xmath100 and @xmath101 such that , for all @xmath102 with @xmath103 and all @xmath104 , @xmath105 and @xmath106 the initial values for the nodal vector @xmath107 corresponding to @xmath108 and the nodal vector @xmath77 of the initial positions are taken as the exact initial values at the nodes @xmath68 of the triangulation of the given initial surface @xmath62 : @xmath109 we now show that the nodal vectors @xmath110 and @xmath111 of the finite element functions @xmath108 and @xmath112 , respectively , together with the surface nodal vector @xmath113 satisfy a system of differential - algebraic equations ( daes ) . using the above finite element setting , we set ( omitting the argument @xmath36 ) @xmath114 , \qquad u_h(x_j)=u_j \in { \mathbb{r } } , \\ v_h = & \ \sum_{j=1}^n v_j \phi_j[{{\mathbf x } } ] , \qquad v_h(x_j)=v_j \in { \mathbb{r}}^3 , \end{aligned}\ ] ] and collect the nodal values in column vectors @xmath115 and @xmath116 . we define the surface - dependent mass matrix @xmath117 and stiffness matrix @xmath118 on the surface determined by the nodal vector @xmath77 : @xmath119 \phi_k[{{\mathbf x } } ] , \\ { { \mathbf a}}({{\mathbf x}})\vert_{jk } = & \ \int_{{\gamma}_h({{\mathbf x } } ) } \ ! { \nabla}_{{\gamma}_h } \phi_j[{{\mathbf x } } ] \cdot { \nabla}_{{\gamma}_h } \phi_k[{{\mathbf x } } ] , \end{aligned } \qquad ( j , k = 1 , \dotsc , n ) .\ ] ] we further let @xmath120 the right - hand side vectors @xmath121 and @xmath122 are given by @xmath123 , \\ { { \mathbf g}}({{\mathbf x}},{{\mathbf u}})\vert_{3(j-1)+\ell } & = \int_{{\gamma}_h({{\mathbf x } } ) } g(u_h,{\nabla_{\gamma_h}}u_h ) \,\bigl(\nu_{{\gamma}_h({{\mathbf x}})}\bigr)_\ell \ , \phi_j[{{\mathbf x } } ] , \end{aligned } \ ] ] for @xmath124 and @xmath125 . we then obtain from the following coupled dae system for the nodal values @xmath126 and @xmath77 : @xmath127 with the auxiliary vector @xmath128 , this system becomes @xmath129 this is of a form to which standard dae time discretization can be applied ; see , e.g. , ( * ? ? ? * chap.vi ) . in the error analysis we need to compare functions on three different surfaces : the _ exact surface _ @xmath75 , the _ discrete surface _ @xmath130 , and the _ interpolated surface _ @xmath131 , where @xmath132 is the nodal vector collecting the grid points @xmath133 on the exact surface . in the following definitions we omit the argument @xmath36 in the notation . a finite element function @xmath134 ( @xmath135 or 3 ) on the discrete surface , with nodal values @xmath136 , is related to the finite element function @xmath137 on the interpolated surface that has the same nodal values : @xmath138.\ ] ] the transition between the interpolated surface and the exact surface is done by the _ lift operator _ , which was introduced in @xcite and further investigated in @xcite . the lift operator @xmath139 maps a function on the interpolated surface @xmath140 to a function on the exact surface @xmath141 , provided that @xmath140 is sufficiently close to @xmath141 . using the oriented distance function @xmath142 ( ( * ? ? ? * section 2.1 ) ) , the lift of a continuous function @xmath143 is defined as @xmath144 where for every @xmath145 the point @xmath146 is uniquely defined via @xmath147 for functions taking values in @xmath12 the lift is componentwise . by @xmath148 we denote the function on @xmath149 whose lift is @xmath150 . we denote the composed lift @xmath151 from finite element functions on @xmath152 to functions on @xmath141 via @xmath140 by @xmath153 we are now in the position to formulate the main result of this paper , which yields optimal - order error bounds for the finite element semidiscretization of a surface pde on a solution - driven surface as specified in , for finite elements of polynomial degree @xmath154 . we denote by @xmath75 the exact surface and by @xmath155 the discrete surface at time @xmath36 . we introduce the notation @xmath156 [ theorem : coupled error estimate ] consider the space discretization of the coupled problem , using evolving surface finite elements of polynomial degree @xmath154 . we assume quasi - uniform admissible triangulations of the initial surface and initial values chosen by finite element interpolation of the initial data for @xmath11 . suppose that the problem admits an exact solution @xmath157 that is sufficiently smooth ( say , in the sobolev class @xmath158 ) on the time interval @xmath44 . then , there exists @xmath159 such that for all mesh widths @xmath160 the following error bounds hold over the exact surface @xmath161 for @xmath44 : @xmath162 and @xmath163 the constant @xmath164 is independent of @xmath36 and @xmath64 , but depends on bounds of the @xmath158 norms of the solution @xmath165 , on local lipschitz constants of @xmath166 and @xmath17 , on the regularization parameter @xmath19 and on the length @xmath167 of the time interval . we note that the last error bound is equivalent to @xmath168 a key issue in the proof is to ensure that the @xmath169 norm of the position error of the curves remains small . the @xmath170 error bound and an inverse estimate yield an @xmath171 error bound in the @xmath169 norm . this is small only for @xmath154 , which is why we impose the condition @xmath154 in the above result . our assumptions guarantee that the triangulations formed by the nodes @xmath133 remain admissible uniformly for @xmath172 $ ] ( though the admissibility bounds may degrade with growing @xmath36 ) . since @xmath154 , the position error estimate implies that for sufficiently small @xmath64 also the triangulations formed by the numerical nodes @xmath70 remain admissible uniformly for @xmath172 $ ] . this can not be concluded for @xmath173 . the error bound will be proven by clearly separating the issues of consistency and stability . the consistency error is the defect on inserting a projection ( interpolation or ritz projection ) of the exact solution into the discretized equation . the defect bounds involve _ geometric estimates _ that were obtained for higher order @xmath154 in @xcite , by combining techniques of dziuk & elliott @xcite and demlow @xcite . this is done with the esfem formulation of section [ subsection : semidiscretization ] . the main issue in the proof of theorem [ theorem : coupled error estimate ] is to prove _ stability _ in the form of an @xmath64-independent bound of the error in terms of the defect . the stability analysis is done in the matrix - vector formulation of section [ subsection : dae ] . it uses energy estimates and transport formulae that relate the mass and stiffness matrices and the coupling terms for different nodal vectors @xmath77 . _ no geometric estimates _ enter in the proof of stability . in section [ section : preliminaries ] we prove important auxiliary results for the stability analysis . the stability is first analysed for the discretized velocity law without coupling to the surface pde in section [ section : uncoupled problem ] and is then extended to the coupled problem in section [ section : coupled problem ] . the necessary geometric estimates for the consistency analysis are collected in section [ section : geometric estimates ] , and the defects are then bounded in section [ section : defect bounds ] . the proof of theorem [ theorem : coupled error estimate ] is then completed in section [ section : proof completed ] by putting together the results on stability , defect bounds and interpolation error bounds . the finite element matrices of section [ subsection : dae ] induce discrete versions of sobolev norms . for any @xmath174 with corresponding finite element function @xmath175 \in s_h({{\mathbf x}})$ ] we note @xmath176 in our stability analysis we need to relate finite element matrices corresponding to different nodal vectors . we use the following setting . let @xmath177 be two nodal vectors defining discrete surfaces @xmath78 and @xmath178 , respectively . we let @xmath179 . for the parameter @xmath180 $ ] , we consider the intermediate surface @xmath181 and the corresponding finite element functions given as @xmath182\ ] ] and , for any vectors @xmath183 , @xmath184 { \quad\hbox { and } \quad}z_h^\theta=\sum_{j=1}^n z_j \phi_j[{{\mathbf y}}+\theta{{\mathbf e } } ] .\ ] ] [ lemma : matrix differences ] in the above setting the following identities hold : @xmath185 with @xmath186 for @xmath187 . using the fundamental theorem of calculus and the leibniz formula we write @xmath188 in the last formula we used that the material derivatives ( with respect to @xmath189 ) of @xmath190 and @xmath191 vanish , thanks to the transport property of the basis functions . the second identity is shown in the same way , using the formula for the derivative of the dirichlet integral ; see @xcite and also ( * ? ? ? * lemma 3.1 ) . @xmath192 a direct consequence of lemma [ lemma : matrix differences ] is the following conditional equivalence of norms : [ lemma : cond - equiv ] if @xmath193 for @xmath194 , then @xmath195 if @xmath196 for @xmath194 , then @xmath197 by lemma [ lemma : matrix differences ] we have for @xmath198 @xmath199 and the first result follows from gronwall s inequality . the second result is proved in the same way . @xmath192 to check the conditions of the previous lemma , the following result is very useful , since it reduces the condition for @xmath194 to just the case @xmath200 when it is applied with @xmath190 chosen as the components of @xmath201 . [ lemma : theta - independence ] in the above setting , assume that @xmath202 } e_h^0 \|_{l^\infty(\gamma_h[{{\mathbf y } } ] ) } \le \frac12.\]]then , for @xmath194 the function @xmath203 $ ] on @xmath204 $ ] is bounded by @xmath205 where @xmath206 depends only on @xmath25 ( we have @xmath207 ) . we describe the finite element parametrization of the discrete surfaces @xmath208 in the same way as in section [ section : esfem ] , with @xmath189 instead of @xmath36 in the role of the time variable . we set @xmath209(q_h),\quad\ q_h\in\gamma_h[{{\mathbf y}}]\ ] ] so that @xmath210=\gamma_h^\theta.\ ] ] since @xmath211 for all @xmath212 $ ] , the above formula can be rewritten as @xmath213 tangent vectors to @xmath208 at @xmath214 are therefore of the form @xmath215 where @xmath216 is a tangent vector to @xmath67 at @xmath217 , or written more concisely , @xmath218 . letting @xmath219 denote the euclidean norm of a vector in @xmath12 , we have at @xmath214 @xmath220 by construction of @xmath190 and the transport property of the basis functions , we have @xmath221(y_h^\theta(q_h ) ) = \sum_{j=1}^n w_j \phi_j[{{\mathbf y}}](q_h ) = w_h^0(q_h).\]]by the chain rule , this yields @xmath222 under the imposed condition @xmath223 ) } \le \frac12 $ ] we have for @xmath194 @xmath224 hence we obtain @xmath225 this yields the stated result for @xmath226 . for @xmath227 we note in addition that in using the integral transformation formula we have a uniform bound between the surface elements , since @xmath228 is close to the identity matrix by our smallness assumption on @xmath229 . @xmath192 the arguments of the previous proof are also used in estimating the changes of the normal vectors on the various surfaces @xmath230 $ ] . [ lemma : normal vector perturbation ] suppose that condition is satisfied . let @xmath231 be related by the parametrization of @xmath208 over @xmath67 , for @xmath194 . then , the corresponding unit normal vectors differ by no more than @xmath232 with some constant @xmath164 . let @xmath233 and @xmath234 be two linearly independent tangent vectors of @xmath67 at @xmath235 ( which may be chosen orthogonal to each other and of unit euclidean norm ) . with @xmath236 for @xmath237 we then have , for @xmath194 , @xmath238 since this expression is a locally lipschitz continuous function of the two vectors , the result follows . ( the imposed bound is sufficient to ensure the linear independence of the vectors @xmath239 . ) @xmath192 we denote by @xmath240 the material derivative of a function @xmath241 depending on @xmath180 $ ] and @xmath242 : @xmath243 from lemma [ lemma : normal vector perturbation ] together with lemma [ lemma : theta - independence ] we obtain the following bound : [ lemma : mat normal vector ] if condition is satisfied , then @xmath244 where @xmath164 is independent of @xmath194 and @xmath245 . by lemma [ lemma : normal vector perturbation ] with @xmath208 in the role of @xmath67 , we obtain @xmath246 which implies @xmath247 and lemma [ lemma : theta - independence ] completes the proof . we finally need a result that bounds the time derivatives of the mass and stiffness matrices corresponding to nodes on the exact smooth surface @xmath0 . the following result is a direct consequence of ( * ? ? ? * lemma 4.1 ) . [ lemma : matrix derivatives ] let @xmath75 , @xmath248 $ ] , be a smoothly evolving family of smooth closed surfaces , and let the vector @xmath249 collect the nodes @xmath133 . then , @xmath250 for all @xmath183 and @xmath251 $ ] . the constant @xmath164 depends only on a bound of the @xmath252 norm of the surface velocity . in this section we consider the velocity law without coupling to a surface pde : @xmath253 where @xmath254 is a given continuous function of @xmath255 , and @xmath19 is a fixed parameter . this problem is considered together with the ordinary differential equations for the positions @xmath38 determining the surface @xmath60 . initial values are specified for @xmath38 . we consider the weak formulation : find @xmath256 such that the following weak equation holds for every test function @xmath257 : @xmath258 alongside with the ordinary differential equations for the positions @xmath38 . then the finite element spatial semidiscretization of this problem reads as : find the unknown nodal vector @xmath99 and the unknown finite element function @xmath101 such that the following semidiscrete equation holds for every @xmath259 : @xmath260 together with the ordinary differential equations . as before , the nodal vector of the initial positions @xmath261 is taken from the exact initial values at the nodes @xmath68 of the triangulation of the given initial surface @xmath62 : @xmath262 for @xmath85 . as in section [ subsection : dae ] , the nodal vectors @xmath111 of the finite element function @xmath112 together with the surface nodal vector @xmath113 satisfy a system of differential - algebraic equations ( daes ) . we obtain from and the following coupled dae system for the nodal values @xmath263 and @xmath77 : @xmath264here the matrix @xmath265 is from , and the driving term @xmath266 is given by @xmath267 \qquad ( j = 1 , \dotsc , n , \ \ell=1,2,3 ) .\ ] ] we denote by @xmath268 the nodal vector of the _ exact _ positions on the surface @xmath269 . this defines a discrete surface @xmath270 that interpolates the exact surface @xmath269 . we consider the projected exact velocity @xmath271\ ] ] with the corresponding nodal vector @xmath272 we assume that @xmath273 inserting @xmath274 and @xmath275 in place of the numerical solution @xmath112 and @xmath77 into yields a defect @xmath276 : for every @xmath277 , @xmath278 with @xmath279 $ ] and the corresponding nodal vector @xmath280 we then have @xmath281 as the defect on inserting @xmath275 and @xmath282 in the first equation of , and the defect in the second equation is denoted @xmath283 : with @xmath284}({{{\mathbf x}}^\ast})= i_3\otimes{{\mathbf m}}({{{\mathbf x}}^\ast})$ ] , @xmath285}({{{\mathbf x}}^\ast}){{{\mathbf d}}_{{\mathbf v } } } , \\ { \dot{{\mathbf x}}^\ast}=&\ { { { \mathbf v}}^\ast}+ { { { \mathbf d}}_{{\mathbf x}}}. \end{aligned}\ ] ] we denote the errors in the surface nodes and in the velocity by @xmath286 and @xmath287 , respectively . we rewrite the velocity law in as @xmath288 then , by subtracting from the above version of , we obtain the following error equations for the uncoupled problem : @xmath289}({{{\mathbf x}}^\ast}){{{\mathbf d}}_{{\mathbf v } } } , \\ { \dot{{\mathbf e}}_{{\mathbf x}}}=&\ { { { \mathbf e}}_{{\mathbf v}}}- { { { \mathbf d}}_{{\mathbf x}}}. \end{aligned}\ ] ] when no confusion can arise , we write in the following @xmath290 for @xmath284}({{{\mathbf x}}^\ast})$ ] and @xmath291 for @xmath292 , etc . we recall that @xmath265 and , for @xmath293 and the corresponding finite element function @xmath175 \in s_h({{\mathbf x}})^3 $ ] , we consider the norm @xmath294 for convenience , we will take @xmath295 in the remainder of this section , so that the last norm equivalence becomes an equality . for the defect @xmath296 we use the dual norm ( cf . * proof of theorem 5.1 ) ) @xmath297 we denote @xmath298 so that @xmath299 the following stability result holds for the errors @xmath300 and @xmath301 , under an assumption of small defects . it will be shown in section [ section : defect bounds ] that this assumption is satisfied if the exact solution is sufficiently smooth and the projection @xmath302 is chosen as piecewise polynomial interpolation of degree @xmath154 . [ proposition : stability - uncoupled problem ] suppose that the defects are bounded as follows , with @xmath303 : @xmath304 .\ ] ] then there exists @xmath305 such that the following error bounds hold for @xmath306 and @xmath44 : @xmath307 the constant @xmath164 is independent of @xmath36 and @xmath64 , but depends on the final time @xmath167 and on the regularization parameter @xmath308 . we note that the error functions @xmath309 with nodal vectors @xmath310 and @xmath311 , respectively , are then bounded by @xmath312 .\ ] ] the proof relies on the results of section [ section : preliminaries ] , otherwise it mainly follows the usual steps of stability proofs based on energy estimates ; cf . in the course of this proof @xmath313 and @xmath164 will be generic constants that take on different values on different occurrences . in view of condition for @xmath314 , we will need to control the @xmath169 norm of the position error @xmath315 . let @xmath316 be the maximal time such that @xmath317.\ ] ] at @xmath318 either this inequality becomes an equality , or else we have @xmath319 . we will first prove the stated error bounds for @xmath320 . then the proof will be finished by showing that in fact @xmath321 coincides with @xmath167 . by testing the first equation in with @xmath300 , and dropping the omnipresent argument @xmath322 $ ] , we obtain : @xmath323 we separately estimate the four terms on the right - hand side in an appropriate way , with lemmas [ lemma : matrix differences ] [ lemma : normal vector perturbation ] as our main tools . \(i ) we denote , for @xmath324 , by @xmath325 and @xmath326 the finite element functions in @xmath327 for @xmath328 with nodal vectors @xmath300 and @xmath282 , respectively . lemma [ lemma : matrix differences ] then gives us @xmath329 using the cauchy schwarz inequality , we estimate the integral with the product of the @xmath330 norms of the three factors . we thus have @xmath331 by and lemma [ lemma : theta - independence ] , this is bounded by @xmath332 where the last factor is bounded independently of @xmath64 . by the young inequality , we thus obtain @xmath333 \(ii ) similarly , estimating the three factors in the integrals by @xmath334 , we obtain @xmath335 where in the last inequality we used the bound . \(iii ) in the following estimate we use lemma [ lemma : mat normal vector ] . with the finite element function @xmath336 $ ] on the surface @xmath328 , for @xmath194 , we write @xmath337 using the leibniz formula , this becomes @xmath338 here we have , noting that @xmath339 , @xmath340 with lemmas [ lemma : theta - independence ] and [ lemma : mat normal vector ] we therefore obtain via the cauchy - schwarz inequality @xmath341 and again with lemma [ lemma : theta - independence ] , @xmath342 in total , we obtain a bound of the same type as for the terms in ( i ) and ( ii ) : @xmath343 the combination of the estimates of the three terms ( i)(iii ) with absorptions ( for sufficiently small @xmath306 ) , and a simple dual norm estimate , based on , for the defect term , yield the bound @xmath344 using this estimate , together with taking the @xmath345 norm of both sides of the second equation in , we obtain @xmath346 in order to apply gronwall s inequality , we connect @xmath347 and @xmath348 as follows : @xmath349 where we use the cauchy - schwarz inequality and lemma [ lemma : matrix derivatives ] in the estimate . inserting , we obtain @xmath350 a gronwall inequality then yields , using @xmath351 for @xmath352 . inserting this estimate in , we can bound @xmath310 for @xmath353 by . now it only remains to show that @xmath319 for @xmath64 sufficiently small . for @xmath320 we use an inverse inequality and to bound the left - hand side in : @xmath354 for sufficiently small @xmath64 . hence , we can extend the bound beyond @xmath321 , which contradicts the maximality of @xmath321 unless we have already @xmath319 . now we turn to the stability bounds of the original problem , or in dae form , which is the formulation we will actually use for the stability analysis . similarly as before , in order to formulate the stability estimates we consider the dae system when we insert the nodal values @xmath355 of the exact solution @xmath11 , the nodal values @xmath356 of the exact positions @xmath38 , and the nodal values @xmath357 of the exact velocity @xmath9 . plugging them into yields the defects @xmath358 , @xmath359 and @xmath360 : @xmath361 we denote the pde error by @xmath362 , and just as before @xmath287 and @xmath286 denote the velocity error and surface error , respectively . subtracting from , we obtain the following error equation : @xmath363 we now formulate the stability result for the errors @xmath364 , @xmath300 and @xmath301 of the surface motion coupled to the surface pde . [ proposition : stability - coupled problem ] assume that the following bounds hold for the defects , for some @xmath303 : @xmath365 .\ ] ] then there exists @xmath305 such that the following stability estimate holds for all @xmath306 and @xmath44 : @xmath366 the constant @xmath164 is independent of @xmath36 and @xmath64 , but depends on the final time @xmath167 and on the regularization parameter @xmath308 . the proof is an extension of the proof of proposition [ proposition : stability - uncoupled problem ] , again based on the matrix - vector formulation and the auxiliary results of section [ section : preliminaries ] . we handle the surface pde and the surface equations separately : we first estimate the errors occurred in the pde , while those for the surface equation are based on section [ section : uncoupled problem ] . finally we will combine the results to obtain the stability estimates for the coupled problem . let @xmath367 be the maximal time such that the following inequalities hold : @xmath368,\ ] ] note that @xmath369 since initially both @xmath370 and @xmath371 . we first prove the stated error bounds for @xmath320 . at the end , the proof will be finished by showing that in fact @xmath321 coincides with @xmath167 . testing the first two equations of with @xmath372 and @xmath300 , and dropping the omnipresent argument @xmath322 $ ] , we obtain : @xmath373 _ estimates for the surface pde : _ we estimate the terms separately , again by using lemmas [ lemma : matrix differences ] and [ lemma : cond - equiv ] as our main tools . \(i ) the symmetry of @xmath290 and a simple calculation yields @xmath374 where the last term is bounded by lemma [ lemma : matrix derivatives ] as @xmath375 \(ii ) by the definition of the @xmath376-norm we have @xmath377 \(iii ) with the product rule we write @xmath378 with the notation as in lemma [ lemma : matrix differences ] with @xmath379 $ ] , that lemma yields for the first term @xmath380 using the cauchy - schwarz inequality we obtain for the first term @xmath381 under condition we obtain from lemmas [ lemma : cond - equiv ] and [ lemma : theta - independence ] that for @xmath353 , @xmath382 using lemma [ lemma : matrix differences ] together with the leibniz formula , the second term becomes @xmath383 here we further use , cf . * lemma 2.6 ) , that @xmath384 where @xmath385 symbolizes the euclidean inner product of the vectorization of two matrices . proceeding as for the first term we then obtain the estimate @xmath386 \(iv ) we obtain similarly @xmath387 where in the last inequality we used the third bound from . \(v ) lemma [ lemma : matrix differences ] and the cauchy - schwarz inequality yield @xmath388 \(vi ) similarly we estimate @xmath389 where in the last inequality we used the second bound from . \(vii ) the coupling term is estimated similarly to ( iii ) in the proof of proposition [ proposition : stability - uncoupled problem ] : @xmath390 with @xmath391\ ] ] we therefore have @xmath392 and with the leibniz formula we rewrite this as @xmath393 here we note the chain rule @xmath394 and the relation , cf . * lemma 2.6 ) , @xmath395 with a bounded bilinear map @xmath396 whose precise form is irrelevant here , and further @xmath397 in the same way as before , and using the assumed boundedness of @xmath166 , @xmath398 and @xmath399 , this yields the bound @xmath400 the above estimates combined , yield the following inequality : @xmath401 estimating further , using young s inequality and absorptions into @xmath402 ( using @xmath306 ) , we obtain the following estimate : @xmath403 _ estimates in the surface equation : _ based on section [ section : uncoupled problem ] , we obtain @xmath404 where the coupling term can be estimated based on ( iii ) in the proof of proposition [ proposition : stability - uncoupled problem ] and ( vii ) above . by taking the @xmath405-norm of the third equation in and using the above estimate , we obtain @xmath406 which , similarly as before , provides the estimate @xmath407 _ combination : _ by using all three inequalities , we first insert into , then we sum up and to obtain @xmath408 now we integrate both sides over @xmath409 $ ] , for @xmath320 , to get @xmath410 the middle term can be further bounded using lemmas [ lemma : matrix differences][lemma : theta - independence ] and an @xmath334 estimate , as @xmath411 where in the last inequality we used the first bound from . absorbing this to the left - hand side and using gronwall s inequality yields the stability estimates : @xmath412 further estimating the defects using the assumed bounds , we obtain the @xmath413 estimates for @xmath414 . finally , the fact that @xmath319 is shown by the same argument as in the proof of proposition [ proposition : stability - uncoupled problem ] . @xmath192 in this section we give further notations and some technical lemmas from @xcite that will be used later on . most of the results are high - order and time - dependent extensions of geometric approximation estimates shown in @xcite and @xcite . we return to the setting of section [ section : problem ] , where @xmath415 defines a smooth surface @xmath75 . for an admissible triangulation of @xmath0 with nodes @xmath133 and the corresponding nodal vector @xmath416 , we define the interpolating surface by @xmath417(p_h ) , \qquad p_h \in \gamma_h^0,\ ] ] which has the properties that @xmath418 for @xmath85 , and @xmath419 in the following we drop the argument @xmath36 when it is not essential . the velocity of the interpolating surface @xmath149 , defined as in section [ section : esfem ] , is denoted by @xmath274 . the lift of a function @xmath420 is again denoted by @xmath421 , defined via the oriented distance function @xmath142 between @xmath422 and @xmath0 provided that the surfaces are sufficiently close ( which is the case if @xmath64 is sufficiently small ) . * ( equivalence of norms ( * ? ? ? * lemma 3 ) , @xcite ) * [ lemma : equivalence of norms ] let @xmath423 with lift @xmath424 . then the discrete and continuous @xmath425 and @xmath426 norms are equivalent for @xmath245 , uniformly in the mesh size @xmath427 ( with sufficiently small @xmath305 ) and in @xmath428 $ ] . in particular , there is a constant @xmath313 such that for @xmath427 and @xmath44 , @xmath429 later on the following estimates will be used . they have been shown in @xcite , based on @xcite and @xcite . [ lemma : geometric est ] we have the estimates @xmath430 with constants independent of @xmath427 and @xmath428 $ ] . we use surface - dependent bilinear forms defined similarly as in @xcite : let @xmath38 be a given surface with velocity @xmath9 , with interpolation surface @xmath431 with velocity @xmath274 . for arbitrary @xmath432 and for their discrete analogs @xmath433 : @xmath434 m(x;z,{\varphi } ) & = \int_{{\gamma}(x)}\!\!\!\ ! z { \varphi } , \\ a(x;z,{\varphi } ) & = \int_{{\gamma}(x)}\!\!\!\ ! { \nabla_{\gamma}}z \cdot { \nabla_{\gamma}}{\varphi } , \\ q(x;v;z,{\varphi } ) & = \int_{{\gamma}(x)}\!\!\!\ ! ( { \nabla_{\gamma}}\cdot v ) z{\varphi } , \\ \end{aligned } \quad \begin{aligned}[c ] m({x_h^\ast};z_h,\phi_h ) & = \int_{{\gamma}({x_h^\ast})}\!\!\!\ ! z_h \phi_h , \\ a({x_h^\ast};z_h,\phi_h ) & = \int_{{\gamma}({x_h^\ast})}\!\!\!\ ! { \nabla_{\gamma_h}}z_h \cdot { \nabla_{\gamma_h}}\phi_h , \\ q({x_h^\ast};v_h^*;z_h,\phi_h ) & = \int_{{\gamma}({x_h^\ast})}\!\!\!\ ! ( { \nabla_{\gamma_h}}\cdot v_h^ * ) z_h \phi_h , \\ \end{aligned}\ ] ] where the discrete tangential gradients are understood in a piecewise sense . for more details see ( * ? ? ? * lemma 2.1 ) ( and the references in the proof ) , or ( * ? ? ? * lemma 5.2 ) . we start by defining a discrete velocity on the smooth surface , denoted by @xmath435 . we follow sections 4.3 and 5.3 of @xcite . using the lifted elements , @xmath436 is decomposed into curved elements whose edges move with the velocity @xmath435 defined by @xmath437 discrete material derivatives on @xmath438 and @xmath436 are given by @xmath439 in ( * ? ? ? * lemma 4.1 ) it was shown that the transport property of the basis functions carries over to the lifted basis functions @xmath440 $ ] : @xmath441^l = ( { \partial^{\bullet}}_{{v_h^\ast } } \phi_j[{{{\mathbf x}}^\ast}])^l = 0 \qquad ( j=1,\dotsc , n ) .\ ] ] therefore , the above discrete material derivatives and the lift operator satisfy , for @xmath442 , @xmath443 * ( transport properties ( * ? ? ? * lemma 4.2 ) ) * [ lemma : transport prop ] for any @xmath444 , @xmath445 the same formulas hold when @xmath6 and @xmath9 are replaced by @xmath446 and @xmath435 , respectively : @xmath447 similarly , in the discrete case , for arbitrary @xmath448 we have : @xmath449 where @xmath450 is the velocity of the surface @xmath438 . the following estimates , proved in lemma 5.6 of @xcite , will play a crucial role in the defect bounds later on . [ lemma : geometric perturbation errors ] for any @xmath451 where @xmath438 is the interpolation surface of piecewise polynomial degree @xmath61 , we have the following bounds @xmath452 the constant @xmath313 is independent of @xmath64 and @xmath172 $ ] . for any @xmath453 , there is a unique piecewise polynomial surface finite element interpolation of degree @xmath61 in the nodes @xmath454 , denoted by @xmath455 . we set @xmath456 . error estimates for this interpolation are obtained from ( * ? ? ? * proposition 2.7 ) by carefully studying the time dependence of the constants , cf . @xcite . [ lemma : interpolation error ] there exists a constant @xmath457 independent of @xmath64 and @xmath36 such that for @xmath458 , for @xmath459 , @xmath460 the same result holds for vector valued functions . as it will always be clear from the context we do not distinguish between interpolations for scalar and vector valued functions . in this section we show that the assumed defect estimates of proposition [ proposition : stability - uncoupled problem ] and [ proposition : stability - coupled problem ] are indeed fulfilled when the projection @xmath461 is chosen to be the piecewise @xmath61th - degree polynomial interpolation operator @xmath462 for @xmath463 . the interpolations satisfy the discrete problem only up to some defects . these defects are denoted by @xmath464 , with @xmath465 the vector of exact nodal values @xmath466 , and are given as follows : for all @xmath467 with @xmath468 and @xmath469 , @xmath470 later on the vectors of nodal values of the defects @xmath471 and @xmath472 are denoted by @xmath473 and @xmath474 , respectively . these vectors , together with @xmath475 , satisfy . [ lemma : semidiscrete residual ] let the solution @xmath11 , the surface @xmath38 and its velocity @xmath9 be all sufficiently smooth . then there exists a constant @xmath457 such that for all @xmath172 $ ] , the defects @xmath471 and @xmath472 of the @xmath61th - degree finite element interpolation are bounded as @xmath476 where the @xmath477-norm is defined in . the constant @xmath313 is independent of @xmath64 and @xmath172 $ ] . \(i ) we start from an identity for the dual norm as in , ( omitting the argument @xmath478 of the matrices ) : @xmath479 in order to estimate the defect in @xmath11 , we subtract from the above equation , and perform almost the same proof as in ( * ? ? ? * section 7 ) . we use the bilinear forms and the discrete versions of the transport properties from lemma [ lemma : transport prop ] . we obtain , for any @xmath467 with @xmath468 , @xmath480 and @xmath481 subtracting the two equation yields @xmath482 we bound all the terms pairwise , by using the interpolation estimates of lemma [ lemma : interpolation error ] and the estimates for the geometric perturbation errors of the bilinear forms of lemma [ lemma : geometric perturbation errors ] . for the first pair , using that @xmath483 , we obtain @xmath484 for the second pair we obtain @xmath485 the third pair is estimated by @xmath486 for the last pair we use the fact that @xmath487 and the local lipschitz continuity of the function @xmath166 , to obtain @xmath488 the first term is estimated , using the local lipschitz continuity of @xmath166 and equivalence of norms , by @xmath489 where the first two terms are bounded by @xmath413 using interpolation estimates , while the third term is bounded , using remark 4.1 in @xcite and lemma [ lemma : geometric est ] , as @xmath490 thus for the fourth pair we obtained @xmath491 altogether , we have @xmath492 which , by the equivalence of norms given by lemma [ lemma : equivalence of norms ] , shows the first bound of the stated lemma . \(ii ) in order to estimate the defect in @xmath9 , similarly as previously we subtract from the above equation and use the bilinear forms to obtain @xmath493 similarly as in the previous part , these three pairs are bounded pairwise . for the first pair we have @xmath494 for the second pair we use the interpolation estimate to bound @xmath495 the third pair we estimate , similarly to the nonlinear pair above , by @xmath496 where we have used the local lipschitz boundedness of the function @xmath17 , the interpolation estimate , lemma [ lemma : geometric est ] , and lemma [ lemma : geometric perturbation errors ] , through a similar argument as above for the semilinear term with @xmath166 . finally , the combination of these bounds yields @xmath497 providing the asserted bound on @xmath498 . @xmath192 the errors are decomposed using interpolations and the definition of lifts from section [ section : lifts ] : omitting the argument @xmath36 , @xmath499 the last terms in these formulas can be bounded in the @xmath500 norm by @xmath501 , using the interpolation bounds of lemma [ lemma : interpolation error ] . to bound the first terms on the right - hand sides , we first use the defect bounds of lemma [ lemma : semidiscrete residual ] , which then together with the stability estimate of proposition [ proposition : stability - coupled problem ] proves the result , since by the norm equivalences of lemma [ lemma : equivalence of norms ] and equations we have ( again omitting the argument @xmath36 ) @xmath502 and similarly for @xmath503 and @xmath504 . in this section we consider the extension of our results to different velocity laws : adding a mean curvature term to the regularized velocity law considered so far , and a dynamic velocity law . we concentrate on the velocity laws without coupling to the surface pde , since the coupling can be dealt with in the same way as previously . we only consider the stability of the evolving surface finite element discretization , since bounds for the consistency error are obtained by the same arguments as before . we next extend our results to the case where the velocity law contains a mean curvature term : @xmath505 where @xmath254 is a given lipschitz continuous function of @xmath255 , and @xmath19 and @xmath506 are fixed parameters . here @xmath507 is a suggestive notation for @xmath508 , where @xmath21 denotes the mean curvature of the surface @xmath436 . ( more precisely , @xmath509 . ) the corresponding differential - algebraic system has the following form : @xmath510 similarly as before the corresponding error equation is given as @xmath511 [ proposition : stability - with mc term ] under the assumptions of proposition [ proposition : stability - uncoupled problem ] , there exists @xmath305 such that the following stability estimate holds for all @xmath306 , for @xmath44 : @xmath512 the constant @xmath164 is independent of @xmath36 and @xmath64 , but depends on the final time @xmath167 , and on the parameters @xmath308 and @xmath513 . we detail only those parts of the proof of proposition [ proposition : stability - uncoupled problem ] where the mean curvature term introduces differences , otherwise exactly the same proof applies . in order to prove the stability estimate we again test with @xmath300 , and obtain @xmath514 every term is estimated exactly as previously in the proof of proposition [ proposition : stability - uncoupled problem ] , except the terms corresponding to the mean curvature term , involving the stiffness matrix @xmath376 . they are estimated by the same techniques as previously : @xmath515 altogether we obtain the error bound @xmath516 which is exactly . the proof is then finished as before . @xmath192 with proposition [ proposition : stability - with mc term ] and the appropriate defect bounds , theorem [ theorem : coupled error estimate ] extends directly to the system with mean curvature term in the regularized velocity law . let us consider the dynamic velocity law , again without coupling to a surface pde : @xmath517 where again @xmath254 is a given lipschitz continuous function of @xmath255 , and @xmath19 is a fixed parameter . this problem is considered together with the ordinary differential equations for the positions @xmath38 determining the surface @xmath60 . initial values are specified for @xmath38 and @xmath9 . the weak formulation and the semidiscrete problem can be obtained by a similar argument as for the pde on the surface in section [ section : coupled problem ] . therefore we immediately present the ode formulation of the semidiscretization . as in section [ subsection : dae ] , the nodal vectors @xmath111 of the finite element function @xmath112 , together with the surface nodal vector @xmath113 satisfy a system of odes with matrices and driving term as in section [ section : uncoupled problem ] : @xmath518 by using the same notations for the exact positions @xmath519 , the projected exact velocity @xmath520 , and for the defects @xmath521 and @xmath522 , we obtain that they fulfill the following equation @xmath523 by subtracting this from , and using similar arguments as before , we obtain the error equations for the surface nodes and velocity : @xmath524 essentially we repeat the stability estimates of section [ subsection : stability estimate - pde ] , but we track the @xmath376- and @xmath525-norms more carefully in the proof of proposition [ proposition : stability - coupled problem ] . [ proposition : stability - dynamic velocity laws ] under the assumptions of proposition [ proposition : stability - uncoupled problem ] , there exists @xmath305 such that the following error estimate holds for all @xmath306 , uniformly for @xmath44 : @xmath526 the constant @xmath527 is independent of @xmath36 and @xmath64 , but depends on the final time @xmath167 and the parameter @xmath308 . by testing the error equation with @xmath300 we obtain @xmath528 the terms are bounded in the same way as the corresponding terms in the proofs of propositions [ proposition : stability - uncoupled problem ] and [ proposition : stability - coupled problem ] . with these estimates , a gronwall inequality yields the result . @xmath192 with proposition [ proposition : stability - dynamic velocity laws ] and the appropriate defect bounds , theorem [ theorem : coupled error estimate ] extends directly to the parabolic surface pde coupled with the dynamic velocity law . in this section we complement theorem [ theorem : coupled error estimate ] by showing the numerical behaviour of piecewise linear finite elements , which are not covered by theorem [ theorem : coupled error estimate ] , but nevertheless perform remarkably well . moreover , we compare our regularized velocity law with regularization by mean curvature flow . our test problem is a combination of with a mean curvature term as in : @xmath529 for non - negative parameters @xmath530 . the velocity law here is a special case of for @xmath531 , and reduces to for @xmath532 . the matrix - vector form reads @xmath533 , \\ { { { \mathbf k}}}\bigl({{{\mathbf x}}}(t)\bigr ) \dot{{{\mathbf x}}}(t ) + \beta { { { \mathbf a}}}\bigl({{{\mathbf x}}}(t)\bigr ) { { { \mathbf x}}}(t)= { } & \delta \mathbf{n}\bigl({{{\mathbf x}}}(t)){{\mathbf u}}(t ) + { { \mathbf g}}\bigl(t , { { { \mathbf x}}}(t ) \bigr ) & & t\in [ 0,t],\end{aligned}\ ] ] for given @xmath261 and @xmath534 , where @xmath535,\ ] ] for @xmath536 and @xmath537 . in our numerical experiments we used a linearly implicit euler discretization of this system with step sizes chosen so small that the error is dominated by the spatial discretization error . [ example:1 ] we consider and choose @xmath166 and @xmath17 such that @xmath538 with @xmath539 and @xmath540 are the exact solution of the problem . the parameters are set to be @xmath541 , @xmath295 , @xmath542 , @xmath543 , @xmath544 , @xmath545 and @xmath546 . we choose @xmath547 as a series of meshes such that @xmath548 . in table [ table : ex1 ] we report on the errors and the corresponding experimental orders of convergence ( eoc ) . using the notation of section [ section : lifts ] , the following norms are used : @xmath549 } \lvert \widehat u_{h}({\:\cdot\ : } , t ) - { \widetilde{i}_h}u({\:\cdot\ : } , t ) \rvert_{l^{2}({\gamma}_h^*(t ) ) } , \\ \|{\rm err}_u \| _ { l^{2}(h^1)}&= \left ( \int_{0}^{t } \bigl\lvert \widehat u_{h}({\:\cdot\:},s)- { \widetilde{i}_h}u({\:\cdot\:},s ) \bigr\rvert^{2}_{h^{1}({\gamma}_h^*(s ) ) } \mathrm{d}s \right)^{\frac{1}{2 } } , \\ \|{\rm err}_v \| _ { l^{\infty}(h^1 ) } & = \sup_{[0,t]}\lvert \widehat v_{h}({\:\cdot\ : } , t ) - { \widetilde{i}_h}v({\:\cdot\ : } , t ) \rvert_{h^1({\gamma}_h^*(t ) ) } , \\ \|{\rm err}_x \| _ { l^{\infty}(h^1 ) } & = \sup_{[0,t]}\lvert \widehat x_{h}({\:\cdot\ : } , t ) - { \mathrm{id}}_{{\gamma}_h^*(t ) } \rvert_{h^1({\gamma}_h^*(t ) ) } . \end{aligned}\ ] ] the eocs for the errors @xmath550 and @xmath551 with mesh sizes @xmath552 are given via @xmath553 the degree of freedoms ( dof ) and maximum mesh size at time @xmath167 are also reported in the tables . in table [ table : ex1 ] we report on the errors and eocs observed using example [ example:1 ] . the eocs in the pde are expected to be @xmath27 for the @xmath554 norm and @xmath555 for the @xmath556 norm , while the errors in the surface and in the surface velocity are expected to be @xmath555 in the @xmath557 norm . + [ example:3 ] again we consider , but this time we quantitatively compare the two different regularized velocity laws . hence , we let @xmath558 vanish . we use a @xmath17 like in example [ example:1 ] and run two tests with the common parameters @xmath559 , @xmath544 , @xmath545 and @xmath546 , and use the same mesh and time step levels as before . the first test uses @xmath560 and @xmath561 and the second test uses @xmath295 and @xmath542 . the results are captured in table [ table : ex3 ] . + our regularized velocity law provides smaller errors as regularizing with mean curvature flow . the eocs in the errors in the surface and in the errors for the surface velocity are expected to be @xmath555 in @xmath562 and @xmath563 norm , see table [ table : ex3].b . while it can be observed that for this particular example the convergence rates for @xmath564 are higher then for @xmath565 . our next test problem is the coupled system of equations @xmath566 where @xmath567 with non - negative parameters @xmath568 . for @xmath560 this system has been used as a simplified model for tumor growth ; see barreira , elliott and madzvamuse @xcite and @xcite . these authors used the mean curvature term with a small parameter @xmath506 to regularize their velocity law . we used piecewise linear finite elements and the same time discretization scheme as in @xcite . [ example:2 ] we consider and want to compare qualitatively the two different regularized velocity laws @xmath564 and @xmath569 . as common parameters we use @xmath570 , @xmath571 , @xmath572 , @xmath573 and @xmath574 . the initial surface is a sphere and the initial values @xmath575 and @xmath576 are calculated by solving an auxiliary surface pde as follows . we take small perturbations around the steady state @xmath577 where @xmath578 $ ] take random values . we solve the auxiliary coupled diffusion equations with the stationary initial surface until time @xmath579 . we set @xmath580 and @xmath581 , which we used as initial values for . we perform two experiments with @xmath582 and @xmath583 . we present snapshots in figure [ fig : example2 ] . we observe that both velocity laws display the same qualitative behavior , also agreeing with @xcite . the research stay of buyang li at the university of tbingen during the preparation of this paper was funded by the alexander von humboldt foundation . the work of balzs kovcs is funded by deutsche forschungsgemeinschaft , sfb 1173 .

biological and cultural features of human populations have been traditionally studied by separate disciplines , but the parallelisms between biological and cultural evolution have been put forward by a number of researchers . already darwin ( 1871 ) pointed out that _ `` the formation of different languages and of distinct species and the proofs that both have been developed through a gradual process are curiously parallel . '' _ cultural traits are transmitted from ancestors to their descendence , in a process analogous to inheritance , and are subject to changes , similar to mutations , by interaction between individuals such as teaching and imitation . moreover , they usually fulfill a practical purpose , which amounts to being subject to selection . in fact , they enhance the relationships within human groups , defining social entities comparable to certain biological species and populations . the quantification of cultural traits has been attempted only recently . for example , cavalli - sforza _ et al . _ ( 1982 ) applied concepts from the quantitative theory of biological evolution to construct a theory of cultural evolution . they analyzed forty traits , ranging from political preferences to superstitions . many of these traits are subject to high mutability , since they are influenced by fashion , individual acquaintances , and personal experience , and one does not expect their quantitative properties to be directly comparable to those of any biological feature . other traits , on the other hand , are better preserved . among them we find languages and surnames . language is essential to integrate the individual to society ; surnames are historical though recent signs of identity in social groups . quite early , galton and watson ( 1874 ) dealt with the problem of the extinction of surnames . this problem is equivalent to that of the extinction of a mutant allele in a population , although this relation was only established half a century later ( fisher , 1922 ; haldane , 1927 ) , when the first quantitative approaches to biological evolutionary processes took place . comparative methods analogous to biological taxonomy are used to determine the degree of similarity between languages . this method returns the _ genetic classification _ of linguistic diversity ( see for example greenberg , 1992 ; ruhlen , 1992 ) . recently , a quantitative study of the taxonomy of languages has been carried out ( zanette , 2001 ) , showing that the distribution of the number of subtaxa within a taxa displays scaling properties , quantitatively similar to those disclosed in biological taxonomy by burlando ( 1990 , 1993 ) . that is , if @xmath0 is the number of subtaxa belonging to a given taxa say , the number of languages in the indo - european family , or the number of species in the genus _ canis_ the fraction @xmath1 of taxa that have precisely @xmath0 subtaxa scales as @xmath2 . the exponent is found in the interval @xmath3 in both cases . this gives quantitative support to darwin s observation on the `` equivalence '' between the mechanisms behind biological and linguistic evolution . a complementary comparison is that of linguistic abundance , measured as the number of individuals speaking a language , with the number of individuals of a biological species . again , the frequency as a function of the number of individuals has scaling properties both for languages ( gomes _ et al . _ , 1999 ) and for species ( pielou , 1969 ) , see fig . 1 . surnames are cultural traits ( cavalli - sforza & feldman , 1981 ) whose transmission bears strong similarity with that of some biological features . they are inherited from one of the parents , usually the father , much in the same way as the y - chromosome or the mitochondrial dna . the extinction of a surname and the persistence of a non - recombining neutral allele are equivalent problems . this is not only a mathematical fact , but has also practical implications . indeed , to assess the multiple or single origin of a surname one can turn to genetic measures , since males sharing the same surname might also share the same haplotype in the nonrecombining segment of the y chromosome ( sykes & irven , 2000 ) . in a very large population , the statistical properties of the surname distribution can be strongly correlated with genetic diversity ( barrai _ et al . _ , 1996 ) , and may even be used to understand the genetic structure of a population ( yasuda _ et al . _ , 1974 ) . recent reports on actual populations ( miyazima _ et al . _ , 2000 ; zanette & manrubia , 2001 ) show that the distribution of surnames follows the same statistical law observed for languages and biological species . namely , if @xmath0 is the number of individuals bearing a given surname , the fraction @xmath1 of surnames decreases with @xmath0 as @xmath4 ( see fig . 1 ) . here , however , the exponent is always @xmath5 . this paper focuses on a model aimed at predicting this kind of regularities , observed in disparate human populations . [ fig : f1 ] at the mathematical level , several models have been proposed and analysed in order to identify quantitative properties of surnames evolution or , equivalently , those of nonrecombining neutral alleles . the main points addressed by these studies are ( i ) the probability of fixation of a given surname / allele in a closed population ( galton & watson , 1874 ; fisher , 1922 ; haldane , 1927 ; moran , 1962 ; lange , 1981 ; rannala 1997 ; hull 1998 ) , and ( ii ) the distribution of the number of individuals bearing the same surname / allele ( kimura & crow , 1964 ; karlin & mcgregor , 1967 ; fox & lasker , 1983 ; panaretos , 1989 ; gale 1990 ; consul , 1991 ; islam , 1995 ; zanette & manrubia , 2001 ) . indeed , these two questions cover complementary aspects of the same problem . in ( i ) , one deals with a closed population ( no immigrants enter the system ) , and implicitly assumes that the mutation rate is low enough , such that fixation can indeed occur before a mutant form appears . suppose that there are @xmath6 individuals in the population . we know from coalescence theory that the time @xmath7 ( in units of the number of generations ) required for a neutral allele to be fixed is of order @xmath8 . now suppose that the mutation rate per generation and per individual is @xmath9 . then @xmath10 is the average number of mutants after @xmath7 generations have elapsed . only if @xmath11 , that is if @xmath12 , will the fixation of the allele be possible . as soon as this inequality is violated , a new situation arises , in which both neutral drift and mutation play relevant roles . in this case , a broad distribution of surnames or of genetic diversity are expected . actually , the value of @xmath9 will be usually fixed by the nature of the problem , while the size @xmath6 of the considered population can increase enough such that @xmath13 . therefore , the statistical behaviour crosses over to the second regime , where the appearance of mutants can not be discarded . we are then in the assumptions of the models of class ( ii ) . all the models in class ( ii ) represent systems where statistical equilibrium is reached . while the neutral drift drives the less frequent surnames out of the system , mutations generate new surnames . for large times , the number of different surnames in the system is much larger than unity . yasuda and co - workers ( yasuda _ et al . _ , 1974 ) used a stochastic model for a population with a fixed number of individuals . at each evolution step , a new individual with the surname of his father is added , but he acquires a new surname with a given probability . in any case , a randomly chosen individual is eliminated in order to keep the population constant . their analytical results compared successfully to field data despite the restriction of constant population , which forces the elimination of individuals . in this model , the size of the progeny is poisson distributed . other models start with a single individual in the population and new individuals carrying the same surname or a new one are sequentially added ( panaretos , 1989 ; consul , 1991 ; islam , 1995 ) . these fall in the category of branching processes ( harris , 1963 ) with an increasing number of individuals in the population . panaretos ( 1989 ) rephrased a model introduced by simon ( 1955 ) in the context of linguistics , as follows . at each step , an individual is chosen to be the father of a newborn and his surname is transmitted with probability @xmath14 . with the complementary probability @xmath15 , the newborn is assigned a surname not present in the population . recently , we have modified simon s model by ( i ) introducing an additional parameter @xmath16 which represents the death rate of the individuals in the population , and ( ii ) allowing for arbitrary initial conditions . preliminary results have been successfully compared with large data sets ( zanette & manrubia , 2001 ) . this is the starting point of the present work , where we present new analytical and numerical results for this birth - death model . we describe the model in the next section , where our analytical results are derived . questions like the role played by the composition of the founder population and by death events in an exponentially growing population are addressed . we also analyse a system with constant population , which turns out to quantitatively differ from the previous case . in section [ num ] we numerically check our analytical results and run computer simulations of the model in situations not amenable to an analytical description . in section [ empirical ] our results are tested against field data . we finish with an overall discussion in the last section . our model population evolves in discrete steps , each step corresponding to the birth of a new individual . at each time step , moreover , an individual is chosen at random from the whole population and becomes removed with probability @xmath16 , representing a death event . the total population at step @xmath17 , @xmath18 , is therefore a stochastic process governed by the evolution equation @xmath19 where @xmath20 is the dichotomic stochastic process @xmath21 we show in appendix [ ap1 ] that , averaging over realizations of the stochastic process @xmath20 , the average total population @xmath22 grows exponentially in time : @xmath23.\ ] ] here , @xmath24 is the birth rate per individual and unit time , and the product @xmath25 turns out to be the corresponding death rate ; @xmath26 is the initial population . we think of the population as divided into groups the families within which all individuals bear the same surname . at each birth event , the newborn is assigned a new surname , not previously present in the population , with probability @xmath15 . this probability can be seen as a mutation rate for surnames , but could also be interpreted as a measure of migration towards the population of individuals with new surnames . with the complementary probability , @xmath14 , a preexistent individual is chosen at random from the population and becomes the newborn s father , i.e. his surname is given to the newborn . surnames are therefore assigned with probabilities proportional to the size of the corresponding families , allowing however for fluctuations inherent to the random distribution of births among families . consequently , the distribution of surnames is driven by a stochastic multiplicative process ( van kampen , 1981 ) , modulated in turn by the total population growth . this process is analogous to the mechanism proposed by simon ( 1955 ) to model the frequency distribution of words and city sizes , described by zipf s law ( zipf , 1949 ) , among other instances . our model , in fact , reduces to simon s model if mortality is neglected , i.e. for @xmath27 . note that allowing for death events adds a relevant process in the case of surnames , namely , the possibility that a surname disappears if it is borne by a single individual and the individual dies . since neither the probability of becoming father of a newborn nor the death probability depend on the individual s age , the present model population can be thought of as ageless . as a consequence , if @xmath24 and @xmath16 are constant , the probability that an individual dies at an age between @xmath28 and @xmath29 is @xmath30 which implies a life expectancy @xmath31 . moreover , the probability that an individual has exactly @xmath32 children with the same family name during its whole life equals @xmath33^{-m-1},\ ] ] giving a fertility @xmath34 . the exponential distribution in ( [ pm ] ) is in reasonable agreement with actual data collected over relatively short periods , for instance , for the united states ( hull , 1998 ) , but contrasts with the poissonian distribution found in data integrated over several centuries , as in the case of england ( dewdney , 1986 ) . this discrepancy can presumably be ascribed to long - range variations of real birth and death rates ( see fig . we are assuming that an individual s surname is inherited from the father . consequently , the model as presented above describes the evolution of the male population only . the same evolution rules apply however if it is assumed that surnames are transmitted with the same probability by either parent . in this case , there is no sex distinction and the model encompasses the whole population . the real situation is in fact intermediate between these two limiting cases : whereas some societies strongly favor inheritance of the surname either from the father or from the mother , other groups allow for a choice between the two possibilities . in some developed western countries , where the surname is preferentially transmitted by the father , opting for the mother s surname has been proposed as a method to avoid the persistent extinction of surnames , due to the extremely slow population growth ( legay & vernay , 1999 ) . [ fig : f2 ] as stated above , our model reduces to simon s model ( simon , 1955 ) for @xmath35 . in this case , @xmath36 for all @xmath17 and the total population evolves deterministically , @xmath37 , since exactly one individual is added to the population at each step . the number of families with exactly @xmath38 individuals at step @xmath17 , @xmath39 , grows to @xmath40 when an individual is added to a family of size @xmath41 . this occurs with probability @xmath42 . on the average , thus , the number of families with@xmath38 individuals varies according to @xmath43,\ ] ] for @xmath44 . to the families with only one individual , on the other hand , the positive contribution comes from the creation of new surnames with probability @xmath15 . therefore , @xmath45 note that , since for @xmath27 the stochastic process @xmath20 becomes trivial , overlines indicate here average over realizations of the stochastic process by which each newborn s surname is chosen . the system of equations ( [ simon2 ] ) and ( [ simon3 ] ) can be completely solved for arbitrary initial conditions . in fact , ( [ simon3 ] ) is an autonomous equation for @xmath46 , whose solution reads @xmath47 where @xmath48 is the gamma - function ( abramowitz & stegun , 1970 ) . then , eqn ( [ simon2 ] ) can be used to recursively obtain @xmath49 for @xmath44 . for long times , @xmath50 , @xmath46 is essentially given by the first term in the right - hand side of eqn ( [ n1 ] ) , which grows linearly with @xmath17 , as the total population . the second term , which contains all the information about the initial condition , can be seen to decrease as @xmath51 . in this limit , the recursion from eqn ( [ simon2 ] ) can be immediately solved . thus , for @xmath50 , we find @xmath52 the asymptotic number of families of a given size @xmath38 turns out to be proportional to the total population . for fixed @xmath17 and sufficiently large values of @xmath38 , @xmath49 decreases as a power law , @xmath53 , with @xmath54 in the limit of small mutation rate , @xmath55 , we have @xmath56 . as advanced in the introduction , this is the value observed in actual surname distributions . this same result was obtained by simon ( 1955 ) for a special initial condition . elsewhere ( zanette & manrubia , 2001 ) we have already discussed the fact that , though the effects of the initial condition in simon s model fade out for long times , transients can strongly depend on the initial distribution @xmath57 ( see also section [ num ] ) . in our case , this aspect could be relevant in populations where modern surnames have appeared recently , like the japanese ( miyazima _ et al . _ , 2000 ) . we shall return to this specific point later . equations ( [ simon2 ] ) and ( [ simon3 ] ) imply that , in average , the total number of surnames in the population grows linearly : @xmath58 with @xmath59 the initial number of different surnames . as a function of time , the number of surnames increases exponentially : @xmath60.\ ] ] we have already pointed out that , in contrast , the number of surnames in some real populations at present times is decreasing ( cavalli - sforza & feldman , 1981 ; legay & vernay , 1999 ) . under suitable conditions , adding mortality allows to reproduce this particular behaviour . under the action of mortality , the growth of the total population @xmath18 fluctuates stochastically , according to eqn ( [ pest ] ) , depending on the occurrence of death events at each evolution step . the evolution of @xmath49 can be implemented in two substeps , as follows . first , eqns ( [ simon2 ] ) and ( [ simon3 ] ) are used to calculate the intermediate values @xmath61,\ ] ] and @xmath62 and the population is updated to @xmath63 . second , the evolution due to mortality is applied . in terms of the random process @xmath64 of eqn ( [ xi ] ) , we have @xmath65 } \left [ ( i+1 ) { \overline}n_{i+1}'(s)-i{\overline}n_i'(s ) \right],\ ] ] and @xmath66 [ cf . eqn ( [ pest ] ) ] . naive replacement of the stochastic process @xmath20 by its average value , @xmath67 , in order to obtain a deterministic approximation to the problem , would lead to an equation for which positive solutions can not be insured . in fact , it is possible to give initial conditions for such deterministic equation which lead to negative values of @xmath49 at sufficiently large @xmath17 and @xmath38 . a satisfactory deterministic approximation can however be proposed by assuming that the solution @xmath49 varies slowly both in @xmath17 and @xmath38 . this assumption is generically verified for large populations with smooth initial conditions . in this situation , eqns ( [ simon2i ] ) and ( [ simonm ] ) admit a continuous approximation in terms of real variables @xmath68 and @xmath69 , which replace the integer variables @xmath17 and @xmath38 , respectively . meanwhile , @xmath49 is replaced by a continuous function @xmath70 . the continuous approximation can be analyzed at different truncation orders , as discussed in appendix [ ap2 ] . at the first order , the approximate solution to eqns ( [ simon2i ] ) and ( [ simonm ] ) reads @xmath71 for @xmath72 , and @xmath73 for @xmath74 , with @xmath75 in this first - order continuous approximation , thus , the average number of families @xmath70 exhibits two separated regimes . for large values of the size variable , @xmath74 , the distribution is essentially determined by the initial condition . at the first evolution steps , where @xmath76 and @xmath77 , this regime covers practically all the domain of variable @xmath69 . as time elapses and @xmath78 grows , however , this regime recedes and is replaced for @xmath79 by a power - law distribution , @xmath80 , with @xmath81 note that , as the exponent @xmath68 in eqn ( [ z ] ) , this new exponent coincides with the observed value , @xmath82 , in the relevant limit @xmath55 . this result is independent of the death probability @xmath16 . for sufficiently long times , the power - law regime will be observed for all the relevant family sizes . all the contribution of the initial condition will become restricted to the domain of the largest families . the boundary @xmath83 between the two regimes , eqn ( [ yc ] ) , grows as a power of the ratio between the average current population @xmath84 and the initial population @xmath26 . for @xmath85 , the exponent is practically equal to unity . consequently , for the boundary to reach a given value @xmath86 , in such a way that all the families with sizes below @xmath86 are in the power - law regime , the total population must grow by a factor practically equal to @xmath86 . taking into account eqn ( [ pt ] ) , the position of the boundary as a function of real time reads @xmath87.\ ] ] the transient @xmath88 needed for the power - law regime to develop up to a given size @xmath86 is therefore logarithmic , @xmath89 . the average total number of surnames in the continuous approximation is calculated by analogy with eqn ( [ ns ] ) : @xmath90 using the above first - order solution for @xmath70 , we find @xmath91 or , as a function of time , @xmath92 - 1\ } , \ ] ] as shown in appendix [ ap2 ] [ cf . eqn ( [ n0 ] ) ] . in consequence , within this approximation , the surname diversity always grows exponentially . we point out , however , that this conclusion is valid for sufficiently smooth distributions and for @xmath93 , two necessary conditions for the continuous approximation to apply to our problem . it could therefore be argued that having a decreasing number of surnames , as found in some modern developed societies ( cavalli - sforza & feldman , 1981 ; legay & vernay , 1999 ) , requires violation of the above conditions . for instance , an initial condition with sharp variations would violate the smoothness condition during the first evolution stages . a death probability @xmath94 would also threaten the validity of the continuous approximation . this is precisely the case of modern developed societies , where the population growth rate is practically vanishing and , as a matter of fact , reaches occasionally negative values a situation not described by the present model . in section [ mu1 ] we treat the special case @xmath95 and show that , in this limit , the total number of surnames can in fact decrease . though the first - order continuous approximation gives a rather rough description of the solution to our model , as a piecewise function with a moving discontinuity , it provides a quite clear qualitative picture of how the solution behaves . the growth of the power - law regime as the initial condition recedes is in good qualitative agreement with the evolution observed in numerical realizations of the model . a better analytical approximation is obtained from the second - order truncation . second - order derivatives would in fact introduce diffusive - like effects in the variable @xmath69 , with the consequent smoothing of the discontinuities of the first - order approximation . the second - order equation , however , can not be analytically solved for arbitrary initial conditions . nevertheless , it is possible to give an asymptotic approximation for long times , as @xmath96 where @xmath97 is the kummer s function ( see appendix [ ap2 ] ) . with respect to the first - order approximation , eqn ( [ sol1 ] ) , this solution predicts a lower value of @xmath70 for small @xmath69 . for larger family sizes , however , it behaves as a power law , @xmath98 with exactly the same exponent as in eqn ( [ zeta ] ) . the asymptotic behaviour for large @xmath69 in the power - law regime is therefore not modified . it must be pointed out that , for small values of @xmath15 , the profile of @xmath70 given by eqn ( [ secorder ] ) results to be quite independent of the death probability over a considerable range of values of @xmath16 . only for @xmath99 , where the exponent @xmath100 strongly depends on @xmath16 , does @xmath70 change sensibly as @xmath16 is varied . we take advantage of this feature in section [ empirical ] , where the second - order approximation is compared with actual data of surname distributions . as argued above , the limit @xmath95 is relevant to the discussion of the evolution of surname distributions in some modern developed societies . in fact , this limit corresponds to the case where the population growth rate vanishes , a situation which is closely approached , for instance , in many european countries . in this case , eqn ( [ simonm ] ) is again deterministic , with @xmath101 and @xmath102 . the total population at any time is @xmath103 . the main difference between this case and that of @xmath104 is that , now , the distribution @xmath49 becomes independent of time for asymptotically large times . this feature is in agreement with the fact that the asymptotic distribution for @xmath104 is proportional to the total population @xmath18 . the asymptotic surname distribution @xmath105 is given , for @xmath27 , by the recurrence equations @xmath106i{\overline}n_i^\infty -(1-\alpha ) ( p_0 + 1-i)(i-1){\overline}n_{i-1}^\infty}{(i+1)[p_0-(1-\alpha)(i+1)]}\ ] ] for @xmath44 , and @xmath107{\overline}n_1^\infty - \alpha p_0 ^ 2}{2[p_0 - 2(1-\alpha)(i+1 ) ] } .\ ] ] note that , since the total population is always @xmath26 , we have @xmath108 for @xmath109 , as no family can be larger than @xmath26 . the solution reveals two well - defined regimes , depending on how the product @xmath110 compares with unity . for @xmath111 , the asymptotic distribution behaves as @xmath112 for a vast range of family sizes . departures from this behaviour are found very close to @xmath113 only . for @xmath109 , in fact , the distribution must vanish . we stress the remarkable difference between the exponential stationary distribution ( [ est ] ) and the long - time power - law solution obtained for @xmath104 . in this regime , the stationary total number of surnames is @xmath114 for @xmath115 , on the other hand , the distribution behaves as a power law , @xmath116 , over practically the whole range of family sizes . note that the exponent of this power law is in agreement with eqn ( [ sol1 ] ) in the limit @xmath117 . for @xmath118 , however , the distribution deviates from the power law and exhibits a sharp peak . in the limit @xmath119 , the distribution becomes an isolated peak at @xmath113 , namely , @xmath120 for @xmath121 and @xmath122 for @xmath113 . therefore , the total number of surnames is @xmath123 . this special solution describes the well - known case of a closed population with no surname mutations where , by random drift , a single surname survives for asymptotically large times ( cavalli - sforza & feldman , 1981 ) . notice also that this limit corresponds to the birth - death model introduced by moran ( 1962 ) to study probabilities of fixation of alleles when generations overlap . we conclude that , for a given population @xmath26 , the asymptotic number of surnames can be very small if the mutation probability @xmath15 is , in turn , small enough . consequently , in a steady population with many surnames at the initial stage and with a sufficiently low mutation probability , the number of surnames will decrease towards the stationary value as time elapses . numerical simulations , discussed in detail in the next section , show that for @xmath16 just below unity and small @xmath15 , an initial transient where the number of surnames decreases temporarily can be observed . since @xmath104 , however , the population grows steadily and , as a consequence of mutations , @xmath124 will also increase in the long run . the situation in modern developed societies is that the population growth rate has been constantly decreasing , to reach values around zero at present times . starting from a state with a wealth of surnames due to the combined effect , a few centuries ago , of a high population growth and the frequent appearance of new surnames these societies have now reached a regime of almost stationary population where the number of surnames decreases . this situation would be reverted if the growth rate could be maintained above zero during substantial periods . presently , the only mechanism acting in this direction seems to be immigration , which is in addition an effective source of new surnames . in this section we present results of numerical realizations of our model , in order to compare with the analytical approximations presented in section [ analytic ] , and to illustrate the behaviour within the regimes where such approximations do not hold . emphasis is put on the role of the initial conditions , the duration of transients , and the evolution of the number of different surnames . also , we discuss actual surname distributions of modern populations in the light of our results . first , it is worthwhile to illustrate how the shape of the distribution @xmath49 depends on the initial condition @xmath57 [ see also zanette & manrubia ( 2001 ) ] . we focus the attention on initial distributions of the form @xmath125 for @xmath126 , and @xmath127 for @xmath128 , in which there are @xmath59 families of equal size formed by @xmath129 individuals each . consequently , the initial population is @xmath130 . in the following , we denote such an initial condition by the pair @xmath131 . figure 3 shows four normalized distributions of family sizes , @xmath132 obtained both from the numerical realization of the stochastic birth - death model and from the iterative solution of the deterministic eqns ( [ simon2 ] ) and ( [ simon3 ] ) , for @xmath27 . we find a very good agreement between both methods and , at the same time , clearly realize the relevant role of the initial condition in determining the profile of the distribution for large family sizes . these same features are found for other values of the death probability . the solution of the first - order approximation to our model , eqns ( [ sol1 ] ) to ( [ yc ] ) , predicts that in the zone of small family sizes i.e . for @xmath79 in the continuous variables the only dependence on the initial condition appears through the quantity @xmath26 this means that the distribution of family sizes is not sensitive to a variation of @xmath129 and @xmath59 as far as their product is kept constant . note that the three cases @xmath133 , @xmath134 , and @xmath135 of fig . 3 share the same value of @xmath136 . for the parameters of the figure , the crossover between the regions of small and large family sizes , given for the continuous variables by eqn ( [ yc ] ) , should occur at @xmath137 . though two of the distributions indeed have the same profile with the expected power - law decay up to that value , the distribution corresponding to the initial condition @xmath134 deviates considerably below @xmath138 . this deviation can be attributed to the fact that the initial condition @xmath134 corresponds to a quite singular distribution , with a high peak at @xmath139 . a continuous approximation for such a distribution is arguably expected to give a poor description of the real situation . equation ( [ yc ] ) defines , within the first - order approximation , the boundary that separates the asymptotic regime and the zone dominated by the initial condition . alternatively , for a fixed family size , it can be used to determine the transient before the asymptotic distribution builds up at that size . in order to test this aspect of the first - order approximation , we have devised an independent computational method to evaluate the point at which the cut - off between the two regimes actually takes place . according to eqns ( [ sol1 ] ) and ( [ p_i ] ) , the asymptotic normalized distribution of family sizes is @xmath140 . in numerical realizations , the distribution is expected to adopt values very close to @xmath141 for large @xmath38 and large enough @xmath17 . in the simulations , we fix a certain family size @xmath138 and stop the calculation at the step @xmath142 when the measured value of @xmath143 satisfies , for the first time , @xmath144 . the results presented in the following correspond to the choice @xmath145 and @xmath146 . since , as shown in section [ initcond ] , the agreement between the numerical simulations of the stochastic birth - death model and the iterative solution of the corresponding average equations is very good , we use this second description to measure the transient @xmath142 . we focus the analysis on the dependence with the initial conditions , and keep the values of @xmath147 and @xmath27 fixed . as far as @xmath148 , the results are qualitatively the same for other values of @xmath16 . we consider initial conditions of the form @xmath131 , as defined in section [ initcond ] . figure 4 shows the transient @xmath142 measured from the numerical solution to eqns ( [ simon2 ] ) and ( [ simon3 ] ) according to the criterion introduced above , as a function of @xmath129 and @xmath59 . on the one hand , for small @xmath59 and , especially , for small @xmath129 , @xmath142 varies quite irregularly . on the other hand , we see that for larger values of @xmath129 and @xmath59 , say @xmath149 and @xmath150 , @xmath142 exhibits a well - defined linear dependence with both parameters . in this regime , the transient is well approximated by @xmath151 [ fig : i0 ] linear fitting of @xmath142 as a function of @xmath129 and @xmath59 for the data shown in fig . 4 yields @xmath152 . note that eqn ( [ stn ] ) implies that , for large @xmath129 and @xmath59 the transient length is proportional to the initial population @xmath153 . this is to be compared with the analogous result within the first - order approximation , eqn ( [ yc ] ) , which implies @xmath154.\ ] ] the first - order approximation predicts then a proportionality between @xmath142 and @xmath26 independently of the values of @xmath15 and @xmath16 in full agreement with the numerical results . we point out , however , that the corresponding proportionality coefficients can not be directly compared . in fact , the coefficient @xmath155 in eqn ( [ stn ] ) is evaluated following the computational definition of the transient , and thus depends on the parameter @xmath156 , which has no correlate in the analytical approach . note , nevertheless , that for @xmath148 and @xmath27 eqn ( [ sta ] ) predicts a coefficient approximately equal to @xmath157 , which is of the same order as the numerical result . as discussed in section [ model ] , the evolution of the average number @xmath158 of different surnames is driven in our model by two competing mechanisms . the diversity increases due to the appearance of new surnames at rate @xmath15 , and surnames borne by single individuals disappear when the individual in question dies . within the first - order continuous approximation to our model , for @xmath104 , the average number of different surnames increases linearly with @xmath17 as @xmath159 ( see section [ analytic ] ) . in this approximation , the death probability @xmath16 plays a role in the variation of diversity as a function of time , eqn ( [ nmu ] ) . on the one hand , increase of the surname diversity in a steadily growing population is generally expected when new surnames are created at a constant rate . on the other , it is possible to conceive special situations where the number of different surnames should temporarily decrease , violating the first - order approximation . imagine , for instance , that the initial population consists of an ensemble of families with only one individual each . for moderate values of @xmath16 and small @xmath15 , the initial stage would be characterized by the death of some individuals , with the consequent disappearance of their surnames , and no significative appearance of new surnames . since the total population grows , however , @xmath158 will eventually attain a minimum and , from then on , will increase . to illustrate this situation , we numerically solve eqns ( [ simon2i ] ) to ( [ simonm ] ) for the initial condition @xmath134 , with @xmath160 families of one individual each . note that this is the initial condition for which we detected the largest deviations from the first - order approximation in section [ initcond ] . figure 5 shows the evolution of @xmath158 for different values of @xmath16 . as expected , an initial transient where the surname diversity drops is found for @xmath161 . the transient is longer and the minimum in @xmath162 is deeper as @xmath16 grows . in all cases , however , the subsequent growth of @xmath158 is clearly seen . note that the slope of this growth depends slightly on @xmath16 , a feature not predicted by the first - order continuous approximation . these long transients where ( for large @xmath16 and suitable initial conditions ) the number of different surnames is expected to decrease could be relevant to the description of modern populations with declining diversity ( cavalli - sforza & feldman , 1981 ; legay & vernay , 1999 ) . a realistic description of this situation should however take into account that , in recent times , the relative values of birth and death rates in actual populations have considerably changed . the general trend to population growth observed worldwide in the nineteenth century has by now been reversed in many developed areas , such as in europe , where the total population is practically stationary . in our model , this corresponds to an increment in the value of @xmath16 to values close to unity . in particular , the death probability must be allowed to vary with time . additionally , if the description is expected to encompass the periods where the appearance of new surnames was frequent specifically , the middle ages in the case of europe @xmath15 should also change during the evolution . as a qualitative demonstration of the effect of varying the model parameters with time , we focus the attention on a change of @xmath16 from a low value to a high value . for the initial condition @xmath134 we fix @xmath27 during the first @xmath163 evolution steps . then , @xmath16 is left to increase linearly with @xmath17 , such that it reaches unity after @xmath164 additional steps . in the numerical calculations shown in fig . 6 we fix @xmath165 and consider several values of @xmath163 . for @xmath166 , the number of different surnames increases ( cf . fig . 5 ) . then , as the death probability grows , @xmath158 attains a maximum and begins to decrease . for asymptotically large times , it is expected to approach a stationary value , as predicted for the case @xmath95 in section [ mu1 ] . interestingly , @xmath158 responds faster to the effects of growing @xmath16 for smaller values of @xmath163 , where the contribution of the initial condition is still important during the variation of the death probability . the asymptotic surname distribution seems to be , in this sense , quite robust to the action of varying @xmath16 . this feature should be related to the fact that the asymptotic distribution is rather insensitive to the value of @xmath16 , as discussed at the end of section [ analytic ] . finally , we compare some of the analytical results for our model , derived under several approximations in section [ analytic ] , with actual data from three modern populations . specifically , we focus the attention on the second - order approximation , eqn ( [ secorder ] ) , for the asymptotic distribution of surnames in the range of small family sizes , where the effect of the initial condition is negligible . in fact , it is virtually impossible to extract information on the distribution of surnames in historical times from the data presently available . we recall that eqn ( [ secorder ] ) correctly describes the asymptotic power law @xmath98 with @xmath167 , as observed in real data . the validity of our second - order approximation is therefore to be especially evaluated in the region of very small family sizes , where the distribution differs from the power law . our three data sets were obtained from surname counts in telephone books . they correspond to ( i ) the almost 350,000 different surnames of the whole 1996 argentine telephone book , ( ii ) the 6,400 surnames beginning by a in the 1996 berlin telephone book , and ( iii ) the surnames in five japanese cities , with populations ranging between @xmath168 and @xmath169 , reported by miyazima _ et al . _ ( 2000 ) . let us point out that the three populations involved here have considerably different demographic history . the modern argentine population has predominantly european ancestors , who immigrated mainly in the period 1880 - 1915 and just after the world war ii . their surname distribution has therefore to be considered as a combined sample from the countries that contributed the immigrants . from the times of the largest immigration waves to the present , both the birth and death rates have substantially changed . as for berlin , this city was practically abandoned in the late stages of world war ii and subsequently repopulated with a mixture of the ancient inhabitants and newcomers from other cities of germany . in this case , the surname distribution has consequently resulted from a combination of several german regions . the modern population in berlin , in addition , presents the particularity of having been artificially separated into two practically immiscible groups during several decades , ending in 1989 . since european surnames appeared mostly during the middle ages ( legay & vernay , 1999 ) , the populations that contributed surnames to both argentina and berlin are expected to have developed the asymptotic surname distribution on a substantial range of family sizes . this situation contrasts with the case of modern japanese surnames , that were originated some @xmath170 years ago ( miyazima _ et al . _ , 2000 ) . in japan , moreover , the contribution of immigration in the relevant period should have been considerably less important than in berlin and argentina . explicit evaluation of eqn ( [ secorder ] ) requires fixing the values of the mutation rate @xmath15 and the death probability @xmath16 . we know that the mutation rate , i.e. the probability that a new individual acquires a surname not previously present in the population , is very low for any modern society . since , as far as @xmath171 , the limit of eqn ( [ secorder ] ) for @xmath172 is well defined , fixing any sufficiently small value for @xmath15 gives a correct description of @xmath70 . in our comparison with real data , we have taken @xmath173 . as for the death probability @xmath16 , unfortunately , we have found it impossible to fix a reliable value . for all real populations , mortality has considerably changed during the periods relevant to the evolution of surname frequencies . moreover , since @xmath16 measures the relative frequency of death and birth events the latter also including , in real populations , arrival of new individuals by immigration an evaluation of the death probability should also involve a detailed description of immigration effects . on the other hand , as discussed at the end of section [ analytic ] , the asymptotic profile of @xmath70 , given by eqn ( [ secorder ] ) , is practically independent of the death probability as far as @xmath16 is not close to unity . we therefore decided to fit the field data with eqn ( [ secorder ] ) allowing @xmath16 to vary in order to get the best approximation in the relevant zone of small family sizes . figure 7 shows the three data sets and the corresponding fittings with eqn ( [ secorder ] ) . fittings have been optimized in the domain of small family sizes , where our analytical approximation is expected to hold . in the case of argentina , the agreement with real data is excellent up to family sizes above @xmath174 . only for @xmath175 a systematic deviation is observed , where the analytical result overestimates the frequency . according to our discussion in section [ analytic ] , this deviation would be the remnant contribution of initial conditions . for the berlin data , the statistics are poorer . however , it is clear that the analytical approximation fits the data with good precision in the whole range shown here . in the case of the japanese data the fitting is very good for relatively small family sizes , @xmath176 , but , on the other hand , noticeable systematic deviations appear for @xmath177 . this agrees with our expectation on the effect of initial conditions . in the @xmath170 years elapsed from the appearance of modern japanese family names , the population in the cities from which the present data were obtained has increased by a factor of , at most , order @xmath178 . therefore , according to eqn ( [ yc ] ) , initial conditions should contribute to the distribution for relatively small family sizes , @xmath179 , just as observed . we have analysed a birth - death model with overlapping generations , in order to study several statistical properties of monoparental inheritance in large populations . we have focused the analysis on a cultural trait , namely , the distribution of surnames , taking advantage of the availability of big corpora of real data . however , the model applies generally to biological traits associated with non recombining alleles . our model has two parameters , namely , the probability that a new individual carries a surname not present in the population , @xmath15 , and the average number of death events per birth @xmath16 . for any value of @xmath180 the system attains a broad , stationary distribution of surname diversity . if @xmath104 the total population grows exponentially in time and so does the total number of different surnames . the marginal case @xmath95 corresponds to constant total population and , on average , to constant surname diversity for asymptotically large times . our analytical results for the stationary distribution of surnames frequency are in good agreement with field data for modern human populations in different countries . through an analysis of the transient time required for this distribution to reach its asymptotic shape , we have shown that some deviations observed in real data might actually reflect the composition of the founder population . this result has implications in the study of polyphyletism . indeed , if the same surname can have multiple origins and thus the individuals carrying it are not always phylogenetically related , this will affect the shape of the surname distribution . in particular , it is not difficult to estimate the time when surnames originated in a population ( using historical records ) or , in the biological counterpart , when a mutant allele first appeared ( through molecular clock analysis ) . then , the approximate cut - off until which the stationary distribution follows the asymptotic shape can be calculated . if the distribution underestimates the frequency of values larger than the cut - off , the system is mainly polyphyletic . if it overestimates that frequency , then the simultaneous appearance of many individual carrying different surnames took place in the past . the strong resemblance between the cultural inheritance of the surname and the biological process in which non recombining neutral alleles are passed to offspring has justified to apply results from field data in the former case to the latter ( barrai _ et al . _ , 1996 ) . in the few cases where data on genetic diversity was available , it was possible to retrieve information on past populations by comparing both sets of data ( sykes & irven , 2000 ) . a specific example comes from the small island of tristan da cunha , where the fact that its @xmath181 inhabitants represent only seven surnames and five mitochondrial lineages reflects without doubt the small size of the founder population ( soodyall _ et al . _ , 1997 ) . our results could be applied to population genetics under some hypotheses . indeed , we are assuming that the number of different alleles at a given locus is practically infinite ( this is analogous to the assumption made by kimura & crow ( 1964 ) , in their model of infinitely many alleles ) , since the possibility of backward mutations is discarded . nonetheless , this factor could be accounted for just by lowering the value of @xmath15 , because this process implies that the `` new '' mutant is in fact identical to one of the forms present in the population . since , as long as @xmath15 is small , our results are not sensibly modified , we could also work with a finite but large number @xmath182 of different surnames ( equivalently , large genetic polymorphism ) , and use the same model as long as the current diversity is lower than @xmath182 . we have also assumed that the death rate is constant during the life time of individuals , while it is known that the life expectancy depends not only on the age of individuals , but also varies as a function of time ( vaupel _ et al . _ , 1998 ) . a more realistic model of human inheritance could be constructed by taking into account the variation of @xmath16 along the lifetime of each individual . the evolution of surname diversity follows truly neutral evolution : family names do not fulfill any practical purpose but identifying the lineage of each individual . they can not be selected for through natural selection . hence , as we have shown , their statistical distribution closely follows the predictions of a neutral model of monoparental inheritance . it would be interesting to test our theoretical results against the genetic diversity of a large population sample . unfortunately , however , data on frequencies of allelic polymorphisms are still scarce to carry out a massive study like the one presented here for actual surname distributions . moreover , one should make sure that the different haplotypes in the sample do not confer any advantage to the individuals , in which case positive feedbacks and deviations from neutral statistics would be expected . different tests have been proposed in the literature to detect deviations from neutrality ( nielsen , 2001 ; fu & li , 1993 ) , and a number of neutral haplotypes have been positively identified ( snchez mazas _ et al . _ , 1994 ; stenoien , 1999 ) . hopefully , enough data will be available in a near future to calculate reliable diversity distributions in human populations . we thank m. montemurro for kindly suplying the whole surname count from the 1996 argentine telephone book . abramowitz , m. & stegun , i. a. ( 1970 ) . _ handbook of mathematical functions_. new york : dover . barrai , i. , scapoli , c. , beretta , m. , nesti , c. , mamolini , e. & rodriguez - larralde , a. ( 1996 ) . isonymy and the genetic structure of switzerland . i. the distribution of surnames . _ * 23 * _ , 431 - 455 . burlando , b. ( 1990 ) . the fractal dimension of taxonomic systems . . biol . * 146 * _ , 99 ; ( 1993 ) . the fractal geometry of evolution . _ j. theor . biol . * 163 * _ , 161 . cavalli - sforza , l. l. & feldman , m. w. ( 1981 ) . _ cultural transmission and evolution : a quantitative approach_. princeton : princeton university press . cavalli - sforza , l. l. , feldman , m. w. , chen , k. h. & dornbusch , s.m . theory and observation in cultural transmission . _ science * 218 * _ , 19 - 27 . consul , p. c. ( 1991 ) . evolution of surnames . stat . rev . * 59 * _ , 271 - 278 . darwin , c. ( 1871 ) . _ the descent of man and selection in relation to sex . _ london : murray . dewdney , a. k. ( 1986 ) . computer recreations : branching phylogenies of the paleozoic and the fortunes of english family names . * 254 * _ , 12 - 16 . fisher , r.a . ( 1922 ) . on the dominance ratio . _ 42 * _ , 321 - 341 . fox , w.r . & lasker , g.w . the distribution of surname frequencies . stat . rev . * 51 * _ , 81 - 87 . fu , y.x . & li , w.h . ( 1993 ) . statistical tests of neutrality of mutations . _ genetics * 133 * _ , 693 - 709 . gale , j.s . _ theoretical population genetics . _ unwin hyman , london . galton , f. & watson , h.w . ( 1874 ) . on the probability of the extinction of families . . inst . * 4 * _ , 138 - 144 . gomes , m. a. f. , vasconcelos , g. l. , tsang , i. j. & tsang , i. r. ( 1999 ) . scaling relations for diversity of languages . _ physica a * 271 * _ , 489 - 495 . greenberg , j. h. ( 1992 ) . preliminaries to a systematic comparison between biological and linguistic evolution . in : _ the evolution of human languages _ ( hawkins j. a. and gell - mann m. , eds ) reading : addison - wesley . haldane , j. b. s. ( 1927 ) . a mathematical theory of natural and artificial selection . part v : selection and mutation . * 26 * _ , 838 - 844 . harris , th . e. ( 1963 ) . _ the theory of branching processes . _ berlin : springer . hull , d. m. ( 1998 ) . a reconsideration of galton s problem ( using a two - sex population ) . * 54 * _ , 105 - 116 . islam , m. n. ( 1995 ) . a stochastic model for surname evolution . j. * 37 * _ , 119 - 126 . karlin , s. & mcgregor , j. ( 1967 ) . the number of mutant forms maintained in a population . fifth berkeley symp . stat . prob . * 4 * _ , 415 - 438 . kimura , m. & crow , j.f . ( 1964 ) . the number of alleles that can be maintained in a finite population . _ genetics * 49 * _ , 725 - 738 . lange , k. ( 1981 ) . minimum extinction probability for surnames and favorable mutations . biosci . * 54 * _ , 71 - 78 . legay , j. m. & vernay , m. ( 1999 ) . l instructive histoire des noms de familie . _ pour la science _ n. 255 , 58 - 65 . lotka , a. j. ( 1931 ) . population analysis the extinction of families i . _ washington acad . * 21 * _ , 377 - 380 ; _ ibid . _ 453 - 459 . miyazima , s. , lee , y. , nagamine , t. & miyajima , h. ( 2000 ) . power - law distribution of family names in japanese societies . _ physica a * 278 * _ , 282 - 288 . moran , p.a.p . ( 1962 ) . _ the statistical processes of evolutionary theory . _ clarendon press , oxford . nielsen , r. ( 2001 ) . statistical tests of selective neutrality in the age of genomics . _ heredity * 86 * _ , 641 - 647 . panaretos , j. ( 1989 ) . on the evolution of surnames . stat . rev . * 57 * _ , 161 - 167 . pielou , e. c. ( 1969 ) . _ an introduction to mathematical ecology . _ new york : wiley . poore , m. e. d. ( 1968 ) . studies in malaysian rainforest . _ ecology * 56 * _ , 143 - 196 . rannala , b. ( 1997 ) . gene genealogy in a population of variable size . _ heredity * 78 * _ , 417 - 423 . ruhlen , m. ( 1992 ) . an overview of genetic classification . in : _ the evolution of human languages _ ( hawkins , j.a . and gell - mann m. , eds ) reading : addison - wesley . snchez mazas , a. , butler brunner , e. , excoffier , l. , ghanem , n. , salem , m.b . , breguet , g. , dard , p. , pellegrini , b. , tikkanen , m.j . , langaney , a. , lefranc , g. , & butler , r. ( 1994 ) . new data for ag haplotype frequencies in caucasoid populations and selective neutrality of the ag polymorphism . _ human biology * 66 * _ , 27 - 48 . simon , h. a. ( 1955 ) . on a class of skew distribution functions . _ biometrika * 42 * _ , 425 - 440 . soodyall , h. , jenkins , t. , mukherjee , a. du toit , e. , roberts , d.f . & stoneking , m. ( 1997 ) . the founding mitochondrial dna lineages of tristan da cunha islanders . _ am . . anthropol . * 104 * _ , 157 - 166 . sol , r. v. and alonso , d. ( 1998 ) . random walks , fractals , and the origin of rainforest diversity . _ adv . complex systems * 1 * _ , 203 - 220 . stenoinen , h.k . ( 1999 ) . are enzyme loci selectively neutral in haploid populations of nonvascular plants ? _ evolution * 53 * _ , 1050 - 1059 . sykes , b. & irven , c. ( 2000 ) . surnames and the y chromosome . _ am . * 66 * _ , 1417 - 1419 . van kampen , n. g. ( 1981 ) . _ stochastic processes in physics and chemistry . _ amsterdam : north - holland . vaupel , j. w. , carey , j. r. , christensen , k. , johnson , t. e. , yashin , a. i. , holm , n. v. , iachine , i. a. , kannisto , v. , khazaeli , a. a. , liedo , p. , longo , v. d. , zeng , y. , manton , k. g. & curtsinger , j. w. ( 1998 ) . biodemographic trajectories of longevity . _ science * 280 * _ , 855 - 860 . yasuda , n. , cavalli - sforza , l. l. , skolnick , m. & moroni , a. ( 1974 ) . the evolution of surnames : an analysis of their distribution and extinction . biol . * 5 * _ , 123 - 142 . zanette , d. h. ( 2001 ) . statistical regularities in the taxonomic classification of human languages . complex systems * 4 * _ , 281 - 286 . zanette , d. h. & manrubia , s.c . vertical transmission of culture and the distribution of family names . _ physica a * 295 * _ , 1 - 8 . zipf , g. k. ( 1949 ) _ human behavior and the principle of least effort_. cambridge : addison - wesley . as explained in the main text , at each step , the total population @xmath18 in our model either increases by @xmath183 with probability @xmath184 , due to the occurrence of a birth event and no death event , or remains constant with probability @xmath16 , due to the occurrence of both a birth and a death event . equations ( [ pest ] ) and ( [ xi ] ) quantify the corresponding stochastic process . since the total population changes during the evolution , the time interval @xmath185 to be associated with each step which corresponds to the interval between consecutive births must also change . in fact , the birth frequency is proportional to the total population , so that @xmath185 is inversely proportional to @xmath186 . we write , at step @xmath17 , @xmath187 where the frequency @xmath24 fixes time units . in consequence , the real - time average variation in @xmath18 can be obtained from @xmath188 } = \nu { \overline}{[1-w(s)]p(s ) } = \nu ( 1-\mu ) { \overline}p , \ ] ] where overlines indicate average over realizations of the stochastic process @xmath20 . to obtain this average evolution equation we have used eqn ( [ pest ] ) , and have taken into account that @xmath20 and @xmath18 are independent stochastic processes at each step @xmath17 , so that @xmath189 . if @xmath24 is identified with the birth rate per individual and unit time , the product @xmath190 is the mortality rate . for constant @xmath24 and @xmath16 , the solution to eqn ( [ a1 ] ) is @xmath191,\ ] ] with @xmath192 the initial population . under the action of both birth and death events , the evolution of the average number of families of size @xmath38 , @xmath193 , is given by eqns ( [ simon2i ] ) and ( [ simonm ] ) . these equations can be approximately solved assuming that the solution varies slowly on @xmath17 and @xmath38 , so that these two discrete variables admit replacement by continuous variables @xmath68 and @xmath69 , respectively . accordingly , @xmath49 is replaced by a continuous function @xmath70 . the approximation is based on the expansions @xmath194 and @xmath195 , to be introduced in eqns ( [ simon2i ] ) and ( [ simonm ] ) at different truncation orders . to the first order , we obtain the differential equation @xmath196 for @xmath197 . the contribution for @xmath198 , given by eqn ( [ simon3i ] ) , can be incorporated to ( [ e1 ] ) either as a boundary condition or as a singular inhomogeneity in the equation itself . the latter yields @xmath199 where @xmath200 stands for the dirac delta distribution . in the following , we obtain and analyse the solution to this equation . introducing the auxiliary variables @xmath201 eqn ( [ e2 ] ) becomes @xmath202 with @xmath203 . this is a one - dimensional wave equation , with `` spatial '' variable @xmath204 and `` temporal '' variable @xmath205 . it describes shape - preserving advection of the `` density '' @xmath206 with constant velocity @xmath207 , subject to the action of a point source of intensity @xmath208 at @xmath209 . in our problem , thus , this equation makes sense for @xmath210 ( @xmath211 ) since @xmath204 must be non - negative ( @xmath212 ) . the general solution for @xmath213 is given by an arbitrary combination of functions of the form @xmath214 , where @xmath215 is in principle arbitrary . the combination must be chosen in such a way that the boundary and initial conditions are satisfied . in our case , this is achieved with a combination of two such functions . in the original variables , the solution is given piecewise as @xmath216 for @xmath72 , and @xmath217 for @xmath74 . here , @xmath218 is the initial distribution . the transition point between the two pieces is located at @xmath219 these expressions give the first - order continuous approximation @xmath220 to the distribution of families by size . the average total number of surnames in the continuous approximation is defined in ( [ avnc ] ) . for the first - order approximation we find @xmath221 the last integral is clearly equal to the initial number of surnames , @xmath59 , as may be immediately realized by the change of variables @xmath222 . explicit calculation of the first integral shows that @xmath223 , i.e. exactly the same result as for the case of @xmath27 , eqn ( [ ns ] ) . as a function of time , we have @xmath224 - 1\}.\ ] ] truncation to the second order in the continuous approximation to eqns ( [ simon2i ] ) and ( [ simonm ] ) yields @xmath225 } \frac{\partial^2}{\partial y^2 } ( yn)=\alpha \delta ( y-1).\ ] ] unfortunately , this equation can not be analytically solved for arbitrary initial conditions . however , a particular solution can be found in the form of a separate function , @xmath226 $ ] . comparison with eqns ( [ ni ] ) and ( [ sol1 ] ) suggests that this particular solution will correspond to the long - time asymptotic evolution . it reads @xmath227 where @xmath97 is the logarithmic kummer s function ( abramowitz & stegun , 1970 ) .

quantum coherence ( qch ) plays an important role in today s physics , more specifically in physics of low temperature starting from the laws of thermodynamics @xcite and most recently in the emerging area of quantum information theory . it has been useful in understanding the underlying process of efficient energy transfer in complex biological systems @xcite . in the recent past researchers have attempted to devolop a framework to formalize the theory of qch within the realms of quantum information theory and quantum resource theories @xcite . in this framework qch is quantified in terms of the functions of coherence monotones . some of these better known measures are based on @xmath0 norm , relative entropy @xcite , skew information @xcite and there are many others @xcite . several works were done in this context to extend the measures of coherence for multipartite systems@xcite . in another work , researchers studied the creation of qch using unitary operations with limited thermodynamic resources and eventually they found explicitly the unitary transformation that creates the maximal coherence @xcite . other works were done where people found the average relative entropy of coherence of pure quantum states sampled randomly from the uniform haar measure and showed that it to be typical @xcite . it was shown that in the asymptotic limit , the minimum amount of noise that is required to fully decohere a quantum system , is equal to the relative entropy of coherence . this is quantified by the cost of erasing qch @xcite . in another work it was demonstrated that measurements in a coherent basis are needed to make use of coherence , and that such measurements must be generated by coherent unitary interactions with an environment @xcite . recently some authors derive complementarity relations between coherence measures on mutually unbiased bases using various coherence measures @xcite . in a similar kind of work the robustness of coherence was suitably defined and was proven a full monotone in the context of the recently introduced resource theories of qch @xcite . another group of authors showed that all measures of coherence are frozen for an initial state in a strictly incoherent channel if and only if the relative entropy of coherence is frozen for the state @xcite . further work show the study the relation between qch and quantum discord , a kind of quantum correlation which appears even in non entangled states @xcite . a new framework for measuring qch in finite dimensional systems was introduced and a theoretical measure satisfying the reliability criteria was established in the context of quantum resource theories @xcite . + the impossibility to clone quantum states is regarded as one of the most fundamental restriction that comes from nature @xcite . the no cloning theorem " states that there exists no quantum mechanical process , which can take two distinct non - orthogonal quantum states @xmath1 , @xmath2 into states @xmath3 , @xmath4 respectively . even though we can not copy an unknown quantum state perfectly but quantum mechanics never rules out the possibility of cloning it approximately @xcite . it also allows probabilistic cloning as one can always clone an arbitrary quantum state perfectly with some non - zero probability of succsess @xcite . in the year 1996 , buzek et al . went beyond the idea of perfect cloning and introduced the concept of approximate cloning . in this process , they have also introduced the state independent quantum copying machine by keeping the fidelity of cloning independent of the input state parameters . this machine is also popularly known as universal quantum cloning machine ( uqcm ) @xcite . later in another work by gisin and massar this cloner was shown to be optimal @xcite . other than this state independent quantum cloning machine ( qcm ) , there are also state dependent qcms @xcite . state dependent quantum cloning machine can be both dynamic and static depending on the choice of machine parameters @xcite . in this context it was shown that how the fidelity of quantum cloning machine is related with the amount of correlation generated in the process of cloning @xcite . + quantum entanglement @xcite in general quantum correlation is considered as a useful resource as it plays a significant role in computational and communication processes like quantum key generation @xcite , secret sharing @xcite , teleportation @xcite , superdense coding @xcite , entanglement swapping @xcite , remote entanglement distribution @xcite and in many more tasks @xcite . the more pure " is entanglement , more valuable " is the given two - particle state at least from the point of view of information processing tasks . therefore it becomes important to extract pure entanglement from mixed entangled state . researchers had done a lot of work in past years on these extraction processes @xcite . this comes at the cost of reduction of number of entangled states . it is interesting to see that in quantum information processing scenario , there exists tasks which are completely opposite to this from the operational point of view . this comes under the umbrella of broadcasting of inseparability " @xcite . when there is an exigency in increasing the number of available entangled pairs rather than the purity of it , it becomes important to broadcast the amount of entanglement from a given pair to different pairs . one possible way to achieve this is by applying local and non local cloing . in general cloning in this context includes both state dependent and state independent cloning @xcite . in a recent work we have shown that it is impossible to optimally broadcast quantum correlation across different laboratories @xcite . but if we consider non optimal broadcasting it is indeed possible to broadcast quantum correlation between separated parties @xcite . other than application of cloning there are different ways to broadcast quantum correlation . quite recently , many authors showed by using sophisticated methods that correlations in a single bipartite state can be locally or uni locally broadcast if and only if the states are classical ( i.e. having classical correlation ) or classical - quantum respectively @xcite . + in this work our basic objective is to broadcast qch present in a pair of qubits to other pairs . keeping in mind the usefulness of qch in resource theories it is also important to broadcast it to more pair of qubits . here we have considered the @xmath0 norm as a measure of coherence . given a state @xmath5 , with its matrix elements as @xmath6 the amount of coherence present in the state @xmath5 in the basis @xmath7 is given by the quantity @xmath8 , where @xmath9 . it is interesting note that coherence is a basis dependent quantity as the representation of coherence will be different in different basis . more precisely since all these measures are functions of off diagonal elements of a given density matrix representation . the value of the coherence in the eigen basis where there are no off diagonal elements , will be zero . it is intuitive that for all other basis the physical findings followed by implications will be similar . so , without any loss of generality here we have computed the coherence in the computational basis @xmath10 . + let us begin with a situation where we have two distant parties @xmath11 and @xmath12 and they share a two qubit mixed state @xmath13 which can be canonically expressed as , @xmath14 = \left\{\overrightarrow{x},\:\overrightarrow{y},\ : t_{3\times 3}\right\},\:\:\ : \label{eq : mix}\end{aligned}\ ] ] where @xmath15 and @xmath16 are bloch vectors with @xmath17 and @xmath18 . here , @xmath19 s ( @xmath20 = @xmath21 ) are elements of the correlation matrix ( @xmath22_{3 \times 3}$ ] ) , whereas @xmath23 are the pauli matrices and @xmath24 is the identity matrix . + @xmath25{local}\includegraphics[height=4.5cm , width=4cm]{nonlocal } \end{array}\ ] ] we start with two qubit state @xmath13 and then apply universal symmetric optimal buzek hillery cloner both locally and non locally ( @xmath26 ) as a cloning transformation . in short , broadcasting will be possible if we are able to produce more coherent pairs out of initial coherent pair . in principle , broadcasting is both optimal and non optimal and depends upon the choice of cloning machine . + we separately investigate both _ optimal _ and _ non optimal broadcasting_. in principle , in order to broadcast the amount of coherence between the desired pairs @xmath27 and @xmath28 ( see fig 1 ) , we need to maximize the amount of coherence between the non - local output pairs , regardless of the states between @xmath29 and @xmath30 . however , for optimal broadcasting of coherence across parties , we need to minimize the amount of coherence within parties , i.e , in the states @xmath29 and @xmath30.this is because the total amount of coherence ( @xmath31 ) produced after cloning operation is equal to the sum of the coherence within parties ( @xmath32 ) and the coherence across the parties ( @xmath33 ) , i.e @xmath34 . to maximize @xmath33 , we must have @xmath35 . in other words , for optimal broadcasting we should have no coherence between the qubits @xmath29 and @xmath30 . + * definition : optimal broadcasting ( local / nonlocal ) * [ def : broad_local ] we start with two parties a and b sharing a coherent input state @xmath13 , with qubits 1 and 2 on a s side and b s side respectively . a coherent input state @xmath13 is said to be broadcast optimally after the application of local cloning operation @xmath36 , on the qubits 1 and 2 respectively , if for some values of the input state parameters , the non - local output states , given by @xmath37=tr_{23}\left[u_1\otimes u_2\left(\rho_{12}\right)\right],{}\nonumber\\ & & \tilde{\rho}_{23 } = tr_{14}\left[\tilde{\rho}_{1234}\right]=tr_{14}\left[u_1\otimes u_2 \left(\rho_{12}\right ) \right],\end{aligned}\ ] ] of a and b are coherent i.e @xmath38 , while the local output states , given by @xmath39= tr_{24}\left[u_1\otimes u_2\left(\rho_{12}\right)\right],{}\nonumber\\ & & \tilde{\rho}_{24 } = tr_{13}\left[\tilde{\rho}_{1234}\right]= tr_{13}\left[u_1\otimes u_2\left(\rho_{12}\right)\right ] , \end{aligned}\ ] ] of a and b are incoherent i.e @xmath40 . + [ def : broad_nonlocal]similarly , a coherent input state @xmath13 is said to be broadcast optimally after the application of non local cloning operation @xmath41 together on qubits 1 and 2 , if for some values of the input state parameters , the non - local output states , given by @xmath42=tr_{34}\left[u_{12}\left(\rho_{12}\right)\right ] , { } \nonumber\\ & & \tilde{\rho}_{34 } = tr_{12}\left[\tilde{\rho}_{1234}\right]=tr_{12}\left[u_{12 } \left(\rho_{12}\right ) \right],\end{aligned}\ ] ] of a and b are coherent i.e @xmath43 , while the local output states , given by @xmath39= tr_{24}\left[u_{12}\left(\rho_{12}\right)\right],{}\nonumber\\ & & \tilde{\rho}_{24 } = tr_{13}\left[\tilde{\rho}_{1234}\right]= tr_{13}\left[u_{12}\left(\rho_{12}\right)\right],\end{aligned}\ ] ] of a and b are incoherent i.e @xmath40 . here , in this work , we show that it is impossible to broadcast coherence optimally in both the cases of local and non local cloning . we show this for the computational basis @xmath10 starting with most general mixed state ( given in eq . ) as a resource with the help of the following theorems . + [ theo : local_broad_no_optimal ] _ given a two qubit general mixed state @xmath13 ( given by eq . ) and b - h local cloning transformations ( @xmath44 ( see appendix b ) ) , it is impossible to optimally broadcast the coherence within @xmath13 into two lesser coherent states : [ ( a ) ] @xmath45 = @xmath46 $ ] , [ ( b ) ] @xmath47 = @xmath48 $ ] ( here we have @xmath49 ) in the computational basis @xmath10 . _ here , we consider the input state as the most general two qubit mixed state @xmath13 ( given by eq . ) . the b - h local cloning transformation @xmath50 is applied to clone qubits ( @xmath51 and @xmath52 ) of @xmath13 . we have local output states as , @xmath53 , where @xmath54 and the non local output states @xmath55 . here , @xmath56 with @xmath57 and @xmath58 being the cloning machine parameters and @xmath59 represent the bloch vector and @xmath60 represents the correlation matrix of the input state . the coherence calculated by using @xmath0 norm of the local output states in the computational basis @xmath10 are given by , @xmath61 on a s side and @xmath62 on b s side . each of these quantities are positive as the norm @xmath63 are positive quantities lying between @xmath64 and @xmath65 . this is true for all values of @xmath58 lying between @xmath66 . since the coherence of local output states are non vanishing , it is clear enough that it is impossible to optimally brodcast quantum coherence . + now for @xmath67 , the above b - h local cloner @xmath44 reduces to state independent b - h optimal local cloner . subsequently , the coherence of local output states changes to , @xmath68 and @xmath69 . [ theo : nonlocal_broad ] _ given a two qubit general mixed state @xmath13 ( given by eq . ) and b - h non - local cloning transformations ( @xmath70(see appendix b ) ) , it is impossible to optimally broadcast the coherence within @xmath13 into two lesser coherent states : [ ( a ) ] @xmath71 = @xmath72 $ ] , [ ( b ) ] @xmath73 = @xmath74 $ ] . ( here we have @xmath75 ) in the computational basis @xmath10 . _ similar to the previous proof , here also we start with most general version of the two qubit mixed state @xmath13 ( given by eq . ) . when b - h non - local cloning transformation @xmath76 is applied non locally to clone qubits ( @xmath51 and @xmath52 ) of @xmath13 , then , we have local output states as , @xmath77 , where @xmath78 and the non local output states @xmath79 . here , @xmath80 with @xmath57 and @xmath58 being the cloning machine parameters and @xmath59 represent the bloch vector and @xmath60 represents the correlation matrix of the input state . the coherence calculated by using @xmath0 norm measure in the computational basis @xmath81 . hence , for the local output states , we have , @xmath82 on a s side and @xmath83 on b s side . each of these quantities are positive as the norm @xmath63 are positive quantities lying between @xmath64 and @xmath65 . this is true for all values of @xmath58 lying between @xmath84 . similarly here also we see that it is impossible to broadcast quantum coherence optimally as we have non vanishing quantum coherence for local output states . + now for @xmath85 , the above b - h non - local cloning machine @xmath70 reduces to state independent b - h optimal non - local cloner . subsequently , the coherence of local output states changes to , @xmath86 and @xmath87 . + at this point it is interesting to note that though it is impossible to broadcast optimally qch between the desired qubits . however , it does not rule out the entire possibility of broadcasting . if we relax the condition @xmath40 , then we show that subsequently it will be possible to broadcast qch . we show this for the most general two qubit mixed state as a resource . + * definition : non optimal broadcasting ( local / non local ) * a coherent state @xmath13 is said to be broadcast non optimally after the application of local cloning operation @xmath36 , on the qubits 1 and 2 respectively or by non local cloning operation @xmath41 together on qubits 1 and 2 , if for some values of the input state parameters , the coherence of non - local output states are greater than coherence of local output states , i.e , @xmath88 and @xmath89 in the local cloning case ; and , @xmath90 and @xmath91 in the non - local cloning case . + under this relaxation , the quantum coherence of the cloned state @xmath92 can either increase , or remain same or even decrease from the initial resource state @xmath13 , which implies the following three scenarios after broadcasting : * @xmath93 $ ] , @xmath94 $ ] @xmath95 @xmath96 * @xmath93 $ ] , @xmath94 $ ] @xmath97 @xmath96 * @xmath93 $ ] , @xmath94$]@xmath98 @xmath96 interestingly , in this work we show that it is indeed impossible to increase the quantum coherence by local and non local cloning operation though in general it is possible to increase the quantum coherence by local operations . we present this result in form of two following theorems , where we have considered general local and non local cloning transformations . [ theo : local_broad ] _ given a two qubit general mixed state @xmath13 ( given by eq . ) and b - h local cloning transformations ( @xmath44 ) , we can always create two states @xmath45 and @xmath99 , such that : ( a ) @xmath100 ( b ) @xmath101 in the computational basis @xmath10 . _ we start with most general representation of two qubit mixed state @xmath13 ( given by eq . ) . when buzek hillary local cloning transformation @xmath50 is applied to @xmath13 , we get the non local output states as @xmath102 . here , @xmath56 with @xmath57 and @xmath58 being the cloning machine parameters and @xmath59 represent the bloch vector and @xmath60 represents the correlation matrix of the input state . using @xmath0-norm as measure of coherence we find out the coherence in the computational basis @xmath10 , @xmath103 $ ] and @xmath104 $ ] . where , @xmath105 if we can show that the conditions @xmath106 , @xmath107 and @xmath108 hold true , then we can always conclude that @xmath109 @xmath98 @xmath110 . + at first , if we observe the terms @xmath111 and @xmath112 carefully , we find that @xmath112 @xmath97 @xmath113 @xmath95 @xmath111 because @xmath114 + now , we need to compare @xmath115 and @xmath116.if we observe the terms in @xmath115 and @xmath116 , and ignoring the constant @xmath117 , they can be thought of as distance between two points in a cartesian plane . let us denote the distance between two points a and b in a 2-dimensional space ( like xy cartesian plane ) by @xmath118.let us consider @xmath119 and as @xmath120 , so , we have , @xmath121 . now , assume 4 points in cartesian plane , namely , @xmath122 , @xmath123,@xmath124 and @xmath125 without @xmath126 as the origin . now , as @xmath120 , hence , it is always true that @xmath127 ; that is , @xmath128 . so , d always lies on @xmath129 , inside @xmath130 . @xmath131{triangle } \end{array}\ ] ] now , from the figure 2 , we have , @xmath132 and @xmath133 . now , we have proved @xmath134 in the appendix a by plane geometry . hence , we have,@xmath135 similarly , we can show @xmath136 hence , we have proved that @xmath137 [ theo : non_local_broad ] _ given a two qubit general mixed state @xmath13 and b - h non local cloning transformations ( @xmath70 ) , we can always create two states @xmath71 and @xmath73 , such that : ( a ) @xmath138 ( b ) @xmath139 in the computational basis @xmath10 , where @xmath140 , is the machine parameter . _ first of all we apply buzek hillary non - local cloning transformation @xmath76 to the 2- qubit general mixed quantum states @xmath13 ( eq . ) . the non local output states obtained as aresult of this are given by , @xmath79 . here , @xmath80 with @xmath57 and @xmath58 being the cloning machine parameters , @xmath59 represent the bloch vector and @xmath60 represents the correlation matrix of the input state . the qch present in these states are calculated in the computational basis @xmath10 by using the @xmath0 norm are given by , @xmath141 and @xmath142 where , @xmath143\end{aligned}\ ] ] now , we know , @xmath144 and @xmath145 . so , clearly , @xmath146 and @xmath147 . hence , we have created two weaker ( in terms of coherence ) correlations from input correlation after non - local cloning . in this section , we consider the standard examples of two qubit mixed state as a resource state . these include werner like states ( wls ) and bell diagonal states ( bds ) . we apply the standard bh cloner both locally and non locally , to find the range where at least non optimal broadcasting is possible , in terms of input state parameters . in both the examples , we have chosen computational basis @xmath10 for measuring coherence via @xmath0-norm . + * _ example : a. werner like states ( wls ) : _ * + as our first example , we consider werner - like states ( wls ) , which can be formally expressed in terms of bloch vectors as , @xmath148 , where @xmath149 is the bloch vector and the correlation matrix is given by , @xmath150 with the condition @xmath151 and @xmath152 . + + the local output states obtained after applying general local cloning operation ( @xmath44 ) on both the qubits @xmath65 and @xmath153 are given by , @xmath154 , where @xmath155 is the bloch vector of the state @xmath156 , @xmath157 $ ] and @xmath158 , where @xmath57 and @xmath58 being the machine state parameter ( see appendix b ) . also , we have the non - local output states as , @xmath159 , where @xmath155 is bloch vector , @xmath160 is the correlation matrix of the state @xmath156 . + now , coherence for the local output state is independent of @xmath161 and @xmath162 and it is given by , @xmath163 . on the other hand , coherence of non - local output state depends on @xmath161 , @xmath162 , @xmath58 and it is given by , @xmath164 . for non optimal broadcasting of coherence , our region of interest is the one where the condition @xmath165 or @xmath166 is satisfied . this condition eventually reduces to @xmath167 for general local cloning case . + in order to obtain the state independent local cloner , we substitute the value of @xmath58 as @xmath168 . the corresponding values of coherence of the local output states , if we perform state independent cloning , are given by , @xmath169 and the coherence of non - local output state is given by , @xmath170 . subsequently , the condition under which non optimal broadcasting would be possible reduces to @xmath171 . + .broadcasting ranges for local state independent cloning in case of werner like states . [ cols="<,<",options="header " , ] in table [ tab : bell_sidnl_table ] , we have shown some of the trends in broadcasting ranges obtained in terms of @xmath172 , @xmath173 and @xmath174 after state independent non - local cloning transformations . we fix the values of @xmath172 and @xmath174 and substitute their respective values in the expression @xmath175 > 1.$ ] here , from @xmath176 onwards , discontinuities in the intervals corresponding to the ranges of @xmath173-parameter are observed , which imply the presence of a no - broadcast zone in between . + or @xmath177 ) after normalised in a scale of 0 to 1.,title="fig:",width=283,height=245 ] + in figure [ fig : bell_sidnl ] , the region of broadcasting is superimposed on the tetrahedral - region hosting all bell - diagonal states in @xmath178-coordinate system , with axes @xmath172 , @xmath173 and @xmath174.the light light regions represent the no - broadcast zones , while the regions on the dark blue - magenta spectrum , occupying a significant fraction of the tetrahedron s volume , highlight the regions where we can broadcast coherence . in the bluish - magenta region , the change in hue signifies how the values of the non - local coherence , @xmath179 or @xmath177 , changes with respect to the values of the @xmath178-parameters . as the value of @xmath179 or @xmath177 increases , the hue changes from blue to magenta . for instance , the concentrated presence of magenta - pink hue , especially in the regions along the tetrahedron s edge and in the regions surrounding the corners , suggests that non - local coherence is much higher in those regions , compared to the other regions in the broadcasting zone . also , we find that there is a light purple cavity - like zone sandwiched in between the bluish - magenta regions , which points to a no - broadcast zone . this cavity - like zone accounts for discontinuities int the intervals corresponding to ranges of @xmath173-paramter in table [ tab : bell_sidnl_table ] . moreover , we observe in the figure that in terms of coverage of the tetrahedron s volume , the zone of broadcasting in the non - local cloning scenario is significantly larger when compared to that of the local cloning scenario ( figure [ fig : bell_sidl ] ) , thereby suggesting that non - local state independent cloning performs much better when it comes to broadcasting of coherence in the bell - diagonal states . + _ acknowledgements : _ authors acknowledge mr . tushant jha and mr . sourav chatterjee for their suggestions . 99 m.horodecki , j. openheim , http://www.nature.com/ncomms/2013/130626/ncomms3059/full/ncomms3059.html[nat . commun * 4 * 2059 ( 2013 ) ] ; c. a. rodriguez - rosario _ et . al _ , https://arxiv.org/abs/1308.1245[arxiv : 1308.1245 ] ; m. lostaglio _ et . al _ , http://dx.doi.org/10.1103/physrevx.5.021001[phys . x * 5 * , 0210001 ( 2015 ) ] ; l. a. correa _ et . al _ http://www.nature.com/articles/srep03949[scientific reports * 4 , 3949 * ( 2014 ) ] ; p. skrzypczyk , a. j. short , and s. popescu , http://www.nature.com/ncomms/2014/140627/ncomms5185/abs/ncomms5185.html[nature communications * 5 , 4185 * ( 2014 ) ] ; m. lostaglio , d. jennings , and t. rudolph , http://www.nature.com/ncomms/2015/150310/ncomms7383/full/ncomms7383.html?wt.ec_id=ncomms-20150311[nat . commun . * 6 , 6383 * ( 2015 ) ] ; m. lostaglio , k. korzekwa , d. jennings , and t. rudolph , http://dx.doi.org/10.1103/physrevx.5.021001[phys . x * 5 * , 021001 ( 2015 ) ] ; k. korzekwa , m. lostaglio , j. oppenheim , and d. jennings , http://iopscience.iop.org / article/10.1088/1367 - 2630/18/2/023045/meta[new j. phys . 18 , 023045 ( 2016 ) ] ; p. kammerlander and j. anders , http://www.nature.com/articles/srep22174[scientific reports * 6 , 22174 * ( 2016 ) ] ; g. vacanti , c. elouard , and a. auffeves , ( 2015 ) , https://arxiv.org/abs/1503.01974[arxiv:1503.01974 ] ; a. l. malvezzi , g. karpat , b. akmak , f. f. fanchini , t. debarba , and r. o. vianna , http://dx.doi.org/10.1103/physrevb.93.184428[phys . rev . b 93 , 184428 ( 2016 ) ] ; s. bhattacharya , s. banerjee , and a.k . pati , ( 2016 ) , https://arxiv.org/abs/1601.04742[arxiv:1601.04742 ] ; y. yang and g. chiribella , ( 2015 ) , http://arxiv.org/abs/1502.00259v2[arxiv:1502.00259v2 ] . s. llyod , http://iopscience.iop.org/article/10.1088/1742-6596/302/1/012037/meta[j phys : conf . * 302 * , 012037 ( 2011 ) ] ; s. huelga , m. plenio , http://www.tandfonline.com / doi / abs/10.1080/00405000.2013.829687[contemp . phys . , * 54 * , 181 ( 2013 ) ] ; f. levi , f. mintert , http://iopscience.iop.org/article/10.1088/1367-2630/16/3/033007/meta[new . j. phys , * 16 * , 033007 ( 2014 ) ] ; c.m . li , n. lambert , y.n . chen , g.y . chen , and f. nori , http://www.nature.com/articles/srep00885[sci . * 2 , 885 * ( 2012 ) ] ; v. singh poonia , d. saha , and s. ganguly , ( 2014 ) , https://arxiv.org/abs/1408.1327[arxiv:1408.1327 ] . e. chitambar , a. streltsov , s. rana , m. n. bera , g. adesso , and m. lewenstein , http://dx.doi.org/10.1103/physrevlett.116.070402[phys . lett . * 116 * , 070402 ( 2016 ) ] ; a. winter and d. yang , http://dx.doi.org/10.1103/physrevlett.116.120404[phys . lett . * 116 * , 120404 ( 2016 ) ] ; i. marvian and r.w . spekkens , http://dx.doi.org/10.1103/physreva.90.062110[phys . rev . a * 90 * , 062110,(2014 ) ] ; j. aberg , ( 2006 ) , https://arxiv.org/abs/quant-ph/0612146[arxiv:quant-ph/0612146v1 ] ; g. gour , i. marvian , and r.w . spekkens , http://dx.doi.org/10.1103/physreva.80.012307[phys . a * 80 * , 012307 ( 2009 ) ] ; j. chen , j. cui , and h. fan , ( 2015 ) , https://arxiv.org/abs/1509.03576[arxiv:1509.03576v1 ] ; j. chen , n. johnston , c .- k . li , and s. plosker , ( 2016 ) , https://arxiv.org/abs/1601.06269[arxiv:1601.06269 ] ; s. cheng and m.j.w . hall , http://dx.doi.org/10.1103/physreva.92.042101[phys . a * 92 * , 042101 ( 2015 ) ] ; p. deb , http://dl.acm.org/citation.cfm?id=2915444[quantum information processing , * 15(4 ) * 1629 - 1638(2016 ) ] ; s. du , z. bai , and y. guo , http://dx.doi.org/10.1103/physreva.91.052120[phys . rev . a * 91 * , 052120 ( 2015 ) ] ; m. n. bera , t. qureshi , m. a. siddiqui , and a. k. pati , http://dx.doi.org/10.1103/physreva.92.012118[phys . rev . a * 92 * , 012118 ( 2015 ) ] ; x. liu , z. tian , j. wang , and j. jing , http://www.sciencedirect.com/science/article/pii/s0003491615004340[annals of physics , * 366 * 102 - 112 ( 2016 ) ] ; k. von prillwitz , l. rudnicki , and f. mintert , http://dx.doi.org/10.1103/physreva.92.052114[phys . a * 92 * , 052114 ( 2015 ) ] ; m. hillery , http://dx.doi.org/10.1103/physreva.93.012111[phys . a * 93 * , 012111 ( 2016 ) ] ; k. bu , u. singh , and j. wu , ( 2015 ) , http://dx.doi.org/10.1103/physreva.93.042326 [ phys . rev . a 93 , 042326 ( 2016)];. j. m. matera , d. egloff , n. killoran , and m. b. plenio , ( 2015 ) , http://arxiv.org/abs/1512.07486[arxiv:1512.07486v1 ] ; y.r . zhang , l .- h . shao , y. li , and h. fan , http://dx.doi.org/10.1103/physreva.93.012334[phys . a * 93 * , 012334 ( 2016 ) ] ; u. singh , m. n. bera , h. s. dhar , and a. k. pati , http://dx.doi.org/10.1103/physreva.91.052115[phys . rev . a * 91 * , 052115 ( 2015 ) ] . t. baumgratz . al _ , http://dx.doi.org/10.1103/physrevlett.113.140401[phys . * 113 * , 140401 ( 2014 ) ] . davide girolami , http://dx.doi.org/10.1103/physrevlett.113.170401[phys . 113 , 170401 ( 2014 ) ] . x. yuan , h. zhou , z. cao , and x. ma , http://dx.doi.org/10.1103/physreva.92.022124[phys . a * 92 * , 022124 ( 2015 ) ] ; d. p. pires , l. c. cleri , and d. o. soares - pinto , http://dx.doi.org/10.1103/physreva.91.042330[phys . rev . a * 91 * , 042330 ( 2015 ) ] ; b. yadin and v. vedral , http://dx.doi.org/10.1103/physreva.93.022122[phys . rev . a * 93 * , 022122 ( 2016 ) ] ; s. rana , p. parashar , and m. lewenstein , http://dx.doi.org/10.1103/physreva.93.012110[phys . a * 93 * , 012110 ( 2016 ) ] ; m. allegra , p. giorda , and s. lloyd , http://dx.doi.org/10.1103/physreva.93.042312[phys . rev . a * 93 * , 042312 ( 2016)];r . chandrashekar , p. manikandan , j. segar , and t. byrnes , http://dx.doi.org/10.1103/physrevlett.116.150504[phys . lett . * 116 * , 150504 ( 2016 ) ] ; e. chitambar and g. gour , http://dx.doi.org/10.1103/physrevlett.117.030401[phys . 117 , 030401 ( 2016 ) ] ; m. piani , m. cianciaruso , t. r. bromley , c. napoli , n. johnston , and g. adesso , http://dx.doi.org/10.1103/physreva.93.042107[phys . rev . a 93 , 042107 ( 2016 ) ] . x. hu , a. milne , b. zhang , and h. fan , http://www.nature.com/articles/srep19365[scientific reports 6 , 19365 ( 2016 ) ] ; e. chitambar and m .- h . hsieh , http://dx.doi.org/10.1103/physrevlett.117.020402[phys . 117 , 020402 ( 2016 ) ] ; a. streltsov , s. rana , m. nath bera , and m. lewenstein,(2015 ) , http://arxiv.org/abs/1509.07456[arxiv:1509.07456v2 ] ; t. r. bromley , m. cianciaruso , and g. adesso , http://dx.doi.org/10.1103/physrevlett.114.210401[phys . lett . * 114 * , 210401 ( 2015 ) ] ; a. streltsov , u. singh , h. s. dhar , m. n. bera , and g. adesso , http://dx.doi.org/10.1103/physrevlett.115.020403[phys . lett . * 115 * , 020403 ( 2015 ) ] ; y. yao , x. xiao , l. ge , and c. p. sun , http://dx.doi.org/10.1103/physreva.92.022112[phys . a * 92 * , 022112 ( 2015 ) ] . avijit misra , uttam singh , samyadeb bhattacharya , arun kumar pati , https://arxiv.org/abs/1602.08437v2[arxiv:1602.08437v2 ] . uttam singh , l. zhang , and a. k. pati , ( 2015 ) , http://dx.doi.org/10.1103/physreva.93.032125[phys . a 93 , 032125 , ( 2016 ) ] . uttam singh , manabendra nath bera , avijit misra , arun kumar pati , http://arxiv.org/abs/1506.08186[arxiv:1506.08186v1 ] . b. yadin , j. ma , d. girolami , m. gu , and v. vedral , ( 2015 ) , https://arxiv.org/abs/1512.02085[arxiv:1512.02085v1 ] . carmine napoli , thomas r. bromley , marco cianciaruso , marco piani , nathaniel johnston , gerardo adesso , http://dx.doi.org/10.1103/physrevlett.116.150502[phys . 116 , 150502 ( 2016 ) ] . debasis mondal , tanumoy pramanik , arun kumar pati , http://arxiv.org/abs/1508.03770[arxiv:1508.03770v2 ] . xiao - dong yu , da - jian zhang , c. l. liu , d. m. tong , http://dx.doi.org/10.1103/physreva.93.060303[phys . a 93 , 060303(r ) , ( 2016 ) ] . jiajun ma , benjamin yadin , davide girolami , vlatko vedral , mile gu , http://dx.doi.org/10.1103/physrevlett.116.160407[phys . lett . 116 , 160407 ( 2016 ) ] . w. k. wootters and w. h. zurek , http://www.nature.com/nature/journal/v299/n5886/abs/299802a0.html[nature * 299 * , 802 ( 1982 ) ] . v. buzek and m. hillery , http://dx.doi.org/10.1103/physreva.54.1844[phys . rev . a * 54 * , 1844 ( 1996 ) ] . d. bruss , p.d . divincenzo , a. ekert , c. a. fuchs , c. macchiavello and j. a. smolin , http://dx.doi.org/10.1103/physreva.57.2368[phys . a * 57 * , 2368 ( 1998 ) ] . v. buzek and m. hillery , http://dx.doi.org/10.1103/physrevlett.81.5003[phys . lett . * 81 * , 5003 ( 1998 ) ] . a. chefles and s. m. barnett , http://dx.doi.org/10.1103/physreva.60.136[phys . a * 60 * , 136 ( 1999 ) ] ; h. fan , k.matsumoto and m.wadati , http://journals.aps.org/pra/abstract/10.1103/physreva.64.064301[phys a * 64 * , 064301 ( 2001 ) ] ; j. fiurasek , s. iblisdir , s. massar , and n. j. cerf , http://dx.doi.org/10.1103/physreva.65.040302[phys . a * 65 * , 040302 ( 2002 ) ] ; h. p.yuen , phys . lett . a * 113 * , 405 ( 1986 ) ; k. azuma , j. shimamura , m. koashi and n. imoto , phys . rev . a * 72 * , 032335 ( 2005 ) ; s. adhikari , n. ganguly , i. chakrabarty and b. s. choudhury , http://iopscience.iop.org / article/10.1088/1751 - 8113/41/41/415302/meta [ j. phys . a * 41 * , 415302 ( 2008 ) ] ; s adhikari , ak pati , i chakrabarty , bs choudhury , http://link.springer.com/article/10.1007/s11128-007-0053-6[quantum information processing * 6 * ( 4 ) , 197 - 219 ] . n. gisin and s. massar , http://dx.doi.org/10.1103/physrevlett.79.2153[phys . lett . * 79 * , 2153 ( 1997 ) ] . n. j. cerf , http://dx.doi.org/10.1103/physrevlett.84.4497[phys . 84 * , 4497 ( 2000 ) ] . n. j. cerf , http://www.tandfonline.com/doi/abs/10.1080/09500340008244036[j . mod . opt . * 47 * , 187 ( 2000 ) ] . v. scarani , s. iblisdir , n.gisin and a. acin , http://dx.doi.org/10.1103/revmodphys.77.1225[rev . phys . * 77 * , 1225 ( 2005 ) ] , and references therein . l. m. duan and g. c. guo , http://dx.doi.org/10.1103/physrevlett.80.4999[phys . lett . * 80 * , 4999 ( 1998 ) ] ; l.hardy and d. d. song , http://www.sciencedirect.com/science/article/pii/s037596019900448x[phys . a * 259 * , 331 ( 1999 ) ] . n. gisin , doi:10.1016/s0375 - 9601(98)00170 - 4[phys . a * 242 * , 1 ( 1998 ) ] . s. adhikari , i. chakrabarty and b. s. choudhary , http://iopscience.iop.org/article/10.1088/0305-4470/39/26/011/meta[j . a * 39 * , 8439 ( 2006 ) ] . m. k. shukla , i. chakrabarty , s . chatterjee , http://arxiv.org / pdf/1511.05796.pdf[arxiv : 1511.05796 ] . sk sazim , i. chakrabarty , a. datta and a. k. pati , http://dx.doi.org/10.1103/physreva.91.062311[phys . rev a * 91 * , 062311 ( 2015 ) ] . a. einstein , b. podolsky and n. rosen , http://dx.doi.org/10.1103/physrev.47.777[phys . , 47 777 ( 1935 ) ] . a. ekert , http://dx.doi.org/10.1103/physrevlett.67.661[phys . 67 * , 661 ( 1991 ) ] . c. h. bennett and g. brassard , http://researcher.watson.ibm.com/researcher/files/us-bennetc/bb84highest.pdf[proceedings of ieee international conference on computers , system and signal processing , bangalore , india , pp.175 - 179 ( 1984 ) ] ; p. w. shor and j. preskill , http://dx.doi.org/10.1103/physrevlett.85.441[phys . lett . * 85 * , 441 ( 2000 ) ] . m. hillery , v. buzek and a. berthiaume , http://dx.doi.org/10.1103/physreva.59.1829[phys . a * 59 * , 1829 ( 1999 ) ] ; s. k. sazim , i. chakrabarty , c. vanarasa and k. srinathan , http://link.springer.com/article/10.1007/s11128-015-1109-7[quant . inf . process * 14 * 4651 ( 2015 ) ] ; s. adhikari , i. chakrabarty and p. agrawal , http://arxiv.org/abs/1012.5570[quant . inf . comp . * 12 * , 0253 ( 2012 ) ] ; m.ray , s. chatterjee and i. chakrabarty , http://link.springer.com/article/10.1140/epjd/e2016-60683-x[eur . j. d * 70 * : 114 ( 2016 ) ] . c. h. bennett , g. brassard , c. crepeau , r. jozsa , a. peres and w. k. wootters , http://dx.doi.org/10.1103/physrevlett.70.1895[phys . lett . * 70 * , 1895 ( 1993 ) ] ; r. horodecki , m. horodecki and p. horodecki , http://www.sciencedirect.com/science/article/pii/0375960196006391[phys . a * 222 * , 21 ( 1996 ) ] . c.h.bennett and s.j.weisner , http://dx.doi.org/10.1103/physrevlett.69.2881[phys.rev.lett . * 69 * , 2881 ( 1992 ) ] ; r. nepal , r. prabhu , a. sen de and u. sen , http://dx.doi.org/10.1103/physreva.87.032336[phys . a * 87 * , 032336 ( 2013 ) ] . r. horodecki , p. horodecki , m. horodecki and k. horodecki , http://dx.doi.org/10.1103/revmodphys.81.865[rev . phys . * 81 * , 865 ( 2009 ) ] , and references therein . s. bose , v. vedral and p. l. knight , http://dx.doi.org/10.1103/physreva.57.822[phys . a * 57 * , 822 ( 1998 ) ] ; m. zukowski , a. zeilinger , m. a. horne and a. k. ekert , http://dx.doi.org/10.1103/physrevlett.71.4287[phys . rev . lett . * 71 * , 4287 ( 1993 ) ] . sk sazim and i. chakrabarty , http://link.springer.com/article/10.1140/epjd/e2013-30746-9[eur . j. d * 67 * , 174 ( 2013 ) ] . d. deutsch , a. ekert , r. jozsa , c. macchiavello , s. popescu and a. sanpera , http://dx.doi.org/10.1103/physrevlett.77.2818[phys . . lett . * 77 * , 2818 ( 1996 ) ] ; l. goldenberg and l.vaidman , http://dx.doi.org/10.1103/physrevlett.75.1239 [ phys . lett . * 75 * , 1239 ( 1995 ) ] ; a. cabello , http://dx.doi.org/10.1103/physreva.61.052312[phys . a * 61 * , 052312 ( 2000 ) ] ; c. li , h - s song and l. zhou , http://iopscience.iop.org/article/10.1088/1464-4266/5/2/306/pdf[journal of optics b : quantum semiclass . opt . * 5 * , 155 ( 2003 ) ] ; a. orieux , g. ferranti , a. darrigo , r. lo franco , g. benenti , e. paladino , g. falci , f. sciarrino and p. mataloni , http://www.nature.com/articles/srep08575[scientific reports 5 , 8575 ( 2015 ) ] ; a. darrigo , r. lo franco , g. benenti , e. paladino and g. falci , link:10.1016/j.aop.2014.07.021[ann . phys . * 350 * , 211224 ( 2014 ) ] . c. h. bennett , h. j. bernstein , s. popescu and b. schumacher , http://dx.doi.org/10.1103/physreva.53.2046[phys . a * 53 * , 2046 ( 1996 ) ] . v. buzek , v. vedral , m. b. plenio , p. l. knight and m. hillery , http://dx.doi.org/10.1103/physreva.55.3327[phys . rev . a * 55 * , 3327 ( 1997 ) ] . s. bandyopadhyay and g. kar , http://dx.doi.org/10.1103/physreva.60.3296 [ phys . a * 60 * , 3296 ( 1999 ) ] . i. chakrabarty , http://arxiv.org/abs/0803.3393[int . inf , * 7 * ( 02 ) , 559 - 565 ] . s. adhikari and b. s. choudhury , http://dx.doi.org/10.1103/physreva.74.032323[phys a * 74 * , 032323 ( 2006 ) ] ; i. chakrabarty and b. s. choudhary http://arxiv.org/abs/0804.2568[arxiv:0804.2568 ] . s. chatterjee , s. sazim , i. chakrabarty , http://dx.doi.org/10.1103/physreva.93.042309[phys . rev . a * 93 * , 042309 ( 2016 ) ] . a. kheirollahi . et.al , https://arxiv.org/abs/1604.04995[arxiv : 1604.04995 ] . h. barnum , c. m. caves , c. a. fuchs , r. jozsa and b. schumacher , http://dx.doi.org/10.1103/physrevlett.76.2818[phys . lett . * 76 * , 2818 ( 1996 ) ] . m. piani , p. horodecki and r. horodecki , http://dx.doi.org/10.1103/physrevlett.100.090502[phys . . lett . * 100 * , 090502 ( 2008 ) ] . h. barnum , j. barrett , m. leifer and a. wilce , http://dx.doi.org/10.1103/physrevlett.99.240501[phys . 99 * , 240501 ( 2007 ) ] . s. luo , http://link.springer.com/article/10.1007/s11005-010-0389-1[lett . phys . * 92 * , 143 ( 2010 ) ] . s. luo , n. li and x. cao , http://dx.doi.org/10.1103/physreva.79.054305[phys . rev . a * 79 * , 054305 ( 2009 ) ] . i. ghiu , http://dx.doi.org/10.1103/physreva.67.012323[phys . a * 67 * , 012323 ( 2003 ) ] . @xmath131{proof } \end{array}\ ] ] _ let m be a point in the interior of @xmath180 then , @xmath181 , @xmath182 and @xmath183 . _ as shown in figure 5 , @xmath184 is extended to point e outside @xmath130 intersecting @xmath185 at @xmath186 . by applying triangle inequality on @xmath187 and @xmath188 , we get @xmath189 + @xmath190 @xmath98 @xmath191 and , @xmath192 + @xmath193 @xmath98 @xmath194 respectively . + also , as @xmath195 lies on @xmath196 , we have , @xmath184 + @xmath192 = @xmath197 + now , adding these two inequalities , we have , @xmath189 + @xmath190 + @xmath193 + @xmath192 @xmath98 @xmath191 + @xmath194 + @xmath198 @xmath189 + @xmath199 + @xmath200 + @xmath192 @xmath98 @xmath201 + @xmath202 + @xmath194 + @xmath198 @xmath189 + @xmath185 + @xmath192 @xmath98 @xmath184 + @xmath192 + @xmath194 + cancelling out @xmath192 , we finally get,@xmath203 + @xmath185 @xmath98 @xmath184 + @xmath204 similarly , the other two inequalities can be proved . the cloning of arbitrary quantum states is forbidden by no - cloning theorem . however , there still exists the possibility of cloning arbitrary quantum states approximately with certain fidelity , or , imperfection. quantum cloning transformations are ( cptp ) completely positive trace preserving map between two quantum systems , supported by an ancilla state . in this section , at first , we briefly describe the buzek - hillery ( b - h ) qcm . + the b - h cloning machine ( @xmath205 ) is a @xmath206-dimensional quantum copying transformation acting on a state @xmath207 ( @xmath208 = 1 , ... , @xmath206 ) , which is to be copied onto a blank state @xmath209 . initially , the copy is prepared in state @xmath210 which subsequently gets transformed into another set of state vectors @xmath211 and @xmath212 as a result of application of the cloning machine . here @xmath213 , @xmath111 and @xmath214 represent the input , blank and machine qubits respectively . in this case , these transformed state vectors belong to the orthonormal basis set in the @xmath206-dimensional space . the transformation scheme @xmath205 is given by , @xmath215 where @xmath20 = @xmath216 , and the coefficients @xmath217 and @xmath218 are real . an optimal state independent version of the b - h cloner can be obtained from eq . by imposing the unitarity and normalization conditions which give rise to the following set of constraints , + @xmath219 , + @xmath220 and @xmath221 here , @xmath222 where @xmath223 is the number of qubits in a given quantum register . in this transformation , the fidelity of the output copies does nt depend on the input state because the shrinking or scaling property is independent of the input states . + now , by substituting @xmath217 = @xmath218 = 1 , the unitarity constraints on the b - h cloning transformation in eq . give rise to the following set of conditions : @xmath224 where @xmath9 and @xmath225 for @xmath226 ; and the output states are no longer necessarily orthonormal . + let us assume that , @xmath227 where again @xmath9 for @xmath228 ; @xmath57 and @xmath58 are the machine parameters . the optimal cloning machine described above with @xmath229 becomes a local cloning machine ( @xmath230 ) . by substituting the values of the coefficients @xmath217=@xmath231 and @xmath218=@xmath232 in eq . , we obtain the optimal state independent cloning machine which can be used for local copying purposes.in this case , the value of @xmath58 is @xmath168 . + similarly , the optimal cloning machine described above becomes a non local cloning machine ( @xmath233 ) for @xmath234 . then the corresponding values of the coefficients @xmath217 and @xmath218 in eq.10 become @xmath235 and @xmath236 respectively.now , by substituting the values of the coefficients @xmath217=@xmath235 and @xmath218=@xmath236 in eq . , we obtain the optimal state independent non local cloning machine.in this case , the value of @xmath58 is @xmath237 .

the local coulomb interaction @xmath6 of electrons occupying quantum dots leads to a variety of effects . many of them can conveniently be studied in transport through the dot within the linear regime . theoretically as well as experimentally well - investigated examples are the coulomb blockade ( cb ) peaks of the transmission ( conductance ) @xcite as well as the plateaus of width @xmath2 of the transmission ( conductance ) induced by the kondo effect @xcite . additional features of interacting multi - level dots that have recently attracted considerable theoretical attention are the population inversions of the dot levels @xcite , the phase lapses of the transmission phase @xmath1 or , equivalently , the zeros of the transmission amplitude @xmath7 ( transmission zeros ) @xcite and additional correlation - induced resonances of @xmath8 @xcite . they appear in certain parts of the parameter space when the level occupancies and the transmission amplitude are investigated as functions of the level positions , which can be tuned via a nearby plunger gate voltage . such effects were hitherto mostly studied in a minimal model involving only two levels . a very important step towards a unified understanding of population inversions , phase lapses and correlation - induced resonancess in spin - polarized two - level dots was recently taken by a multi - stage mapping of the problem on a generalized kondo model and a subsequent renormalization group and bethe ansatz analysis of the effective hamiltonian @xcite . theoretical studies of phase lapses ( transmission zeros ) are of primary interest in connection with a series of linear response transmission measurements by the weizman group @xcite on aharonov - bohm rings containing a quantum dot in one arm . under suitable conditions both the phase @xmath1 and magnitude @xmath8 of the transmission amplitude @xmath9 of the dot can be extracted from the aharonov - bohm oscillations of the current through the ring @xcite . when this is done as function of a plunger gate voltage @xmath10 that linearly shifts the dot s single - particle energy levels downward , @xmath11 ( @xmath12 is a level index ) , a series of well - separated cb peaks of rather similar width and height was observed in @xmath13 , across which @xmath14 continuously increased by @xmath0 , as expected for breit - wigner - like resonances . in each cb valley between any two successive peaks , @xmath1 always jumped sharply downward by @xmath0 . this phase lapse behavior was found to be `` universal '' , occurring in a large succession of valleys for every _ many_-electron dot studied in refs . this universality is puzzling , since naively the behavior of @xmath15 is expected to be `` mesoscopic '' , i.e. to show a phase lapse in some cb valleys and none in others , depending on the dot s shape , the parity of its orbital wavefunctions , etc . only recently @xcite , also the _ few_-electron regime was probed experimentally : as @xmath10 was increased to successively fill up the dot with electrons , starting from electron number @xmath16 , @xmath14 was observed to behave mesoscopically in the _ few_-electron regime , whereas the above - mentioned universal phase lapse behavior emerged only in the _ many_-electron regime @xmath17 . it was suggested in ref . @xcite that a generic difference between the few- and many - electron dots may be that for the latter , transport might simultaneously occur through several partially filled single - particle levels in parallel . a possible reason could be that the mean ( non - interacting ) level spacing @xmath4 of the topmost filled levels decreases as the number of electrons increases , while the charging energy @xmath2 still implies well separated transmission resonances @xcite . this scenario forms the basis of a recent systematic study by us of the interplay of level spacing , level width and charging energy on the phase lapses for up to four interacting levels and spin - polarized electrons @xcite . we showed that the universal phase lapse and transmission zero behavior appearing at small @xmath4 can be understood as resulting from a fano - type interference effect @xcite involving transport through two or more effective dot levels , whose position and width have been renormalized by the coulomb interaction and coupling to the leads . the importance of several overlapping levels for phase lapses had earlier been pointed out by silvestrov and imry in a rather specific model of a single wide and several narrow levels with strong interaction @xcite . here we supplement our earlier study @xcite by discussing the relation between phase lapses and population inversions and by investigating the role of finite temperatures @xmath18 as well as spin , focusing on @xmath19 levels . when spin is included , the kondo effect plays a role for an odd average occupation of the dot , but the phase lapse scenario is unaffected by this . experimentally the behavior of the phase in the presence of the kondo effect was investigated in refs . @xcite and @xcite . as in ref . @xcite we are concerned with the _ generic _ behavior and thus investigate the entire parameter space , going beyond subspaces of higher symmetry [ such as left - right ( l - r ) or 1 - 2 symmetry of the coupling between the left and right leads and the two levels ] . for low temperatures and sufficiently small single - particle spacings @xmath4 of the dot levels , we find @xmath0 phase lapses between two transmission peaks in an overwhelmingly large part of the parameter space of the level - lead couplings . we point out that the two level case is special compared to models with @xmath20 , as for generic level - lead couplings a transmission zero and phase lapse occurs between the two transmission peaks _ even at _ u=0 . the effect of the interaction is merely to increase the separation of the transmission peaks . for large @xmath4 the appearance or not of phase lapses between transmission peaks depends on the relative sign of the level - lead couplings in analogy to the noninteracting case @xcite . for spin - polarized dots we in addition compare our @xmath21 results with the ones of recent mean - field studies @xcite . in these works level - lead couplings beyond the subspaces with increased symmetry were studied , and a remarkably more complex behavior was found once the symmetries were broken . the importance of considering such generic parameter sets was independently pointed out in ref . we here elucidate how the phase lapse behavior is affected by quantum fluctuations , which are expected to be strong in low - dimensional systems . we find that upon including quantum fluctuations , the part of parameter space exhibiting universal @xmath0 phase lapses between well separated cb peaks becomes larger than suggested by the mean - field study . in particular , we do not recover certain peculiar features of the mean - field results of refs . @xcite and @xcite , namely the occurrence , in certain regimes of parameter space , of a phase lapse of less than @xmath0 ( instead of precisely @xmath0 ) , accompanied by the disappearance of the corresponding transmission zero @xcite . these features thus turn out to be artifacts of the mean - field approximation , which misses the rather simple scenario for the phase lapse behavior of a two - level dot at small @xmath4 : for generic level - lead couplings a phase lapse and transmission zero between two transmission peaks is already present at @xmath5 ; increasing the coulomb interaction the peaks become well separated while the phase lapse and transmission zero remain in the valley between them . in the model of a single wide and several narrow levels @xcite a relation between phase lapses and population inversions was discussed . therefore , in phase lapse studies quite often also the level occupancies @xmath22 , @xmath23 , are investigated . we emphasize that the generic appearance of a phase lapse and transmission zero even at @xmath5 renders the two - level model unsuitable for establishing a general relation between phase lapses and population inversions , as the latter only appear at sufficiently large @xmath2 . furthermore , we show that discontinuities of the @xmath22 as a function of @xmath10 are an artifact of the mean - field solution ( see refs . @xcite and @xcite ) . within our approaches discontinuities are only found for l - r symmetric level - lead couplings with a relative plus sign of the underlying hopping matrix elements and degenerate levels , a case which was earlier identified as being nongeneric @xcite , because the transmission shows only a single peak . this paper is organized as follows . in sec . [ model ] we introduce our model for the spin - polarized and spin - degenerate two - level dot . we discuss the relation between the measured magnetic flux dependence of the interferometer s linear conductance and the magnitude and phase of the dot s transmission amplitude . the latter can be computed from the one - particle green function of the dot . we present a brief account of our techniques to obtain the latter , the numerical ( nrg ) @xcite and functional renormalization group ( frg ) methods . for an introduction to the use of the frg to quantum dots see refs . @xcite and @xcite . in contrast to the implementation of the nrg used in ref . @xcite , where we were restricted to l - r symmetric dots at @xmath21 , we have now adopted the full density matrix ( fdm ) nrg method of @xcite , which enables us to investigate dots with arbitrary level - lead state overlap matrix elements @xmath24 ( with @xmath25 ) as well as to study finite temperatures . in secs . [ resultsnon]-[resultsspin ] we present our results of the @xmath10 dependence of @xmath8 and @xmath1 . first we briefly discuss the noninteracting two - level dot with generic level - lead couplings and point out that the phase lapse scenario differs from the one for more than two levels . we then investigate interacting , spin - polarized dots , study the relation between phase lapses and population inversions and compare to the mean - field results for the phase lapses . the issue of continuous versus discontinuous @xmath10 dependence of the level occupancies @xmath26 and @xmath27 is commented on . next we study the role of finite temperatures . finally , we consider spin - degenerate levels at small @xmath3 which implies the appearance of kondo physics at odd average dot filling . using nrg and frg we show that the spin does not alter the universal phase lapse scenario . our findings are summarized in sec . [ summary ] . in this section we introduce our model for the two - level dot . we argue that it is the energy dependent ( effective ) transmission amplitude @xmath28 which one has to compute if one is interested in comparing to the measurements of refs . @xcite of the magnitude of the transmission amplitude and its phase . the amplitude @xmath29 can be determined from the matrix elements of the dot s interacting one - particle green function . we furthermore discuss aspects of the nrg and the frg specific to our problem . our hamiltonian consists of three parts @xmath30 the two semi - infinite leads are modeled as noninteracting one - dimensional tight - binding chains and for simplicity are assumed to be equal @xmath31 the hopping strength in the leads is @xmath32 . we use standard second quantized notation with @xmath25 indicating the left and right leads , where the quantum numbers @xmath33 and @xmath34 label wannier states and spin , respectively . the dot is described by @xmath35 where the term with @xmath36 and @xmath37 is excluded from the sum in the interacting part . we define @xmath38 . in experimental systems the inter- and intra - level coulomb repulsion can be expected to be comparable in size and to avoid a proliferation of parameters we assumed them to be equal . this assumption is not essential ; by relaxing it we have checked that our results are robust against inter - level variations of the interaction strengths . finally , the coupling between dot and lead states is given by @xmath39 with real overlap matrix elements @xmath24 . for later purposes we define @xmath40 and @xmath41 . for simplicity , part of our studies will be performed on a model of spinless electrons , for which the spin index will be dropped . the resulting model may be regarded as a spin - polarized version of the spinful model obtained if the latter is put in a very large magnetic field . the experimental two - path interferometer is characterized by wide base regions and very narrow point contacts towards the emitter and collector . electrons can thus only pass once through either the upper or lower arm ( one of them containing the dot ) before reaching the collector . it is furthermore reasonable to assume that the voltage drops at the point contacts and the dot remains in equilibrium @xcite . under these conditions the magnetic flux @xmath42 dependent part of the conductance of the aharonov - bohm interferometer @xmath43 is given by @xcite @xmath44 } \end{aligned}\ ] ] with the flux quantum @xmath45 , the derivative @xmath46 of the fermi function as well as the transmission amplitudes @xmath47 of the arm containing the dot and @xmath48 of the reference arm . the transmission @xmath49 is the product of the transmission @xmath50 through the dot and the transmission @xmath51 through the rest of the interferometer arm containing the dot . it is reasonable to assume that @xmath51 as well as @xmath52 are only weakly energy and gate voltage dependent and thus the @xmath10-dependence of the aharonov - bohm oscillations of the measured linear conductance of the interferometer is dominated by the @xmath10 dependence of the magnitude and phase of the transmission amplitude through the dot . as usual @xcite , we compute the energy - averaged transmission phase @xmath1 and magnitude @xmath8 of the dot for a fixed spin direction as the phase and absolute value of @xmath53 where the frequency integral represents an energy average , weighted by the derivative of the fermi function , as appropriate for a finite electron temperature in the leads . in the limit @xmath54 , @xmath55 reduces to a @xmath4-function and @xmath56 is equal to @xmath57 . we here take the chemical potential @xmath58 . using scattering theory @xmath28 ( for fixed spin direction ) can be related to the spin - independent matrix elements ( in the @xmath23 indices of the wannier states ) of the dot s one - particle retarded green function @xmath59 @xmath60 with ( after taking the wide band limit ; see below ) @xmath61 where @xmath62 denotes the local density of states at the end of each semi - infinite lead . without loss of generality we have assumed that @xmath63 , @xmath64 and @xmath65 . the spin - independent dot occupancies @xmath22 ( per spin direction ) , that we will also investigate , follow from the green function @xmath66 by integrating over frequency ( or can be computed directly when using nrg ) . for l - r symmetry of the level - lead couplings and at temperature @xmath21 , @xmath67 can also be expressed in terms of the spin independent occupancies @xcite @xmath68\pi ) } e^{i(n_e+n_o)/\pi } \ ; , \label{t_lr}\ ] ] where @xmath69 for @xmath70 and @xmath71 for @xmath72 , respectively . here we will compute @xmath59 in two ways , using both a truncated , that is approximate , frg scheme , and a numerically exact method , the nrg . the truncated frg is an approximation scheme to obtain the self - energy @xmath73 ( and thus the one - particle green function ) and higher order vertex functions for many - body problems @xcite . as a first step in the application of this approach to quantum dots one integrates out the noninteracting leads within the functional integral representation of our many - body problem @xcite . the leads provide a frequency dependent one - particle potential on the dot levels . on the imaginary frequency axis it is given by @xmath74 where @xmath75 denotes the spin - independent green function of the isolated semi - infinite leads taken at the last lattice site @xmath76 as we are not interested in band effects we take the wide band limit . the potential then reduces to @xmath77 after this step , instead of dealing with an infinite system we only have to consider the dot of two interacting levels . in the computation of the interacting one - particle green function projected onto the dot system the sum of the dot hamiltonian with @xmath78 and @xmath79 can be interpreted as a frequency dependent `` single - particle hamiltonian '' and in the following will be denoted by @xmath80 . for the spin - polarized case it is a @xmath81 matrix in the quantum number @xmath23 . including spin , because of the additional quantum number @xmath82 , @xmath83 is a @xmath84 matrix which is block - diagonal in @xmath34 ( spin conservation ) . as we are here not interested in the role of a magnetic field lifting the spin - degeneracy of each level the @xmath85 and @xmath86 blocks are equivalent . the resolvent @xmath87^{-1}$ ] obtained from @xmath88 is equivalent to the noninteracting propagator of our two - level many - body problem projected on the dot levels . in the generating functional of the one - particle irreducible vertex functions we replace @xmath89 by @xmath90^{-1}\ ] ] with @xmath91 being an infrared cutoff running from @xmath92 down to @xmath93 . taking the derivative with respect to @xmath91 one can derive an exact , infinite hierarchy of coupled differential equations for vertex functions , such as the self - energy and the one - particle irreducible two - particle interaction . in particular , the flow of the self - energy @xmath94 ( one - particle vertex ) is determined by @xmath94 and the two - particle vertex @xmath95 , while the flow of @xmath95 is determined by @xmath94 , @xmath95 , and the flowing three - particle vertex . the latter could be computed from a flow equation involving the four - particle vertex , and so on . at the end of the frg flow @xmath96 is the self - energy @xmath73 of the original , cutoff - free problem we are interested in @xcite from which the green function @xmath59 can be computed using the dyson equation . a detailed derivation of the frg flow equations for a general quantum many - body problem that only requires a basic knowledge of the functional integral approach to many - particle physics @xcite and the application of the method for a simple toy problem is presented in ref . @xcite . for an overview of the application to quantum dots see refs . @xcite and @xcite . we here truncate the infinite hierarchy of flow equations by only keeping the self - energy and the frequency - independent part of the two - particle vertex . higher order terms can be neglected if the bare two - particle interaction is not too large . by comparison to nrg data this approximation scheme was earlier shown to provide excellent results for a variety of dot systems @xcite . for further comparison see fig . [ fig7 ] below . the present scheme leads to a frequency - independent self - energy ( see below ) . as finite frequency effects ( inelastic processes ) become important at temperatures @xmath18 , but these are not accurately treated by the level of approximation used here , in the present paper we shall show frg results only for @xmath21 . the truncation leads to the coupled differential flow equations @xmath97 and @xmath98 \big\ } \label{finalflowgamma}\end{aligned}\ ] ] where @xmath99 , @xmath100 , etc . are multi - indices representing the quantum numbers @xmath101 and @xmath102^{-1 } \ ; .\ ] ] in the model with spin - degenerate levels each index @xmath99 , @xmath100 etc . can take four different values @xmath23 and @xmath103 which gives 16 equations for @xmath94 and 256 for the two - particle vertex . for a spin - polarized two - level dot the multi - indices take two values and one obtains 4 equations for @xmath94 and 16 for the two - particle vertex . the number of independent equations can be significantly reduced ( see below ) taking into account the antisymmetry of the two - particle vertex and the spin symmetry ( for spin - degenerate levels ) both being preserved by eqs . ( [ finalflowsigma ] ) and ( [ finalflowgamma ] ) . the initial conditions at @xmath104 are given by @xmath105 while @xmath106 is given by the bare antisymmetrized two - body interaction . in the spin - polarized case the only nonzero components of the two - particle vertex at @xmath107 are @xmath108 in the model including spin the initial conditions take the form @xmath109 all other components which do not arise out of these by permutations ( @xmath110 and @xmath111 ) are zero . the self - energy matrix and thus the one - particle green function is completely independent of the spin direction and in the following we suppress the spin indices . as already mentioned the present approximation leads to a frequency - independent self - energy . this allows for a simple single - particle interpretation of its matrix elements . the sum of the @xmath112 and the bare level position correspond to the flowing effective level positions , @xmath113 , while @xmath114 is a hopping between the levels 1 and 2 generated in the frg flow . the frg formalism then reduces to a set of coupled differential flow equations for @xmath115 , @xmath116 and a few ( one in the spin - polarized case and seven for spin - degenerate levels ) independent components of the two - particle vertex . these flow equations can easily be integrated numerically using standard routines . it is important to note that although we start out with intra- and inter - level coulomb interactions of equal strengths they generically become different during the frg flow ( because of the different @xmath117 ) . furthermore , additional interaction terms which are initially zero will be generated in the flow . the set of equations significantly simplifies if the flow of the vertex is neglected while the results remain qualitatively the same . within this additional approximation and for a spin - polarized dot the flow equations for @xmath115 and @xmath116 are explicitly given in ref . @xcite . in certain limiting cases it is even possible to analytically solve the differential equations @xcite . however , in the present work , the flow of the vertex is retained which clearly improves the quality of approximation @xcite . at the end of the frg flow , the full green function takes the form @xmath118^{-1}_{j , j ' } = i \omega \delta_{j , j ' } - h_{j , j'}(i\omega)$ ] with an effective , _ noninteracting _ ( but @xmath10- , @xmath2- and @xmath119-dependent ) `` hamiltonian '' @xmath120 in a last step we have to perform the analytic continuation to the real frequency axis @xmath121 . this is straightforward , as the only frequency dependence of @xmath122 is the trivial one of the lead contribution eq . ( [ leadpotdefwideband ] ) . then @xmath28 can be computed using eq . ( [ transformula ] ) . the numerical renormalization group was invented by k. g. wilson in 1974 as a nonperturbative renormalization scheme for the kondo model @xcite . it was later extended to the fermionic @xcite anderson model which describes a localized electronic state coupled to a fermionic bath . the nrg allows thermodynamic and dynamic properties of such strongly correlated systems to be calculated at zero and finite temperature @xcite . the key idea of nrg is to discretize the conduction band of the bath logarithmically , leading to a tight - binding chain for which the hopping matrix elements between the successive sites fall off exponentially with @xmath123 , where @xmath124 is the discretization parameter , typically @xmath125 , and @xmath126 is the site index . this energy scale separation ensures that the problem can be solved iteratively by adding one site at a time and diagonalizing the enlarged system at each step , thereby resolving successively smaller and smaller energy scales . thus , by choosing the length @xmath127 of the chain so large that the corresponding energy scale @xmath128 is smaller than all other energies in the problem , all relevant energy scales can be resolved and treated properly . since the dimension of the hilbert space of the chain increases exponentially with the length of the chain , a truncation scheme has to be adopted , according to which only the lowest @xmath129 eigenstates of the chain are retained at each iteration . recently , it was shown that by also keeping track of discarded states a complete , but approximate , basis of states can be constructed @xcite . this can be used to calculate spectral functions which rigorously satisfy relevant sum rules @xcite . in order to obtain the transmission through the dot @xmath130 eq . ( [ transformula ] ) we follow @xcite and @xcite to compute the imaginary part of the local green functions at temperature @xmath3 , using the lehmann representation @xmath131 ) \nonumber \\ * & -\pi & \frac{e^{-\omega_n / t}}{z}\ , \langle n| d_{j\prime,\sigma}^\dag |m\rangle \ , \langle m| d_{j,\sigma } |n\rangle \ , \mathcal{\delta}(\omega+[\omega_m-\omega_n]),\end{aligned}\ ] ] with @xmath132 , the many - body eigenstates @xmath133 and eigenenergies @xmath134 . since these are causal functions , the real part can be accessed by performing a kramers - kronig transformation @xcite . in the next three sections we present our results . in sec . [ resultsnon ] for the @xmath5 case . in sec . [ resultsgeneric ] we present the generic phase lapse scenario for interacting spin - polarized dots , compare to the mean - field results and investigate the role of finite temperatures . finally , in sec . [ resultsspin ] we study the spinful two - level dot . the large number of parameters makes it essential to analyze the transmission for the noninteracting case before considering the effect of two - particle interactions . we focus on @xmath21 . a closed expression for @xmath13 and @xmath135 ( for a fixed spin direction ) at @xmath5 can be obtained from eqs . ( [ transamp ] ) and ( [ transformula ] ) by replacing @xmath136 by @xmath137 , @xmath138^{1/2}}{\left [ \left ( \gamma_1^l \gamma_2^r + \gamma_2^l \gamma_1^r - 2 s \sqrt{\gamma_1^l \gamma_1^r \gamma_2^l \gamma_2^r}- \varepsilon_{1 } \varepsilon_{2}\right)^2 + \left ( \varepsilon_{1 } \gamma_2 + \varepsilon_{2 } \gamma_1 \right)^2\right]^{1/2 } } \ ; , \\ \label{u0phase } \mbox { } \hspace{-1.5 cm } \theta(v_g ) & = & \arctan { \left [ \frac{\varepsilon_1\gamma_2+\varepsilon_2\gamma_1}{\varepsilon_1\varepsilon _ 2-\left(\sqrt{\gamma_1^l\gamma_2^r}-s\sqrt{\gamma_1^r\gamma_2^l}\right)^2 } \right ] } \ , \mbox{mod } \ , \pi \ ; , \end{aligned}\ ] ] with @xmath139 . for later use we define @xmath140 for a fixed set of @xmath117 the @xmath4 dependence of @xmath13 and @xmath135 is shown in the first columns of figs.[fig1 ] and [ fig2 ] for @xmath141 . the results are qualitatively the same as those obtained for @xmath5 . for generic level - lead couplings @xmath117 the gate voltage dependence of eq . ( [ u0conduct ] ) in the limit of small and large @xmath142 is dominated by two peaks ( of height @xmath143 ) and a transmission zero . associated with the transmission zero is a @xmath0 phase lapse at the same gate voltage . the transmission zero ( and phase lapse ) follows from perfect destructive interference at a particular @xmath10 and in the limit of a strong asymmetry in the coupling of the two levels to the leads , @xmath144 or vice versa , can be understood as resulting from a fano anti - resonance @xcite . apparently the fano anti - resonance with vanishing transmission is robust if one goes away from this limit towards more symmetric level - lead couplings . across each of the transmission resonances @xmath1 increases roughly by @xmath0 as expected for a breit - wigner resonance . further details of @xmath13 and @xmath135 depend on @xmath145 . for @xmath70 the transmission zero ( and phase lapse ) is located between the two conductance peaks for all @xmath4 . for @xmath146 the resonance peak positions depend on the asymmetry of the @xmath117 and the separation of the peaks is small if the @xmath117 are close to l - r symmetry , that is close to @xmath147 . for l - r symmetric dots and @xmath148 the transmission zero ( and phase lapse ) disappears ( not shown in the figures ) . this is an example of a submanifold in parameter space with nongeneric behavior . a complete account of such cases ( which also remain nongeneric for @xmath6 ) is given in refs . @xcite and @xcite . as they require fine tuning these parameter sets are presumably irrelevant in connection with the experiments and we will here only briefly mention results obtained in such cases . for @xmath72 and fixed @xmath117 the position of the transmission zeros and phase lapses with respect to the cb peaks is different for small or large @xmath142 ( see fig . [ fig2 ] ) . at small @xmath142 it is located between the two conductance peaks , whereas for large @xmath142 it lies on one of the outer sides of these peaks @xcite . in the crossover regime between these limiting cases the height of one of the peaks decreases , while the other becomes broader and splits up into two resonances separated by a minimum with non - vanishing conductance [ see fig . [ fig2 ] ( g ) and ( j ) ] . the crossover scale @xmath149 depends on the choice of @xmath117 . for large @xmath142 , @xmath8 has three local maxima , although the height of one of the maxima is significantly smaller than the height of the other two ( not shown in fig . [ fig2 ] ) . for fixed , asymmetric @xmath117 and @xmath146 the separation of the two conductance peaks for @xmath72 is significantly larger than for @xmath70 [ compare figs . [ fig1 ] ( a ) and [ fig2 ] ( a ) ] . it is important to note that for small @xmath142 essential features of the universal phase lapse regime established in the experiments are already found at @xmath5 : regardless of the sign @xmath145 for generic @xmath117 ( that is with the exception of a few cases with increased symmetry ) two transmission resonances are separated by a transmission zero and @xmath0 phase lapse . at @xmath5 the peak separation is too small and the shape of the @xmath10 dependence of the transmission and phase close to the peaks is qualitatively different from those observed experimentally ( namely lorentzian - like for the magnitude of the transmission , s - shaped for the phase ) . as we show next the latter problems do not arise for sufficiently large interaction @xmath2 , which in particular leads to an increased separation @xmath150 of the transmission peaks . in ref . @xcite it was shown that frg and nrg results for @xmath13 and @xmath135 agree quantitatively up to fairly large @xmath2 . for a generic set of couplings @xmath151 we present frg data for @xmath135 and @xmath13 together with the occupancies of the levels @xmath22 for different @xmath2 and @xmath4 in figs . [ fig1 ] ( @xmath70 ) and [ fig2 ] ( @xmath72 ) . increasing @xmath152 the separation of the transmission peaks in the limit of small and large @xmath142 increases and is eventually given by @xmath150 . even though this charging effect appears to be straightforward it is important to note that in particular the groundstate at small @xmath142 is highly correlated . this becomes explicit from the mapping of the present problem on a generalized single impurity anderson and kondo model as discussed in ref . an indication of strong correlation effects are the correlation - induced resonances of the transmission found in ref . @xcite , which we briefly mention below . with increasing @xmath2 , even at small @xmath142 the gate voltage dependence of @xmath135 across the transmission resonances becomes s - shaped and the resonances more lorentzian - like [ see third columns of figs . [ fig1 ] and [ fig2 ] ] . obviously , for @xmath70 the transmission zero and phase lapse remain between the two transmission peaks for all @xmath4 and @xmath2 ( see fig . [ fig1 ] ) . for @xmath72 this only holds for sufficiently small level spacings as , similar to the @xmath5 case , with increasing @xmath142 a crossover sets in to a regime in which the transmission zero and phase lapse are no longer between the peaks . analogously to the @xmath5 case , the crossover scale @xmath149 depends on the particular choice of @xmath117 . as can be seen from the second row of fig . [ fig2 ] ( the cb peaks at large @xmath2 have still almost equal height ) , with increasing @xmath152 , @xmath149 is pushed towards larger values . the coulomb interaction thus stabilizes the parameter regime of universal phase lapses . this shows that the effect of the coulomb interaction leading to universal @xmath0 phase lapses between separated cb peaks in a two - level dot is rather straightforward : for small @xmath142 the phase lapse and transmission zero are already present at @xmath5 , and the effect of finite @xmath2 is simply that the cb peaks become well separated because of charging effects . they also lead to a lorentzian - like lineshape of the peaks and an s - like variation of @xmath1 across them.the present scenario has to be contrasted to the one obtained for @xmath20 levels discussed in ref . @xcite . the generic appearance of @xmath153 transmission zeros and phase lapses separating the transmission peaks at small @xmath142 and @xmath5 is specific to the case with @xmath19 levels . for @xmath20 the number of transmission zeros and phase lapses at @xmath5 strongly depends on the parameters and the mechanism leading to universal @xmath0 phase lapses at sufficiently large @xmath2 ( at small @xmath142 ) is much more involved @xcite . this shows that although important insights can be gained from studying two - levels to achieve a _ complete _ understanding of the phase lapse scenario it is essential to study dots with coulomb interaction and more than two levels @xcite . and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=35.2%,height=141 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=29.2%,height=141 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=33.0%,height=141 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=35.2%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=29.2%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=33.0%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=35.2%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=29.2%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=33.3%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=35.2%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=29.2%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=33.0%,height=117 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=35.2%,height=137 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=29.2%,height=137 ] and @xmath142 that govern the gate voltage @xmath10 dependence of the magnitude of the transmission @xmath8 ( black ) , the transmission phase @xmath1 ( red ) and the level occupancies ( green and blue ) of a spin - polarized two - level dot at @xmath21 . the parameters are @xmath154 and @xmath70 . for better visibility @xmath155 were shifted by @xmath156 . the depicted behavior is the generic one and in particular qualitatively independent of the actual choice of @xmath151 ( up to certain cases of increased symmetry ; for examples see the text ) . the behavior at @xmath141 is qualitatively the same as the one at @xmath5 . the results were obtained using the truncated frg.,title="fig:",scaledwidth=33.0%,height=137 ] , but for @xmath72.,title="fig:",scaledwidth=35.2%,height=141 ] , but for @xmath72.,title="fig:",scaledwidth=29.2%,height=141 ] , but for @xmath72.,title="fig:",scaledwidth=33.0%,height=141 ] , but for @xmath72.,title="fig:",scaledwidth=35.2%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=29.2%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=33.0%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=35.2%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=29.2%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=33.3%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=35.2%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=29.2%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=33.0%,height=117 ] , but for @xmath72.,title="fig:",scaledwidth=35.2%,height=137 ] , but for @xmath72.,title="fig:",scaledwidth=29.2%,height=137 ] , but for @xmath72.,title="fig:",scaledwidth=33.0%,height=137 ] the lineshape of @xmath8 shows characteristic differences in the limits of small and large @xmath142 . in the universal regime at small @xmath142 and for sufficiently large @xmath152 the two cb peaks have equal width of order @xmath157 ( not @xmath158 ) and equal height which is consistent with the expectation that at each peak the transport occurs through both bare levels simultaneously . a similar behavior is observed in the experiments . in the mesoscopic regime ( @xmath159 ) the width of the @xmath160th peak is given by @xmath158 and the relative height @xmath161 by @xmath162 , independent of the value of @xmath2 . this effectively noninteracting lineshape can be understood from the gate voltage dependence of the effective level positions @xmath163 at the end of the frg flow . when one level is charged the effective level position of the other level is pushed downwards by @xmath2 . besides this , the gate voltage dependence of the @xmath163 remains linear , leading to two transmission peaks at gate voltages @xmath164 with separation @xmath150 , but with the same width and height as for @xmath5 . the hopping between the two effective levels generated in the frg flow is small and can be neglected . apart from the cb peaks , for sufficiently large @xmath152 the transmission shows additional features at small @xmath165 [ see figs . [ fig1 ] ( b ) , ( c ) , ( e ) and [ fig2 ] ( c ) ] . these are the correlation - induced resonances mentioned in the introduction , which have been found to be most pronounced at @xmath148 and in this case occur for interactions larger than a critical @xmath166 which depends on the @xmath117 and @xmath145 @xcite . their appearance indicates that the groundstate at small @xmath142 is strongly correlated ( as mentioned in ref . @xcite the correlation - induced resonances are not captured by a mean - field analysis ; see below ) . associated with the correlation - induced resonances is a sharp increase of @xmath1 [ see figs . [ fig1 ] ( b ) and ( e ) ] . at large @xmath152 the correlation - induced resonances are exponentially ( in @xmath152 ) sharp features that vanish quickly with increasing @xmath3 ( see below ) , which might be one of the reasons why up to now they have not been observed in experiments . the correlation - induced resonances are not directly linked to the universal phase lapse scenario . for increasing @xmath152 at fixed @xmath142 and decreasing @xmath142 at fixed @xmath152 we observe an increased tendency towards population inversion of the @xmath22 . we define that a population inversion occurs if ( 1 ) @xmath167 at a certain @xmath168 and ( 2 ) one @xmath22 has positive and one negative slope at @xmath168 so that the filling of one level causes a tendency for the other to empty . for large @xmath152 and small @xmath142 it is mainly the more strongly coupled level ( in figs . [ fig1 ] and [ fig2 ] , this is the level 2 shown in blue ) whose population increases across both cb peaks while it is depopulated in between . this behavior is reminiscent of the one discussed in the model with a broad and several narrow levels @xcite , where a relation between population inversion and phase lapse behavior was proposed . remarkably , for sufficiently large @xmath152 , we find population inversion even for small asymmetries @xmath169 ( which is only 1.5 in the example of figs.[fig1 ] and [ fig2 ] ) . we emphasize that despite this resemblance to the observation of ref . @xcite , the @xmath19 model is not appropriate to establish a general relation between the appearance of population inversions and @xmath0 phase lapses at small @xmath142 @xcite . while the latter are already present at @xmath5 , the former only develop with increasing @xmath2 [ compare figs . [ fig1 ] ( a ) , ( d ) or [ fig2 ] ( a ) , ( d ) to figs . [ fig1 ] ( c ) , ( f ) or [ fig2 ] ( c),(f ) ] . note that the gate voltage @xmath168 at which the population inversion occurs is generically _ not _ identical to the position of the phase lapse and transmission zero [ see figs . [ fig1 ] ( b ) , ( c ) , ( f ) and [ fig2 ] ( c ) , ( f ) ] @xcite . however , for l - r symmetric @xmath117 eq . ( [ t_lr ] ) ensures that if a population inversion occurs its position is identical to the one of the phase lapse and transmission zero . as can be seen in figs . [ fig1 ] ( c ) , ( f ) and [ fig2 ] ( c ) , ( f ) for small @xmath142 and large @xmath152 the @xmath22 show a rather strong gate voltage dependence between the cb peaks . nevertheless the total dot occupancy @xmath170 is only weakly @xmath10 dependent and close to 1 within the entire cb valley . this is reminiscent of the plateau - like occupancy in the local moment regime of the single impurity anderson model showing the kondo effect . as discussed in ref . @xcite a relation to this model can indeed be established . we note in passing that with the exception of the nongeneric case of l - r symmetric @xmath117 , @xmath70 and @xmath148 , the @xmath22 are continuous functions of @xmath10 . in refs . @xcite and @xcite golosov and gefen ( gg ) analyze the phase lapse scenario of the spin - polarized interacting two - level dot within the mean - field approximation . however , they anticipated themselves that quantum fluctuations not captured in the mean - field approach could be important . examples of this had been pointed out already in refs . @xcite and @xcite . thus , gg emphasized that the effects of such fluctuations on their results need to be studied in subsequent work . the present subsection is devoted to this task . and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=45.5%,height=143 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=43.5%,height=143 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=45.5%,height=173 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=43.5%,height=173 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=45.5%,height=143 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=43.5%,height=143 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=45.5%,height=173 ] and different @xmath171 moving from ggs `` phase '' 2 into `` phase '' 1 . figures ( e)-(h ) are for constant @xmath171 and different @xmath172 moving from ggs `` phase 3 '' into `` phase 1 '' . the inset in fig . ( e ) shows a zoom in of the gate voltage region around the phase lapse . for a detailed comparison to the mean - field results see the text . the results were obtained at @xmath21 using the truncated frg.,title="fig:",scaledwidth=43.5%,height=173 ] gg consider the subspace of level - lead couplings defined by @xmath173 . performing a unitary transformation on the dot states the part of the hamiltonian eq . ( [ generalham ] ) containing dot operators can be transformed to ( see refs . @xcite and @xcite ) @xmath174 \;,\end{aligned}\ ] ] with the transformed operators and parameters indicated by a hat . the change of basis leads to a direct hopping @xmath175 between the transformed levels . we here focus on the relative sign @xmath176 . gg then introduce the two new dimensionless parameters @xmath172 and @xmath171 as @xmath177 and @xmath178 . varying @xmath172 and @xmath171 gg investigate the phase lapse behavior for fixed level spacing ( in the new basis ) @xmath179 and interaction @xmath180 . as the parameters in the new basis are rather complicated combinations of the original ones , the variation of @xmath171 and @xmath172 corresponds to the variation of the @xmath117 ( within the above specified subspace ) , @xmath4 and even @xmath145 . our simple picture that increasing the level spacing @xmath4 of the untransformed model leads from the universal to the mesoscopic regime can not easily made explicit using the parameters of gg . to make the comparison of our results to the mean - field study definite we nevertheless follow the steps of gg . varying @xmath181 at @xmath176 we move around in the right part of what is called the `` phase diagram '' by gg ( fig . 4 of ref . @xcite ) . in fig . [ fig3 ] we show the behavior of @xmath8 and @xmath1 varying @xmath171 at fixed @xmath172 [ figs . [ fig3 ] ( a)-(d ) ] and @xmath172 at fixed @xmath171 [ figs.[fig3 ] ( e)-(h ) ] , respectively . in the first case , upon increasing @xmath171 at constant @xmath182 , we move from ggs `` phase 2 '' , ( red in fig . 4 of ref . @xcite ) , with the transmission zero and phase lapse outside the two cb peaks , into `` phase 1 '' ( blue in fig . 4 of ref . @xcite ) , with the transmission zero and phase lapse between the peaks . the mean - field approximation correctly captures the presence of these two regimes . we find a smooth crossover between them ( which is why we prefer the notion of different `` regimes '' rather than `` phases '' ) : the @xmath10 value at which the transmission zero and phase lapse occur _ smoothly _ crosses from lying outside the right cb peak to lying between the two cb peaks . this is similar to the smooth crossover we observe in the @xmath72 case of the untransformed model when moving from the mesoscopic to the universal regime ( see columns of fig . [ fig2 ] ) . for fixed @xmath183 and increasing @xmath172 we move from `` phase 3 '' ( green in fig . 4 of ref . @xcite ) into `` phase 1 '' . in contrast to the mean - field approximation where an abrupt transition from `` phase 3 '' to `` phase 1 '' occurs , figs . [ fig3 ] ( e)-(h ) show a rather smooth evolution . the mean - field `` phase 3 '' is characterized by discontinuous population switching and a phase lapse between the cb peaks which , surprisingly , is smaller than @xmath0 @xcite . furthermore , in this parameter regime the mean - field results show no transmission zero , as discussed in ref . however , from figs . [ fig3 ] ( e)-(h ) it is apparent that `` phase 3 '' , is an artifact of the mean - field approximation : upon taking fluctuations into account via frg , the @xmath0 phase lapse and the transmission zero are found to remain in the cb valley . the evolution with increasing @xmath172 is similar to the @xmath70 case of the untransformed model when the level spacing is increased at fixed @xmath117 ( see the second and third columns of fig . [ fig1 ] ) . `` phase 3 '' then corresponds to the parameter regime with small level spacing and a sizable @xmath152 in which correlations are of particular importance leading e.g. to the correlation - induced resonances [ see figs . [ fig3 ] ( e)-(h ) ] . that the mean - field approximation fails to properly describe this strongly correlated regime is not surprising and has been recognized earlier @xcite . in ref . @xcite a connection between the small @xmath4 regime of the present model and the local moment ( kondo ) regime of the single - impurity anderson model was established . thus , the artifacts of the mean - field approximation for the spinless two - level dot are reincarnations of the the well - known artifacts it produces when applied to the anderson model in the local moment regime . upon taking fluctuations into account , the discontinuities of the @xmath155 in `` phase 3 '' are washed out . the only choice of parameters for which we find discontinuous behavior is the one with l - r symmetric @xmath117 , @xmath70 and @xmath148 , a case which was already identified as being nongeneric @xcite . any arbitrarily small deviation from these conditions leads to a continuous gate voltage dependence of @xmath155 . for parameters close to the nongeneric point the change in @xmath22 at first sight appears to be rather sharp and an extremely high resolution in @xmath10 is required to identify the behavior as continuous . the fact that the mean - field treatment at @xmath6 and small level spacings incorrectly causes the transmission zero to disappear and the corresponding phase lapse to become smaller than @xmath0 is its most consequential problem . by generating such features , the mean - field treatment masks the very simple scenario that emerges upon properly including fluctuations : for increasing @xmath2 , the @xmath0 phase lapse and transmission zero found for small @xmath142 at @xmath5 remain in the cb valley , while the transmission peaks become well separated , lorentzian - like and the phase acquires an s - shape across the resonances . we next investigate how the phase and magnitude of the transmission are affected by finite temperatures . to this end , we use the new fdm - nrg algorithm recently proposed in ref . @xcite . in fig . [ fig4 ] we show nrg data for @xmath13 and @xmath135 at different @xmath3 . we consider generic level - lead couplings @xmath184 , @xmath185 , @xmath186 and @xmath187 ( universal regime ) as well as @xmath188 ( mesoscopic regime ) . as expected , with increasing temperature the sharp @xmath0 phase lapses are gradually smeared out , the transmission zero vanishes and the change of @xmath1 at the phase lapse becomes smaller than @xmath0 . furthermore , the cb peaks decrease and broaden . for @xmath188 and @xmath72 [ fig . [ fig4 ] ( d ) ] the phase lapse lies outside the cb peaks and outside the window of gate voltages shown . and @xmath135 obtained by nrg for @xmath189 , @xmath185 for @xmath186 and in the universal ( @xmath187 ) and the mesoscopic ( @xmath188 ) regime . we used the nrg parameters @xmath190 and @xmath191.,title="fig:",scaledwidth=49.5%,height=166 ] and @xmath135 obtained by nrg for @xmath189 , @xmath185 for @xmath186 and in the universal ( @xmath187 ) and the mesoscopic ( @xmath188 ) regime . we used the nrg parameters @xmath190 and @xmath191.,title="fig:",scaledwidth=47.5%,height=166 ] and @xmath135 obtained by nrg for @xmath189 , @xmath185 for @xmath186 and in the universal ( @xmath187 ) and the mesoscopic ( @xmath188 ) regime . we used the nrg parameters @xmath190 and @xmath191.,title="fig:",scaledwidth=49.5%,height=196 ] and @xmath135 obtained by nrg for @xmath189 , @xmath185 for @xmath186 and in the universal ( @xmath187 ) and the mesoscopic ( @xmath188 ) regime . we used the nrg parameters @xmath190 and @xmath191.,title="fig:",scaledwidth=47.5%,height=196 ] in the mesoscopic regime [ figs . [ fig4 ] ( b ) and ( d ) ] the explicit temperature dependence of the green function entering eq . ( [ transamp ] ) via eq.([transformula ] ) is rather weak and the temperature dependence of the cb peaks and the phase lapse can be understood from the behavior in the noninteracting model , but with level spacing @xmath150 . for small @xmath3 the height @xmath192 of the @xmath160th cb peak scales as @xmath193 and the width @xmath194 of the phase lapse as @xmath195 @xcite . the relevant scale for sizable temperature effects in the peak height is thus @xmath158 while it is @xmath150 in the smearing of the phase lapses . since we have chosen @xmath196 a reduction of @xmath161 is visible for temperatures at which the phase lapse is still fairly sharp [ see fig . [ fig4 ] ( b ) ] . due to the importance of correlation effects at small @xmath142 , the @xmath3 dependence of @xmath13 and @xmath135 in the universal regime is different from the noninteracting case . here the explicit temperature dependence of @xmath197 is much stronger and can not be neglected . the resulting @xmath3 dependence of @xmath13 and @xmath135 is shown in figs . [ fig4 ] ( a ) and ( c ) . a comparison to fig.[fig4 ] ( b ) shows that in the universal regime the smearing of the phase lapse sets in at a lower energy scale than in the mesoscopic regime . this scale depends on the relative sign @xmath145 of the level - lead hopping matrix elements [ compare figs . [ fig4 ] ( a ) and ( c ) ] . furthermore , in contrast to the mesoscopic regime the scales on which the cb peaks and the phase lapse are affected by temperature are comparable . a more detailed investigation of the temperature dependence in the universal regime , which also discusses the fate of the correlation - induced resonances , is beyond the scope of the present work and is left as subject for future studies . , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=49.5%,height=166 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=47.5%,height=166 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=49.5%,height=196 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=47.5%,height=196 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=49.5%,height=166 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=47.5%,height=166 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=49.5%,height=196 ] , but for @xmath198 and @xmath72.,title="fig:",scaledwidth=47.5%,height=196 ] we finally investigate the effect of the spin degree of freedom on the discussed phase lapse scenario , at @xmath21 . in figs . [ fig5 ] ( @xmath70 , @xmath199 ) and [ fig6 ] ( @xmath72 , @xmath198 ) we show frg data for the evolution of @xmath13 , @xmath135 and @xmath200 ( for a fixed spin direction ) with increasing @xmath4 for a generic @xmath201 . the overall dependence of @xmath135 on @xmath4 is similar to the one observed in the spinless case ( compare to figs . [ fig1 ] and [ fig2 ] ) . in particular , the behavior at small @xmath142 appears to be almost unaffected by the presence of the spin degree of freedom [ figs . [ fig5 ] ( a ) , ( b ) and [ fig6 ] ( a ) , ( b ) ] . for large @xmath142 [ figs . [ fig5 ] ( d ) and [ fig6 ] ( d ) ] the transmission resonances are located at odd average total filling of the two - level dot indicated by shoulders in the @xmath22 [ figs . [ fig5 ] ( d ) and [ fig6 ] ( d ) ] . at these fillings and for sufficiently large @xmath152 the kondo effect is active and the resonances can not be regarded as lorentzian - like cb peaks . instead they show a plateau - like shape known from the spinful single - level dot ( see ref . @xcite and references therein ) . across the kondo plateaus of @xmath8 the s - shaped increase of the phase is interrupted by a shoulder at @xmath202 as expected for the kondo effect @xcite . it would be very interesting to study how each of the kondo plateaus of @xmath8 with increasing temperature crosses over to two cb peaks and the phase behaves in the generated cb valley . this question is left for future investigations . the behavior of the phase in the presence of the kondo effect was experimentally investigated at temperatures comparable to the kondo temperature @xcite , and much below the kondo temperature @xcite . as we study the zero temperature case it is proper to compare our calculations to the measurements at low temperatures . indeed fig . [ fig6 ] ( c ) for intermediate @xmath142 qualitatively reproduces the experimental results at low temperature as shown in fig . 3 c of ref.@xcite . in particular the increase of the phase by more then @xmath0 and the absence of clearly developed kondo plateaus are reproduced . in fig . [ fig5 ] ( a ) we left out the frg data around @xmath203 as for these gate voltages some of the components of the flowing two - particle vertex become large . this indicates the breakdown of our present truncation scheme @xcite and the results for @xmath8 , @xmath1 and @xmath22 become unreliable . for an explicit comparison to nrg data of @xmath8 see fig . [ fig7 ] ( a ) . we note in passing that for @xmath70 correlation - induced resonances occur also in the model with spin [ fig . [ fig5 ] ( a ) ] @xcite . in fig . [ fig7 ] we compare frg and nrg results for l - r symmetric level - lead couplings . the computational resources required to obtain nrg data away from l - r symmetry become large and such data are not required for the aim of the present paper . we can then use eq . ( [ t_lr ] ) and must only compute the occupancies @xmath22 , which is numerically less demanding . for @xmath204 ( as exclusively shown ) it is only the absence of the correlation - induced resonances for @xmath70 [ compare figs . [ fig5 ] ( a ) and [ fig7 ] ( a ) ] which is different from the results of the generic @xmath151 shown in figs . [ fig5 ] and [ fig6 ] . with the exception of the @xmath205 regime in the case of small @xmath4 and @xmath70 the frg and nrg data compare quite well . in this case we only show the frg data for @xmath8 as the results for the phase and occupancies become rather erratic . the reason for the breakdown of the currently used truncated frg is explained in refs . @xcite and @xcite and is related to the fact that at small @xmath4 and small @xmath10 the correlations in effect become extremely large . , @xmath135 and @xmath200 ( per spin direction ) in the universal ( small @xmath142 ) and mesoscopic regime ( large @xmath142 ) of a spinful two - level dot at @xmath21 . the parameters are @xmath199 and @xmath206 . in ( a ) the frg data for @xmath135 and @xmath200 around @xmath203 are not shown . already the unsatisfactory comparison between frg and nrg for @xmath8 indicate that the frg becomes unreliable in this regime . for more details on this , see the text . for the nrg parameters we used @xmath207 , @xmath208 for @xmath70 and @xmath209 for @xmath72.,title="fig:",scaledwidth=47.5%,height=166 ] , @xmath135 and @xmath200 ( per spin direction ) in the universal ( small @xmath142 ) and mesoscopic regime ( large @xmath142 ) of a spinful two - level dot at @xmath21 . the parameters are @xmath199 and @xmath206 . in ( a ) the frg data for @xmath135 and @xmath200 around @xmath203 are not shown . already the unsatisfactory comparison between frg and nrg for @xmath8 indicate that the frg becomes unreliable in this regime . for more details on this , see the text . for the nrg parameters we used @xmath207 , @xmath208 for @xmath70 and @xmath209 for @xmath72.,title="fig:",scaledwidth=47.5%,height=166 ] , @xmath135 and @xmath200 ( per spin direction ) in the universal ( small @xmath142 ) and mesoscopic regime ( large @xmath142 ) of a spinful two - level dot at @xmath21 . the parameters are @xmath199 and @xmath206 . in ( a ) the frg data for @xmath135 and @xmath200 around @xmath203 are not shown . already the unsatisfactory comparison between frg and nrg for @xmath8 indicate that the frg becomes unreliable in this regime . for more details on this , see the text . for the nrg parameters we used @xmath207 , @xmath208 for @xmath70 and @xmath209 for @xmath72.,title="fig:",scaledwidth=47.5%,height=196 ] , @xmath135 and @xmath200 ( per spin direction ) in the universal ( small @xmath142 ) and mesoscopic regime ( large @xmath142 ) of a spinful two - level dot at @xmath21 . the parameters are @xmath199 and @xmath206 . in ( a ) the frg data for @xmath135 and @xmath200 around @xmath203 are not shown . already the unsatisfactory comparison between frg and nrg for @xmath8 indicate that the frg becomes unreliable in this regime . for more details on this , see the text . for the nrg parameters we used @xmath207 , @xmath208 for @xmath70 and @xmath209 for @xmath72.,title="fig:",scaledwidth=47.5%,height=196 ] in the present paper we studied the appearance of phase lapses in an interacting two - level quantum dot considering the entire parameter space using nrg and a truncated frg scheme . as a starting point we briefly discussed the noninteracting case at temperature @xmath21 and pointed out that for generic level - lead couplings , that is up to cases with increased symmetry , essential features of the universal phase lapse scenario are already established at @xmath5 . for single - particle level spacings @xmath4 small compared to the level broadenings @xmath158 the transmission is characterized by two transmission peaks of equal width with a transmission zero and an associated @xmath0 phase lapse between them ( universal regime ) . for large @xmath142 at @xmath5 the appearance or not of a transmission zero and phase lapse between the two transmission peaks depends on the relative sign @xmath145 of the level - lead couplings ( mesoscopic regime ) . within a spinless model we have shown that the separation of the two transmission peaks increases linearly with the interaction @xmath2 while the @xmath0 phase lapse and transmission zero remain in the valley between them . furthermore , with increasing @xmath2 the increase of the phase across the peaks takes an s - shape and the peaks become lorentzian - like , thus assuming shapes resembling those observed experimentally @xcite . for @xmath72 and increasing @xmath4 , a crossover occurs to a regime in which the @xmath0 phase lapse and transmission zero lies outside the two cb peaks . the crossover scale @xmath149 increases with increasing interaction and thus the coulomb repulsion stabilizes the universal phase lapse behavior . we have investigated the relation between phase lapses and population inversions of the level occupancies @xmath22 . we have shown that a mean - field treatment of the present problem correctly reproduces certain features of the behavior discussed above , but is not able to produce the universal phase lapse scenario at small @xmath142 due to artifacts of the approximation such as a phase lapse by less than @xmath0 @xcite and a vanishing of the transmission zero @xcite . furthermore , the discontinuous gate voltage dependence of the @xmath22 found in the mean - field approximation turned out to be an artifact . next , we studied how the phase lapse behavior is affect by temperatures @xmath18 . the universal phase lapse ( at small @xmath142 ) is smeared out but remains visible for not too large @xmath3 . in the mesoscopic regime with @xmath210 the smearing of the phase lapse and the decrease of the cb peaks can be understood in detail in analogy to the noninteracting case . for @xmath211 correlations are more important and a detailed understanding of the temperature dependence of the transmission @xmath56 requires further studies . the phase lapse behavior in both the universal and mesoscopic regimes is also stable if the spin degree of freedom is included . for sufficiently large @xmath152 in this case the kondo effect is active at odd average dot filling , leading to minor modifications of the scenario discussed above . in particular , at large @xmath142 the @xmath21 transmission peaks are kondo plateaus rather than lorentzian - like cb peaks . across these kondo plateaus the phase shows a shoulder at @xmath202 . in contrast , the behavior at small @xmath142 appears to be almost unaffected by the spin degree . a study of the combined effect of finite temperature and spin is left for future work . we thank r. berkovits , p. brouwer , y. gefen , l. glazman , d. golosov , m. heiblum , j. imry , v. kashcheyevs , j. knig , f. marquardt , m. pustilnik , h. schoeller , k. schnhammer , and a. silva for valuable discussions . this work was supported in part , for vm , by the dfg through sfb602 ; for th , aw and jvd by the dfg through sfb631 and de730/3 - 2 , and also by spintronics rtn ( hprn - ct-2002 - 00302 ) , nsf ( phy99 - 07949 ) and dip - h.2.1 ; and for yo by dip - h.2.1 , bsf and the humboldt foundation .

solutions of highly - charged polyelectrolytes , such as biopolymers ( dna , proteins , _ etc . _ ) and many synthetic polyelectrolytes , exhibit striking phenomena while multivalent salts or charged molecules are added into the solutions @xcite . if the concentration of the added salt lies within a window of concentration , the solution is separated in two phases : one is a viscous phase containing condensed or precipitated polyelectrolytes , and the other is a dilute phase in which polyelectrolytes are dissolved . if it lies outside the window , the solution is in a homogeneous phase . these phenomena are called `` reentrant condensation '' @xcite , or commonly referred as `` dna condensation '' in the fields of biology and medicine @xcite . this topic has attracted much attention recently due to its potential application in developing nonviral vectors of dna for the purpose of gene therapy @xcite . the amount of salt which results in the phase separation is determined by many factors , such as valence of salt , temperature , chain structure , and concentration of polyelectrolytes @xcite . it is known that the lower boundary of the window depends linearly on the concentration of polyelectrolytes but the upper one of the window is insensitive to it and roughly a constant . both flexible polyelectrolytes and stiff polyelectrolytes display similar phase diagrams @xcite . despite decades of effort , our understanding on polyelectrolytes is still limited and many issues , including the reentrant condensation , are far from being resolved @xcite . the aim of this paper is to investigate the properties of polyelectrolytes in salt solutions by means of computer simulations . there have been many simulations devoted to the study of polyelectrolytes @xcite . however , most of them investigated the properties in salt - free solutions @xcite . only recently , due to the progress of computing power , large - scale simulations for systems with added salt can be realized . the works included , for example , flexible chains in monovalent or multivalent salt solutions @xcite , semiflexible chains with multivalent salt or charged molecules @xcite , and rigid chains in the presence of multivalent salts @xcite . but the salt concentration or the salt valence was not large enough to observe any significant evidence being able to connect to the redissolution of polyelectrolytes . therefore , the reentrant condensation could not be discussed thoroughly in these works . et al . _ @xcite recently studied the effective interaction between two immobile dnas in the presence of tetravalent counterions . they have found that the effective interaction yields an attractive minimum at short distances which then disappears in favor of a shallow minimum at large separations , indicating the occurrence of dna condensation and redissolution and a novel stable mesocrystal . in a more recent study @xcite , we have investigated the properties of mobile polyelectrolytes in tetravalent salt solutions and demonstrated that following a collapsed transition , polyelectrolytes undergo reexpanding transition upon addition of the salt . this phenomenon is just a single - chain version of the condensation and the redissolution of polyelectrolytes . moreover , by varying the ion size , we have shown that the excluded volume of ions plays a decisive role on the reentrant condensation . however , to well understand the phenomena , there is still a far way to go . for example , is it possible to obtain any information related to the phase boundary of reentrant condensation from simulations ? there exist theories to explain the phase boundary @xcite but it has never been studied in simulations . concerning the size of polyelectrolyte in multivalent salts , muthukumar and coworkers have applied single and double screening theories to explain the collapsed transition of chains @xcite . however , no existing theory can predict chain reexpansion in the region of high salt concentration . moreover , chain morphology is essential to understand the properties of polyelectrolytes . it can be described by many quantities , such as asphericity and prolateness parameter @xcite . to our knowledge , no such study has been reported yet for strongly - charged polyelectrolytes . an alternative way to understand polymers is to study scaling behavior . although the concept of scaling has been successfully applied to neutral polymers @xcite , application of this concept to polyelectrolytes @xcite has encountered many difficulties owing to the long - range nature of coulomb interaction . in this paper , we compute the single - chain structure factor . the swelling exponent is then calculated by studying the power - law - like regime of the structure factor . since properties of polyelectrolytes depend strongly on ion size , we investigate in this paper the effect of ion size too . in our previous work @xcite , the sizes of ions were set to identical and varied simultaneously . as a consequence , bjerrum association @xcite was strongly intervened , particularly when ions were small . this setup has a disadvantage that the system was easily jammed by large ions , leading to the impossibility of investigation of the cases of big ions . therefore , in this study we vary only the size of the multivalent counterions and fix unchanged the one of the other ions , so that the ion size effect is restrictedly investigated . in the study of polyelectrolytes , it is tempting to discuss the long range nature of coulomb interaction using the concept of electrostatic persistence length . odijk @xcite , skolnick and fixman @xcite predicted that the electrostatic persistence length is proportional to the square of the screening length @xmath0 . the deduction was based upon debye - hckel theory . consequently , the effective intramolecular interaction is always repulsive , resulting in a positive electrostatic persistence length @xcite . however , experiments have shown that electrostatic persistence length can be negative under certain condition so that electrostatics gives a reversed contribution to chain flexibility @xcite . simulations on this issue @xcite were mainly performed within debye - hckel approximation using a screened coulomb potential @xmath1 to describe the screening effect of salt without involving explicitly salt ions . a negative electrostatic persistence length , hence , has never been reproduced in the simulations . noticeably , it is known that linear screening theory is valid only for a weakly - charged system and breaks down for a highly - charged one , such as dna and many biopolymers , due to strong ion correlations @xcite . recently , by employing a loop expansion method which goes beyond mean field , ariel and andelman were able to predict the negative regime of the electrostatic persistence length for a strongly - charged rodlike polyelectrolyte @xcite . therefore , we expect that to observe a negative electrostatic persistence length , one should incorporate salt ions explicitly and use the bare coulomb potential in simulations . our study fulfills this requirement and hence offer a good chance to verify the above phenomenon and theory . charge inversion is a universal phenomenon , occurred when a strongly charged macroion binds so many counterions that its net charge alters sign @xcite . _ intended to link charge inversion with the reentrant condensation of polyelectrolytes @xcite . the idea was based upon that counterions form strongly correlated liquid on the surface of polyelectrolytes . in an intermediate salt region , the bare charges of the chains are almost neutralized by the condensed counterions ; the correlation - induced attraction results in the condensation of polyelectrolytes . in a high - salt region , the counterions overcompensate the chain bare charges ; coulomb repulsion between chains dominates the attraction and , in consequence , the condensates are redissolved into the solution . solis and olvera de la cruz gave a different point of view using a theory called `` two - state model '' @xcite . they predicted that charge inversion is not necessarily occurred with redissolution of polyelectrolytes . although overcompensated by counterions , the redissolved polyelectrolytes can have either sign due to the association of coions @xcite . this association depends strongly on the ion size which was not considered in the study of nguyen _ et al . _ @xcite . therefore , solis and olvera de la cruz predicted a charge distribution which oscillates from a polyelectrolyte , in accordance with the profile observed in our previous work @xcite . in this paper , we investigate more extreme cases , including much smaller and much larger ions . we also study like - charge attraction between chains by calculating the potential of mean force and determine the conditions under which the attraction takes place . the rest of the article is organized as follows . model and simulation details are given in sec . [ sec_model_method ] . results are discussed in sec . [ sec_result_discussion ] . it is divided into five subsections . the first subsection ( sec . [ sec_rg2_cm ] ) is devoted to the study of the radius of gyration of a polyelectrolyte where salt concentration and polyelectrolyte concentration are varied . the results provide information helpful in understanding the boundaries of the condensation window . [ sec_shape ] studies the morphology of multiple polyelectrolytes in salt solutions . we compute several quantities , including shape factors , asphericity , and prolateness parameter . the effect of size of multivalent ions is discussed . single - chain structure factor is , then , calculated in sec . [ sec_sf ] . by least - square fitting the power - law - like regime , we compute the swelling exponent of a polyelectrolyte , which provides a fundamental vision how strongly - charged polyelectrolytes scale in salt solutions . sec . [ sec_lp ] deals with persistence length . the negative regime of the electrostatic persistence length is successfully reproduced in the simulations . comparison of the results with theoretical predictions is made . [ sec_pmf ] provides a general picture of the effective interaction between polyelectrolytes by computing the potential of mean force . as long as knowing the condition to appear like - charge attraction between chains and the integrated charge distribution around a chain ( sec . [ sec_iondist ] ) , we are able to arbitrate if the phenomena of charge inversion and the redissolution of polyelectrolytes are related . we give our conclusions in sec . [ sec_conclusion ] . we employed a coarse - grained model to simulate a polyelectrolyte system . the system contains four bead - spring chains , one of which is composed of 48 beads ( monomers ) . each monomer dissociates a monovalent cation ( counterion ) into a solution and carries a negative unit charge . there are totally 192 monovalent cations in the solution . solvent molecules are not modeled explicitly . their effect is taken into account implicitly by considering them as a medium of constant dielectric constant . the geometry of the medium is cubic and periodic boundary condition is applied to simulate bulk environment . salt molecules are added into the system . in the solution , they are dissociated into cations ( counterions ) and anions ( coions ) . we assume that the cations are tetravalent and the anions are monovalent . the amount of the cations and anions obeys the condition of charge neutrality . three types of interactions are included in our simulations . the first one is the excluded volume interaction described by a purely repulsive lennard - jones ( lj ) potential , @xmath2 + \varepsilon_{\rm lj } & \mbox{\ for $ r \le \root 6 \of 2 \sigma$\;}\\ 0 & \mbox{\ for $ r > \root 6 \of 2 \sigma$\ ; } \end{array } \right.,\ ] ] where @xmath3 is the distance between two particles , @xmath4 is the interaction strength , and @xmath5 is the collision diameter . we assume that all the particles have identical @xmath4 but the collision diameter may be different . the collision diameters for monomer , monovalent cation , tetravalent cation , and anion are denoted by @xmath6 , @xmath7 , @xmath8 , and @xmath9 , respectively . lorentz - berthelot mixing rule @xcite is applied for the interaction between different kinds of particles . the second interaction is called finitely extensible nonlinear elastic ( fene ) potential @xcite and used to describe the bond connection between adjacent monomers on a chain . it reads as @xmath10 where @xmath11 is the spring constant , @xmath12 is bond length , and @xmath13 is the maximum extension of a bond . all particles interact via the third interaction , coulomb interaction , which is written as @xmath14 where @xmath15 is the charge of the @xmath16th particle , @xmath17 is the temperature , @xmath18 is the boltzmann constant , and @xmath19 is the bjerrum length defined as the separation distance at which the coulomb potential between two unit charges @xmath20 equals to the thermal energy @xmath21 . @xmath19 is equal to @xmath22 where @xmath23 is the dielectric constant and @xmath24 is the vacuum permittivity . we perform molecular dynamics ( md ) simulations in this study sjplimp / lammps.html . ] . we set @xmath7 and @xmath9 equal to @xmath6 and vary the size of tetravalent cation @xmath8 from @xmath25 to @xmath26 to investigate the effect of ion size on properties of the polyelectrolyte solution . simulation parameters are chosen as follows : @xmath27 , @xmath28 , and @xmath29 . the bjerrum length is set to @xmath30 . similar systems in salt - free solutions have been investigated by stevens and kremer @xcite . we assume that the masses of all the particles are identical and equal to @xmath31 , and choose @xmath6 and @xmath32 to be the length unit and the energy unit , respectively . therefore , the natural time unit is @xmath33 . the system is studied in canonical ensemble where the temperature is controlled using nos - hoover thermostat with the frequency of heat transfer with thermostat equal to @xmath34 . coulomb interaction is calculated by the technique of ewald sum . except in sec . [ sec_rg2_cm ] , the monomer concentration is fixed at @xmath35 which corresponds to a box size @xmath36 , whereas the salt concentration is varied . the amount of salt is controlled such that the volume fraction of total particles is smaller than @xmath37 . equations of motion are integrated applying verlet algorithm where time step is set to @xmath38 . each simulation starts with an equilibration phase of @xmath39 md steps and is followed by a production run of @xmath40 md steps in which data are collected every 1000 steps . in order to shorten the notation , we will use @xmath6 as the length unit , @xmath41 as the concentration unit , and @xmath20 as the charge unit in the following text . we started our study by simulating a single polyelectrolyte with added tetravalent salt at various monomer concentrations . the chain consists of 48 monomers and the size of the tetravalent cations is set to be identical to the one of the monomers . we used the radius of gyration to characterize the chain size . the radius of gyration @xmath42 is calculated by the equation @xmath43 where @xmath44 is the number of monomers on the chain , @xmath45 is the position vector of the @xmath16th monomer , and @xmath46 the center of mass of the chain . the results are presented in fig . [ cm_p1n48_rg2.eps ] where the mean square radius of gyration @xmath47 is plotted as a function of salt concentration @xmath48 at four monomer concentrations , @xmath49 , @xmath50 , @xmath51 and @xmath52 . as a function of @xmath48 at four monomer concentrations , @xmath49 , @xmath50 , @xmath51 , and @xmath52 . the symbols used to represent the four @xmath53 are indicated at the left - bottom side of the figure . the value of @xmath47 in a salt - free solution ( @xmath48=0.0 ) is depicted near the left vertical axis . the four curves in the regime @xmath54 can overlap each other ( shown in the inset ) under a suitable transformation described in the text . , scaledwidth=40.0% ] the angular bracket @xmath55 means averaging over chain configurations . two behaviors were observed . the first one is that each curve displays a similar trend of behavior against salt concentration . upon addition of salt , the chain undergoes two structural transitions : a collapsed transition and a followed reexpanding transition , as having been observed in the previous work @xcite . the first transition results from the screening of the coulomb interaction by the added salt . the second transition is still not well understood although it is a result of the competition between energetic gain and entropic loose in counterion condensation and in shape transformation of polyelectrolytes to minimize the free energy . we observed that the minimum of @xmath47 appears near the salt concentration @xmath56 and the more dilute the monomer concentration , the smaller value the minimum of @xmath47 will be . @xmath57 is the equivalence point at which the amount of charges of the tetravalent counterions exactly neutralizes the charges on polyelectrolytes as if the salts and the polyelectrolytes are in the stoichiometric reaction . we will call @xmath57 the equivalence concentration . in a salt - free solution ( @xmath58 ) , we found that the chain size increases with decreasing monomer concentration ( refer to the four data points near the left vertical axis of the figure ) . this effect has been discussed in literatures @xcite and our results are in accordance with them . and the second behavior is that @xmath47 for different @xmath53 attains approximately the same value while the salt concentration is larger than the equivalence concentration . therefore , the curves approximately overlap each other . we noticed , furthermore , that in the region @xmath54 , the @xmath47 curves for different @xmath53 can also lie over each other under an appropriate transformation . we demonstrate this overlap in the inset of fig . [ cm_p1n48_rg2.eps ] where the abscissa is @xmath59 and the ordinate is @xmath60 with @xmath61 the minimum value of @xmath47 . @xmath62 depends smoothly on @xmath53 and , in the figure , is equal to @xmath63 , @xmath64 , @xmath65 , and @xmath66 for @xmath49 , @xmath50 , @xmath51 , and @xmath52 , respectively . the size of polyelectrolytes in multivalent salt solutions has been studied by small - angle scattering experiments @xcite . the results showed that it diminishes with increasing or decreasing @xmath48 toward the region where the condensation of polyelectrolytes occurs ( cf . table 2 of ref . the results of our simulations are in accordance with the experiments . we make a hypothesis that chain size can be served as a qualitative condition to distinguish between a monophasic solution of polyelectrolyte and a diphasic one . more precisely , if the chain size is larger than some value , the chains are dissolved in the solution and the system is in a homogeneous phase ; if it is smaller , the chains are precipitated from the solution and a phase separation occurs . under this hypothesis , polyelectrolyte solutions of different @xmath53 become monophasic as @xmath67 is increased over a concentration in the high - salt region , since the chain size exceeds the some value nearly at the same @xmath48 due to the similarity of the @xmath47 curves against @xmath48 in this region . therefore , the upper boundary of the salt concentration window for the condensation of polyelectrolyte is independent of @xmath53 . similarly , the lower boundary of the window approximately depends linearly on @xmath53 , owing to the similarity of the @xmath60 curves against @xmath59 in the low - salt region ; the weak dependence of @xmath62 and @xmath61 on @xmath53 ensures that the chain size surpasses the some value roughly at the same @xmath59 . therefore , fig . [ cm_p1n48_rg2.eps ] provides an essential information to understand the boundaries of reentrant condensation . the aim of this article is to understand the properties of multiple flexible polyelectrolytes with added salt in an aqueous solution . it is known that the bjerrum length @xmath19 in water is equal to 7.14 at room temperature . our simulation setup , hence , corresponds to a collision diameter @xmath6 of monomer equal to 2.38 . this model represents a polyelectrolyte system with linear charge density equal to a unit charge per @xmath68 . ] , and can be employed to simulate a strongly - charged system such as sodium polystyrene sulfonate . the four monomer concentrations studied in this section , while converted to real unit , are equal to 0.039 m , 0.246 m , 0.986 m , and 1.971 m , respectively . even the largest @xmath53 is still smaller than the overlap threshold @xmath69 estimated by @xmath70 . therefore , our systems are dilute polymer solutions . readers should be aware that the value up to nearly 2 m is a very high monomer concentration for a real polymer system , and our systems are _ dilute _ only in the formal meaning of the world since the chains are very short . in the following sections , we fix @xmath53 at 0.008 and simulate a system comprising multiple chains . this monomer concentration is one or two order of magnitude higher than a typical monomer concentration used in experiments @xcite . however , based on the fact that systems at different monomer concentrations reveal similar evolution of chain size against salt concentration , the results obtained at @xmath71 can be used to understand a system at more dilute @xmath53 . we mention that this @xmath53 is a choice so that the simulation box is not so big . therefore , the amount of the added salt is controllable under limited computing resources . consequently , to study the behavior of polyelectrolytes in salt solutions , covering a broad range of salt concentration , becomes numerically feasible . in this section and in the followings , the studied system contains four polyelectrolytes and the monomer concentration is fixed at @xmath71 . this section is devoted to the study of chain conformation in tetravalent salt solutions . the effects of salt concentration and size of tetravalent counterions are investigated . we vary the size of tetravalent counterion , @xmath8 , from 0.0 to 4.0 and fix the size of the other ions at 1.0 . a large @xmath8 can represent a large cation , a large ion group , or a large charged colloid . it can be also a result of the hydration of a small ion . a small @xmath8 may be less representative for real systems but theorists are generally interested in it because many models are established on the base of vanishing ion size . morphology of a polymer can be quantified by many quantities . the first quantity that we studied is the shape factor @xmath72 , defined as the ratio of the mean - square end - to - end distance , @xmath73 , to the mean - square radius of gyration @xmath47 . @xmath72 attains a value 12 for a rodlike polymer , 6.3 for a flexible chain in a good solvent , and 6 for an ideal chain @xcite . for a compact structure , the value of @xmath72 is small . for example , if the chain has a spherical morphology and the two ends are randomly positioned , @xmath72 is equal to 2 . in fig.[shapep4n48_eta1&xi1.eps](a ) , we show how the shape factor @xmath72 varies with @xmath48 at different @xmath8 . as a function of @xmath48 , and ( b ) @xmath74 as a function of @xmath48 , for the cases of different size @xmath8 of the tetravalent counterions . the symbol used for each @xmath8 is indicated at the left - bottom side of the figures . the error bars are smaller than the size of the symbols . the result for neutral polymers is depicted as reference and marked by `` neutral '' . , title="fig:",scaledwidth=40.0% ] ( a ) as a function of @xmath48 , and ( b ) @xmath74 as a function of @xmath48 , for the cases of different size @xmath8 of the tetravalent counterions . the symbol used for each @xmath8 is indicated at the left - bottom side of the figures . the error bars are smaller than the size of the symbols . the result for neutral polymers is depicted as reference and marked by `` neutral '' . , title="fig:",scaledwidth=40.0% ] ( b ) in the figure , a reference curve , marked by `` neutral '' , is depicted . it represents a system composed of four neutral polymers of 48 monomers in tetravalent salt ( @xmath75 ) solutions . focus firstly on the polyelectrolyte solutions with @xmath76 , 1.4 and 2.0 . we witnessed that @xmath72 decreases at first and then increases upon addition of salt , in analogy of the radius of gyration studied in the previous section , and shows a v - shaped curve against @xmath48 in the semilog plot . in the low - salt region or in the high - salt region , @xmath72 is larger than or close to the value 6.3 , indicating that the chains have elongated or coil - like structures . in between , it appears a minimum near the equivalence concentration @xmath77 . the value of @xmath72 tells us that the chains possess compact structures but not so compact as spheres . for the cases with larger @xmath8 ( @xmath78 and @xmath79 ) , @xmath72 shows a shallow minimum and the whole curve stays basically above the value 6.3 . in these cases , the chains are in an extended state . it is worth to notice that for @xmath80 , @xmath72 , in principle , increases with @xmath8 at a given @xmath48 . now pay attention to the cases with @xmath81 . @xmath72 displays a similar v - shaped curve against @xmath48 but opposite to the previous cases , its value increases with decreasing @xmath8 . moreover , the degree of chain reexpansion in the high - salt region is weakened while @xmath8 is decreased . noticeably , for the case @xmath82 , @xmath72 displays a sharp transition from a value 7.2 in the low - salt solution to a value 6.5 in the vicinity of @xmath57 . in the latter region , the chains behave similarly to neutral polymers . the second quantity that we studied is @xmath83 , which is the ratio of the radius of gyration @xmath84 to the hydrodynamic radius @xmath85 computed by @xmath86 . @xmath83 is equal to @xmath87 for a hard sphere and attains a value 2.25 for a rodlike chain @xcite . renormalization group theory has predicted @xmath88 in a good solvent and @xmath89 at @xmath90 point @xcite , both of which have been shown in agreement with experiments @xcite . we present in fig . [ shapep4n48_eta1&xi1.eps](b ) the variation of @xmath83 against @xmath48 . the referenced curve for the system of neutral polymers is depicted . the curve was found very close to the value 1.24 , which seems to indicate an ideal - chain behavior . however , the neutral chains should behave as random coils and the value of @xmath83 would be larger than 1.24 . the deviation of @xmath83 from the value of a coil is probably due to the short chain length used in our simulations . _ have studied the finite size effect and shown that the dependence of @xmath83 on chain length is not negligible @xcite . therefore , @xmath83 reported here departs slightly from the asymptotic value of an infinite chain length . nonetheless , the results are still indicative and useful to understand how @xmath83 varies with salt concentration . in the small @xmath48 region , the polyelectrolytes are more extended than the neutral polymers . in the middle region of @xmath48 , the chains attain compact structures when @xmath91 . the dependence of @xmath83 upon @xmath8 resembles to that of @xmath72 . for large @xmath8 ( @xmath92 , the chains are elongated . for vanishing @xmath8 , we observed that following a sharp decrease , @xmath83 reattains approximately a constant value , close to that for the neutral polymers . a more fundamental way to study the shape of a polymer is to investigate the tensor of radius of gyration @xmath93 @xcite , defined by : @xmath94 where @xmath95 , @xmath96 denote the three cartesian components and @xmath97 is the @xmath95 component of the position of the @xmath16th monomer . let @xmath98 , @xmath99 and @xmath100 be the three eigenvalues of @xmath101 . a quantity called `` asphericity '' @xcite , which measures the deformation from a spherical geometry , is defined by @xmath102 it takes a value between 0 ( sphere ) and 1 ( rod ) . for a coil polymer , it is 0.431 obtained from simulations @xcite . another quantity describing the degree of prolateness is defined by @xmath103 where @xmath104 @xcite . it is ranged from @xmath105 to @xmath106 and can be used to distinguish between a prolate shape ( @xmath107 ) and an oblate one ( @xmath108 ) . for neutral polymers in the dilute limit , simulations predict a prolate chain with @xmath109 @xcite . we present in figs . [ shapep4n48_asph&sp.eps](a ) and [ shapep4n48_asph&sp.eps](b ) the asphericity @xmath110 and the prolateness parameter @xmath111 as a function of @xmath48 . as a function of @xmath48 , and ( b ) @xmath111 as a function of @xmath48 , for different @xmath8 . the symbol used for each @xmath8 is indicated at the left - bottom side of the figures . the error bars are smaller than the size of the symbols . the result for neutral polymers is depicted as reference and marked by `` neutral '' . , title="fig:",scaledwidth=40.0% ] ( a ) as a function of @xmath48 , and ( b ) @xmath111 as a function of @xmath48 , for different @xmath8 . the symbol used for each @xmath8 is indicated at the left - bottom side of the figures . the error bars are smaller than the size of the symbols . the result for neutral polymers is depicted as reference and marked by `` neutral '' . , title="fig:",scaledwidth=40.0% ] ( b ) resembling to the behaviors of @xmath72 and @xmath83 , @xmath110 and @xmath111 show a minimum value in the vicinity of @xmath57 . the curve associated to @xmath75 roughly separates two different behaviors of evolution : @xmath110 and @xmath111 increase with increasing and with decreasing @xmath8 from 1.0 . the value of @xmath110 lies between 0.275 and 0.55 . therefore , even in a compact structure ( where @xmath48 is near @xmath57 ) , the chain shape is asymmetric . in a low - salt solution , the chains show extended morphology but are not as elongated as rods . on the other hand , @xmath111 is larger than zero for all of the studied cases , indicating that the polyelectrolytes display prolate conformation . at a high salt concentration , @xmath110 is close to 0.45 and @xmath111 is close to 0.54 , suggesting once more that the polyelectrolytes behave as the neutral polymers in the high - salt region . for the case with null @xmath8 , chain reexpansion does not occur ; @xmath110 and @xmath111 stay at a constant value when @xmath48 goes beyond @xmath57 . the results obtained in this section strongly suggest that @xmath75 is the optimal condition to pack polyelectrolytes into the smallest volume . in this section , we investigate the single - chain structure factor . this quantity is experimentally accessible and provides an essential information to theoretical analysis . the single - chain structure factor is defined by @xmath112 where @xmath113 is the scattering wavevector , @xmath114 is the number of polymer chains , @xmath115 is the number of monomers on a chain , and @xmath116 is the position of the @xmath117th monomer on the @xmath118th chain . suppose that the system has spherical symmetry . the dependence of the structure factor on the orientation of @xmath113 can be therefore integrated out : @xmath119 where @xmath120 is the norm of @xmath113 and @xmath121 is the solid angle . this integration was performed numerically by sampling over a set of randomly - oriented @xmath113 vectors . we present , in fig . [ sfp4n48_sxxx&txxx](a ) , @xmath122 at several salt concentrations for the case of @xmath75 . for the reason of clarity , the curves in the figure have been multiplied by @xmath123 with @xmath124 being increased by one for each curve from the bottom to the top . @xmath122 for the neutral polymers is depicted as reference with @xmath125 . for @xmath75 at different @xmath48 and ( b ) @xmath122 at @xmath126 for different @xmath8 . the value of @xmath48 or @xmath8 , at which the simulations were running , is indicated at the left side of each corresponding curve . @xmath122 for the neutral polymers at @xmath126 is depicted as reference and marked by `` neutral '' . for clarity , the curves , from bottom to top , have been multiplied by a factor @xmath127 with @xmath124 started from zero and increased by one for every curve . the dotted curve , near each @xmath122 and terminated at @xmath128 , is the guinier function @xmath129 . two dashed lines , which represent two power - law dependences , @xmath130 and @xmath131 ( marked by `` -1 '' and `` -2 '' , respectively ) , are drawn in the plots.,title="fig:",scaledwidth=40.0% ] + ( a ) + for @xmath75 at different @xmath48 and ( b ) @xmath122 at @xmath126 for different @xmath8 . the value of @xmath48 or @xmath8 , at which the simulations were running , is indicated at the left side of each corresponding curve . @xmath122 for the neutral polymers at @xmath126 is depicted as reference and marked by `` neutral '' . for clarity , the curves , from bottom to top , have been multiplied by a factor @xmath127 with @xmath124 started from zero and increased by one for every curve . the dotted curve , near each @xmath122 and terminated at @xmath128 , is the guinier function @xmath129 . two dashed lines , which represent two power - law dependences , @xmath130 and @xmath131 ( marked by `` -1 '' and `` -2 '' , respectively ) , are drawn in the plots.,title="fig:",scaledwidth=40.0% ] + ( b ) + we witnessed that @xmath122 shows typical behavior . in the guinier regime ( @xmath132 ) , @xmath122 behaves similarly to @xmath129 . it is demonstrated by plotting @xmath129 ( in dotted curve ) near the corresponding @xmath122 with the value of @xmath133 adopted from sec . [ sec_shape ] . the dotted curve terminates at @xmath128 in order to indicate the boundary of the guinier regime . in the regime @xmath134 , @xmath122 shows power - law - like behavior , manifested by a linear dependence of @xmath135 on @xmath136 . the slope @xmath137 of the linear dependence is related to the swelling exponent @xmath138 by the relation @xmath139 , which describes the size extension of a polymer : @xmath140 . in the regime @xmath141 , @xmath122 displays non - universal behavior . however , while @xmath142 , self - scattering of monomer is the only contribution , leading , therefore , @xmath122 to a universal value @xmath63 . we mention that the krathy regime , in which @xmath143 , is not observed in our study because the system is a dilute solution and the chain length is not long enough @xcite . for @xmath8 other than 1.0 , we observed analogous trends of behavior . for example , the variation of @xmath122 for different @xmath8 at the equivalence concentration is presented in fig . [ sfp4n48_sxxx&txxx](b ) , in which we rediscovered the three regimes : the guinier regime , the power - law - like regime and the non - universal regime . we noticed that in the last regime , a small bounce appears on the curve and , with increasing the size of tetravalent ions , gradually moves toward small @xmath120 . the swelling exponent @xmath138 of the polyelectrolytes was calculated by performing least - square fits in the log - log plot of @xmath122 in the power - law - like regime . the fitting region was chosen to be @xmath144 and the fitting reliability has been verified to be excellent . we present the results of the swelling exponent @xmath138 in fig . [ nup4n48.eps ] . as a function of @xmath48 for different @xmath8 . the symbol used for each @xmath8 is indicated at the left - bottom side of the figure . @xmath138 for the referenced neutral - polymer system is depicted and marked by `` neutral '' . , scaledwidth=40.0% ] @xmath138 for the referenced neutral polymers was verified at first and found to attain a value close to the flory s one , @xmath145 . next , we focused on the swelling exponent for the polyelectrolytes . we observed that @xmath138 exceeds the value for the neutral polymers at a low salt concentration . it indicates a strong extension of chain size due to coulomb repulsion between monomers . for the cases with @xmath146 , 1,0 , 1,4 , and 2.0 , the swelling exponent shows a v - shaped curve against @xmath48 : @xmath138 decreases and then increases with increasing salt concentration . a minimum occurs near the equivalence concentration @xmath147 and , at this moment , the chains show compact structures but are less compact than spheres because a spherical structure scales as @xmath148 . we remark that the swelling exponent obtained here for @xmath75 is consistent with that obtained by direct variation of the chain length of a system containing a single polyelectrolyte @xcite except in the mid - salt region where @xmath138 in the latter study is slightly smaller and acquires a value @xmath149 at @xmath147 . this deviation is due to multi - chain aggregation occurred in our system , which deforms the chain shape from a sphere and , consequently , increases @xmath138 . for the case with vanishing @xmath8 , following a drastic decrease , @xmath138 retains roughly a value 0.6 in the high - salt region and the chains , hence , scale like the neutral polymers . for the cases with large @xmath8 ( @xmath78 and 4.0 ) , @xmath138 is greater than 0.6 but still less than 1.0 . the chains , hence , do not reveal rigid rod - like structures . in principle , @xmath75 separates two behaviors of evolution at a salt concentration : @xmath138 increases with increasing and with decreasing @xmath8 from 1.0 . the results demonstrate again that the polyelectrolytes are mostly condensed while the size of the tetravalent counterions is compatible with the one of the monomers . there is a considerable interest to understand the variation of the persistence length of a polyelectrolyte against salt concentration . persistence length is a local property , which measures the stiffness of a polymer , and is distinguished from global conformational quantities . there are many ways to define the persistence length @xcite . in this section , we adopt a microscopic definition to calculate it : @xmath150 where @xmath151 is the @xmath117th bond vector and @xmath152 is the root - mean - square bond length . in this definition , the contribution of the finite bond length , which corresponds to the @xmath153 term in eq . [ persist_lp ] , is considered . the obtaining @xmath154 as a function of @xmath48 for different @xmath8 is plotted in fig . [ lpp4n48.eps ] . as a function of @xmath48 for different @xmath8 . the symbol used for each @xmath8 is indicated at the left - bottom side of the figure . the error bars are smaller than the size of the symbols . the bare persistence length @xmath155 is depicted as reference and marked by `` neutral '' . , scaledwidth=40.0% ] resembling to the quantities of chain conformation , @xmath154 decreases in the chain collapse region and increases in the chain reexpansion region . a minimum value appears in the vicinity of @xmath57 . the degree of increase diminishes with decreasing @xmath8 . for vanishing @xmath8 , it is so weak that @xmath154 is roughly a constant as @xmath156 . similar to the results in the previous sections , @xmath154 increases with both increasing and decreasing @xmath8 from 1.0 . @xmath75 , therefore , denotes a condition to separate the two behaviors of evolution of @xmath154 . the persistence length has generally two contributions : the bare persistence length @xmath155 ( in the absence of electrostatic force ) and the electrostatic persistence length @xmath157 ( owing to the coulomb interaction ) @xcite . in this study , the bare persistence length @xmath155 was calculated by performing simulations on the referenced system where four neutral chains and 192 neutral particles are placed in salt solutions . the result is depicted in fig . [ lpp4n48.eps ] and the curve is marked by `` neutral '' . we found that @xmath155 is not sensitive to the salt concentration and roughly attains a value 2.3 . it slightly decreases in the high - salt region , probably due to the jamming effect by the salt ions . for the polyelectrolyte solutions with large @xmath8 ( @xmath78 and 4.0 ) , the @xmath154 curve thoroughly lies above the @xmath155 curve , indicating a more rigid flexibility than a neutral polymer . the electrostatic persistence length is , hence , positive . on the other hand , for the polyelectrolyte solutions with @xmath158 , @xmath154 is smaller than @xmath155 in the mid - salt region and shows one or two intersections with the @xmath155 curve . therefore , @xmath159 is positive in a salt - free solution and becomes eventually negative upon addition of salt . it increases in the high - salt region and reattains a positive value for @xmath160 or @xmath161 . the negative contribution of electrostatics to the persistence length has been experimentally observed @xcite . recently ariel and andelman ( aa ) pointed out that a negative @xmath157 indicates a mechanical instability ( collapse ) of polyelectrolytes due to the presence of multivalent counterions @xcite . by taking into account the thermal fluctuations and correlations between bound counterions , they derived an expression of @xmath157 for an intrinsically - rigid charged polymer in ( @xmath162:1)-salt solutions in the limit of @xmath163 : @xmath164 where @xmath165 is the coulomb strength parameter between a monomer and a @xmath162-valent counterion , @xmath166 is the inverse debye - hckel screening length , and @xmath167 is the electrostatic persistence length deduced by odijk - skolnick - fixman ( osf ) theory @xcite . in the derivation , they have assumed point - like charges for the ions . we present , in fig . [ lep4n48.eps ] , @xmath157 obtained from the simulations and that predicted by eq . ( [ le_ariel ] ) as a function of @xmath168 . as a function of @xmath168 for different @xmath8 . the symbol used for each @xmath8 is indicated at the left - bottom side of the figure . the error bars are smaller than the size of the symbols . the dashed curve depicts @xmath169 predicted by eq . ( [ le_ariel ] ) . , scaledwidth=40.0% ] we witnessed that @xmath157 depends strongly on the size of the tetravalent ions . the negative regime of @xmath157 has been successfully captured in the simulations . to make a comparison with the theoretical prediction , we focused on the case with vanishing @xmath8 . please notice that this case does not exactly fulfill the assumption of the theory because in the simulations , the monovalent counterions and coions are not point charge - like and , also , the chains are not rigid . however , according to our experiences @xcite , ions other than multivalent counterions do not play a decisive role on the properties of polyelectrolytes . this comparison , therefore , makes sense if the effect of chain rigidness is negligible . we emphasize that the electrostatic correlations are fully taken into account in the simulations . in our study , the region in which eq . ( [ le_ariel ] ) is valid is @xmath170 because @xmath171 , @xmath172 , and @xmath173 is about 1.1 . notice that eq . ( [ le_ariel ] ) has a singularity at @xmath174 and , hence , the curve is divided into two branches . the left branch is located in the valid region but fails to predict @xmath157 . it gives not only a wrong value but also a wrong trend of variation of @xmath157 against @xmath168 . we noticed that eq . ( [ le_ariel ] ) is a modification of @xmath175 multiplied by a factor but the valid region given by aa theory @xcite is incompatible with that of osf theory . osf theory is held when the screening length @xmath176 is much smaller than the bare persistence length @xcite . therefore , aa theory may be true in the region @xmath177 . as shown in the figure , @xmath157 for vanishing @xmath8 is well described by the right branch of the curve of eq . ( [ le_ariel ] ) , which seems to support this conjecture . this topic deserves a further analysis in the future . experiments and simulations have shown that like - charged macroions can attract each other in a solution @xcite . it is hence relevant to know under which conditions the like - charge attraction takes place between polyelectrolytes . to answer this question , we investigate `` potential of mean force '' in this section . potential of mean force describes the effective interaction between two particles ( molecules ) in a medium , which includes two contributions : the direct interaction between two particles ( molecules ) and the indirect interaction via other particles ( molecules ) and the medium . this quantity can be calculated by simulations : @xmath178 where @xmath179 is the radial distribution function between particles of species @xmath95 and @xmath180 . @xmath179 is computed by @xmath181 where @xmath182 and @xmath183 are the number of @xmath95-particles and that of @xmath180-particles , respectively , @xmath184 is the volume of the simulation box , and @xmath185 with @xmath186 denoting the number of @xmath180-particles lying in a spherical shell , centered at the @xmath95-particle @xmath16 , of radius @xmath3 and shell thickness @xmath187 . because our system is composed of only four chains , it is not easy to reduce statistical error if we directly calculate the potential of mean force between chains . we , therefore , calculated the potential of mean force between monomers on different chains , which also provides the information about the chain interaction . we denote this quantity @xmath188 and the results are presented in fig . [ pmf11p4n48txxx ] which contains six plots , from top to bottom and from left to right , corresponding , in orders , to the cases @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 . @xmath188 for the referenced neutral polymers is also depicted in the figure . for @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six plots . each plot includes several @xmath188 at different @xmath48 . the value of @xmath48 is indicated in the plot or near the corresponding curve . @xmath188 for the neutral polymer system is depicted as reference and marked by `` neutral '' . the unit of ordinate axis is @xmath21.,title="fig:",scaledwidth=40.0% ] for @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six plots . each plot includes several @xmath188 at different @xmath48 . the value of @xmath48 is indicated in the plot or near the corresponding curve . @xmath188 for the neutral polymer system is depicted as reference and marked by `` neutral '' . the unit of ordinate axis is @xmath21.,title="fig:",scaledwidth=40.0% ] for @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six plots . each plot includes several @xmath188 at different @xmath48 . the value of @xmath48 is indicated in the plot or near the corresponding curve . @xmath188 for the neutral polymer system is depicted as reference and marked by `` neutral '' . the unit of ordinate axis is @xmath21.,title="fig:",scaledwidth=40.0% ] for @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six plots . each plot includes several @xmath188 at different @xmath48 . the value of @xmath48 is indicated in the plot or near the corresponding curve . @xmath188 for the neutral polymer system is depicted as reference and marked by `` neutral '' . the unit of ordinate axis is @xmath21.,title="fig:",scaledwidth=40.0% ] for @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six plots . each plot includes several @xmath188 at different @xmath48 . the value of @xmath48 is indicated in the plot or near the corresponding curve . @xmath188 for the neutral polymer system is depicted as reference and marked by `` neutral '' . the unit of ordinate axis is @xmath21.,title="fig:",scaledwidth=40.0% ] for @xmath82 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six plots . each plot includes several @xmath188 at different @xmath48 . the value of @xmath48 is indicated in the plot or near the corresponding curve . @xmath188 for the neutral polymer system is depicted as reference and marked by `` neutral '' . the unit of ordinate axis is @xmath21.,title="fig:",scaledwidth=40.0% ] in a salt - free solution , @xmath188 displays a repulsive interaction , and is stronger than that for the neutral polymers owing to the electrostatic repulsion between monomers . this can be seen by comparing the range of repulsion estimated by equaling @xmath188 to the thermal energy @xmath21 . the range of repulsion is at @xmath189 for the polyelectrolytes and at @xmath190 for the neutral polymers . upon addition of salt , @xmath188 responses in two distinct ways , which depends on the ion size . the first one is applied for small or large tetravalent counterions , such as @xmath82 , 0.5 , or 4.0 . in these cases , @xmath188 is purely repulsive in the course of addition of salt . hence , the monomers on different chains repel between themselves and the system exhibits no like - charge attraction . the second way is happened for an intermediate size such as @xmath75 , 2.0 , or 3.0 . at this moment , @xmath188 displays an attractive well in a mid - salt region . therefore , it appears like - charge attraction between chains . in the first case , the range of repulsion of @xmath188 decreases to a minimum value nearly at @xmath126 . further increase of @xmath48 either leaves @xmath188 unchanged ( for vanishing @xmath8 and for @xmath191 ) or increases the range of repulsion of @xmath188 ( for @xmath146 ) . we will call @xmath188 at @xmath126 `` the curve at equivalence '' , which separates two behaviors of evolution of @xmath188 as @xmath48 is increased or decreased from @xmath57 . we noticed that @xmath188 tends toward the referenced curve of the neutral polymers at high salt concentrations except for the deviation happened at small @xmath3 due to the excluded volume of large ions . for vanishing @xmath8 , the curve at equivalence is very close to the reference curve and @xmath188 is roughly unchanged while @xmath192 . for @xmath146 , the curve at equivalence lies below the referenced curve ; @xmath188 is bounded between the two curves and increases with @xmath48 . and in the second case , @xmath188 displays a potential well located near @xmath193 . it suggests an attraction mediated by a tetravalent counterion . the depth of the well can be more profound than @xmath21 . therefore , a stable bound state can be formed between monomers on different chains and it is a prerequisite to occur salt - induced phase separation . the maximum depth of the well is happened while @xmath48 is @xmath57 . in general , we found that @xmath188 is bounded between the salt - free curve and the curve at equivalence while @xmath194 , and between the latter curve and the referenced neutral curve while @xmath192 . we have shown that the occurrence of the like - charge attraction depends on both salt concentration and ion size . in fig . [ pmf11p4n48_boundary.eps ] we present , in the variable space , the region where @xmath188 shows an attractive well . , @xmath8)-variable space . in region ( i ) , @xmath188 shows an attractive well and in region ( ii ) , it is purely repulsive.,scaledwidth=40.0% ] in the figure , the data points ( symbolized by star ) denote the salt concentrations for a given @xmath8 , at which a purely repulsive @xmath188 turns to show an attractive well and _ vice versa _ , and the closed solid curve is sketched to help readers to separate the two regions . it clearly shows that an effective attraction appears only when @xmath8 is neither too small nor too big and salt concentration is intermediate around the equivalence concentration . we mention that the occurrence of an attractive region in the potential of mean force is not sufficient to have a phase separation of the system . phase transitions in charged colloid systems have been largely discussed in literatures but there have been relatively few studies of phase transitions in charged chain systems ( see refs . @xcite and @xcite and references therein ) . fig . [ snapshots ] shows snapshots of the simulations for @xmath146 , 1.0 and 4.0 at two salt concentrations @xmath126 and @xmath195 . , ( b ) @xmath75 , and ( c ) @xmath191 , at @xmath126 ( left ) and @xmath195 ( right ) . in the pictures , the bead - spring chains are represented in gray color ( red color online ) and the monovalent counterions in light one ( white color online ) . the tetravalent counterions are represented in dark - colored ( blue - colored online ) particles whereas the coions are not plotted for the clarity of the pictures . in the snapshots , a chain may appear more than one fragment because of periodic boundary condition.,title="fig:",scaledwidth=40.0% ] , ( b ) @xmath75 , and ( c ) @xmath191 , at @xmath126 ( left ) and @xmath195 ( right ) . in the pictures , the bead - spring chains are represented in gray color ( red color online ) and the monovalent counterions in light one ( white color online ) . the tetravalent counterions are represented in dark - colored ( blue - colored online ) particles whereas the coions are not plotted for the clarity of the pictures . in the snapshots , a chain may appear more than one fragment because of periodic boundary condition.,title="fig:",scaledwidth=40.0% ] , ( b ) @xmath75 , and ( c ) @xmath191 , at @xmath126 ( left ) and @xmath195 ( right ) . in the pictures , the bead - spring chains are represented in gray color ( red color online ) and the monovalent counterions in light one ( white color online ) . the tetravalent counterions are represented in dark - colored ( blue - colored online ) particles whereas the coions are not plotted for the clarity of the pictures . in the snapshots , a chain may appear more than one fragment because of periodic boundary condition.,title="fig:",scaledwidth=40.0% ] , ( b ) @xmath75 , and ( c ) @xmath191 , at @xmath126 ( left ) and @xmath195 ( right ) . in the pictures , the bead - spring chains are represented in gray color ( red color online ) and the monovalent counterions in light one ( white color online ) . the tetravalent counterions are represented in dark - colored ( blue - colored online ) particles whereas the coions are not plotted for the clarity of the pictures . in the snapshots , a chain may appear more than one fragment because of periodic boundary condition.,title="fig:",scaledwidth=40.0% ] , ( b ) @xmath75 , and ( c ) @xmath191 , at @xmath126 ( left ) and @xmath195 ( right ) . in the pictures , the bead - spring chains are represented in gray color ( red color online ) and the monovalent counterions in light one ( white color online ) . the tetravalent counterions are represented in dark - colored ( blue - colored online ) particles whereas the coions are not plotted for the clarity of the pictures . in the snapshots , a chain may appear more than one fragment because of periodic boundary condition.,title="fig:",scaledwidth=40.0% ] , ( b ) @xmath75 , and ( c ) @xmath191 , at @xmath126 ( left ) and @xmath195 ( right ) . in the pictures , the bead - spring chains are represented in gray color ( red color online ) and the monovalent counterions in light one ( white color online ) . the tetravalent counterions are represented in dark - colored ( blue - colored online ) particles whereas the coions are not plotted for the clarity of the pictures . in the snapshots , a chain may appear more than one fragment because of periodic boundary condition.,title="fig:",scaledwidth=40.0% ] at @xmath126 , we saw that the polyelectrolytes favor to be apart from each other for both of the cases @xmath146 and @xmath191 . however , delicate difference was witnessed that the previous case displays a single - chain collapse but the latter one does not . for @xmath75 , the chains favor to aggregate together , in accordance with the results of fig . [ pmf11p4n48txxx ] which show the occurrence of an attraction between the negatively - charged polymers . this attraction is mediated by counterions , mainly the tetravalent counterions , since almost all the tetravalent counterions are condensed on the polyelectrolytes . we point out that the condensation of tetravalent counterion is not a sufficient condition to occur chain aggregation . it can be seen from the case @xmath146 in which most of the tetravalent counterions are condensed but the chains does not aggregate . at the high salt concentration @xmath195 , the attraction between chains is reduced , in comparison with the cases at @xmath126 . the chains segregate or show tendency of segregation from each other . in the simulations , we observed that while @xmath192 , only part of the tetravalent counterions can condense onto the chains . in this section , we investigate the charge distribution around a polyelectrolyte . it enables to understand how ions distribute around a chain and provides useful information for theorists to develop models . we define a worm - like tube of radius @xmath3 around a chain to be a union of @xmath44 spheres of radius @xmath3 centered at each monomer on the chain . we calculated the averaged total charge inside a worm - like tube , including the charge of the ions and the one of the monomers . this quantity is called `` the integrated charge distribution '' and is denoted by @xmath196 . we show in fig . [ iondist ] @xmath196 at various @xmath48 for @xmath8 ranged between 0.0 and 4.0 . at different @xmath48 for @xmath8=0.0 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six figures . the line arts used for different @xmath48 are indicated at the right - bottom side of each figure . , title="fig:",scaledwidth=40.0% ] at different @xmath48 for @xmath8=0.0 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six figures . the line arts used for different @xmath48 are indicated at the right - bottom side of each figure . , title="fig:",scaledwidth=40.0% ] at different @xmath48 for @xmath8=0.0 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six figures . the line arts used for different @xmath48 are indicated at the right - bottom side of each figure . , title="fig:",scaledwidth=40.0% ] at different @xmath48 for @xmath8=0.0 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six figures . the line arts used for different @xmath48 are indicated at the right - bottom side of each figure . , title="fig:",scaledwidth=40.0% ] at different @xmath48 for @xmath8=0.0 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six figures . the line arts used for different @xmath48 are indicated at the right - bottom side of each figure . , title="fig:",scaledwidth=40.0% ] at different @xmath48 for @xmath8=0.0 , 0.5 , 1.0 , 2.0 , 3.0 and 4.0 , shown , in orders , from top to bottom and from left to right , in six figures . the line arts used for different @xmath48 are indicated at the right - bottom side of each figure . , title="fig:",scaledwidth=40.0% ] focus firstly on the case with vanishing @xmath8 . in the region @xmath197 , @xmath196 attains a value equal to the bare charge of a chain , @xmath198 . a stair - like behavior then appears for @xmath196 in the region @xmath199 while @xmath200 . outside the region , @xmath196 monotonically increases and tends toward zero at large @xmath3 due to electroneutrality . since @xmath201 is the depletion zone for the particles other than tetravalent counterions , the behavior of @xmath196 in this region is completely determined by the condensation of tetravalent counterions . the stair - like behavior indicates a _ perfect _ condensation on the surface of the chain , owing to the vanishing size of the tetravalent counterions . for instance , at @xmath202 , we have 24 tetravalent counterions in the simulation box and , therefore , 6 condensed tetravalent counterions on a chain , in average . a perfect condensation will lead to a value @xmath203 for @xmath196 and this is exactly what we have observed in the figure . at the equivalence concentration , @xmath196 is zero while @xmath204 . if @xmath48 is further increased ( @xmath192 ) , @xmath196 reveals a positive peak in the region @xmath205 , which indicates an overcharging , and , then , rapidly decays to zero as @xmath206 . the hight of the peak suggests that only a portion of the tetravalent counterions are condensed on the chain . for the case @xmath146 , the region in which only the tetravalent counterions can intervene is narrowed to @xmath207 . in the region , the step of the stair - like @xmath196 becomes tilted while @xmath194 . if @xmath192 , @xmath196 displays a positive peak , which is more pronounced than that for vanishing @xmath8 . for the case @xmath75 , an oscillatory behavior occurs at high @xmath48 . similar phenomena have been observed in other systems such as charged plate @xcite , colloid system @xcite , and rigid polyelectrolyte @xcite . this oscillatory behavior indicates a multi - layer organization of ions : in the close neighborhood of a polyelectrolyte , tetravalent counterions condense and form a positively - charged layer which overcharges the chain ; exterior to this layer , anionic particles intervene and form a negatively - charged layer which overcompensates the charge inside and , hence , the integrated charge becomes negative , and so forth . notice that the depletion zone around a chain for tetravalent counterions is @xmath208 . it becomes larger than that for the other ions while @xmath209 . in these cases , @xmath196 decreases with increasing @xmath48 in the depletion zone , but shows completely opposed behavior just outside the zone apparently due to the condensation of tetravalent counterions . at very high @xmath48 ( for instance , at @xmath210 for @xmath211 ) , @xmath196 in the zone can even attain a value smaller than the chain bare charge . in this case , negative particles are the major composition of the first ion layer around a chain , which is counterintuitive . the following three observations are worth to be noticed . ( 1 ) the salt concentration at which @xmath196 becomes ( roughly ) zero in the region @xmath212 decreases with increasing @xmath8 . ( 2 ) the oscillatory behavior appears more markedly for large @xmath8 than for small one . ( 3 ) at a given @xmath48 , the larger the @xmath8 , the higher the maximum peak of @xmath196 will be . it is tempting to consider that a polyelectrolyte and the ions condensed on the chain form a complex object . under some conditions , the effective charge of the complex object reverses its sign . this phenomenon is called charge inversion @xcite . _ intended to link the phenomenon of charge inversion with the reentrant condensation @xcite . they proposed that zero effective charge and charge inversion of chains are , respectively , the cause for the condensation and for the redissolution of polyelectrolytes . their theory necessitates further verification either by experiments or by simulations . in the rest of this section , we try to calculate the effective charge of a polyelectrolyte , in hope of being able to clarify the relationship between the charge inversion and the reentrant condensation . we simply define the complex to be the set of particles lying within a worm - like tube of radius @xmath213 around the chain . this definition has been utilized in many places , for example , in refs . @xcite and @xcite . usually , @xmath213 is chosen to be the distance at which the electrostatic attraction between a monovalent counterion and a monomer is equal to the kinetic energy @xmath214 ; in consequence , the counterion inside the tube of radius @xmath213 can not escape from the chain . we , thus , consider the total charge @xmath215 inside the tube as the effective charge of a chain . in our study , @xmath213 is equal to 2 . the results are presented in fig . [ iondist_p4n48_rc200.eps](a ) . as a function of @xmath48 with @xmath216 , and ( b ) @xmath217 as a function of @xmath48 , for different @xmath8 . the symbol used to denote each @xmath8 is indicated in the figures . , title="fig:",scaledwidth=40.0% ] + ( a ) + as a function of @xmath48 with @xmath216 , and ( b ) @xmath217 as a function of @xmath48 , for different @xmath8 . the symbol used to denote each @xmath8 is indicated in the figures . , title="fig:",scaledwidth=40.0% ] + ( b ) two behaviors were observed : ( 1 ) for @xmath78 and 4.0 , @xmath215 decreases with increasing @xmath48 , ( 2 ) for @xmath218 , @xmath215 increases basically with @xmath48 . notice that @xmath215 for @xmath160 or 2.0 attains a large positive value in the mid - salt region . hence , @xmath188 should show strong repulsion in this region , according to the theory of nguyen _ et al_. but , in fact , we observed a strong attractive well for @xmath188 and the chains were observed to aggregate together . therefore , the results do not support the theory . we notice that it is too simple to regard @xmath215 as the effective chain charge , because ion valence is not considered in the definition . for example , the suitable @xmath213 for tetravalent counterions should be 8.0 , based upon the same argument to choice @xmath213 for monovalent counterions . in addition , @xmath216 becomes too small while @xmath219 , since the depletion zone around a chain for the tetravalent counterions is bigger than the tube . therefore , we redefine @xmath220 as the novel condensation region for tetravalent counterions , whereas keeping @xmath216 for monovalent counterions and for monomers on the other chains . considering that the condensation of coions is mainly mediated by tetravalent counterions , we choose @xmath220 for coions . the charges inside the different tubs are computed and their sum @xmath217 is then regarded as an alternative definition for the effective chain charge . the results are presented in fig . [ iondist_p4n48_rc200.eps](b ) . we remark that @xmath217 looks quite differently from @xmath215 . it shows a positive peak in the mid - salt region for @xmath75 , 1.4 , and 2.0 but retains negative value for the other @xmath8 . the obtained results suggest a disconnection between neutralization of chain charge and condensation of polyelectrolytes . our simulating system in the high - salt region reveals a picture more close to the one described by solis @xcite : although _ locally _ overcompensated by condensed counterions , the net charge of a redissolved chain can be positive or negative due to the association of coions . we must point out that the effective chain charge interpreted by @xmath215 or by @xmath217 in this section can be only served as a reference because both @xmath215 and @xmath217 are sensitive to the choice of @xmath213 . a rigorous way to study it , which can circumvent the ambiguity , is to study the electrophoretic mobility of polyelectrolytes in salt solutions . this part of research is currently undergoing . we have performed molecular dynamics simulations in canonical ensemble to investigate the properties of highly - charged polyelectrolytes in tetravalent salt solutions , including chain morphology , single - chain structure factor , swelling exponent , persistence length , like - charge attraction between chains and integrated charge distribution around a polyelectrolyte . we have explored a wide range of salt concentration and studied the effect of size of tetravalent counterions on the above properties . we have found that a polyelectrolyte exhibits , in turns , two conformational transitions upon addition of salt : a collapsed transition and a reexpanding transition , manifested by a v - shaped curve of the square of radius of gyration @xmath47 against salt concentration @xmath48 . the systems at different monomer concentration @xmath53 show similar curves . the curves overlap each other in the chain - reexpansion region . in the chain - collapsed region , we have also observed that @xmath60 against @xmath59 lie over each other for different @xmath53 , where @xmath62 and @xmath221 weakly depends on @xmath53 . these findings are useful in understanding the phase boundaries of the condensation of polyelectrolytes : on the upper boundary , @xmath48 is roughly independent of @xmath53 and on the lower one , @xmath48 depends linearly on @xmath53 . the study of chain morphology shows that for the cases with @xmath8 ( the size of the tetravalent counterions ) around 1.0 , the chains reveal compact and prolate structures in the vicinity of the equivalence concentration @xmath57 . in a low - salt region and in a high - salt region , the chains show extended structures . at a very high salt concentration , the chains behave similarly to neutral polymers . for the cases with large @xmath8 , the chains are elongated and do not display a collapsed - transition in the mid - salt region . for the case with vanishing @xmath8 , the behavior of the chains is analogous to that of neutral polymers as @xmath222 . noticeably , our results show that @xmath75 is the optimal condition to pack polyelectrolytes into the smallest volume . we have shown that the single - chain structure factor @xmath122 of the multi - chain system in salt solutions follows the guinier function @xmath129 while @xmath223 . @xmath122 then exhibits a power - law - like behavior @xmath224 in the region @xmath134 , suggesting that the scaling law , @xmath225 , is held for polyelectrolytes in salt solutions . the swelling exponent @xmath139 attains a minimum value near @xmath57 and behaves in analogy of the morphological quantities . it shows once more that the chains are mostly collapsed while @xmath8 is compatible with 1.0 . our simulations incorporated explicitly the salt ions and , therefore , were able to study the negative regime of the electrostatic persistence length @xmath157 . this regime appears simultaneously with the mechanical instability of the polyelectrolytes induced by salt . we have examined the theory of ariel and andelman @xcite and found that it fails to predict @xmath157 in the _ so - claimed _ valid region @xmath226 but describes @xmath157 well in the unexpected region @xmath227 . since the osf theory is valid in the latter region , we conjectured that the theory of ariel and andelman is applicable in the latter region . this part deserves a detail investigation in the future . we have studied like - charge attraction between polyelectrolytes by calculating the potential of mean force @xmath188 between monomers on different chains . the results show that it depends strongly on salt concentration and ion size . an attractive region appears in @xmath188 only when the salt concentration is intermediate and the tetravalent counterions possess a size compatible with the monomers . the most profound attractive well is happened while @xmath75 and @xmath228 . in the final part , the integrated charge distribution @xmath196 around a polyelectrolyte was investigated . we have found that the chain charge can be locally overcompensated by condensed ions . @xmath196 displays an oscillatory behavior at high salt concentrations while @xmath80 . the oscillatory behavior results from a multi - layer organization of ions around the chain , which is in agreement with the picture predicted by solis @xcite . we have used two different definitions to calculate the effective chain charge . the results suggest that charge inversion is not a necessary condition to occur reentrant condensation . recently wen and tang have provided evidence against the connection between charge inversion and resolubilization of polyelectrolyte bundles @xcite . a more rigorous study by electrophoresis performed by molecular dynamics simulations is undergoing . the author acknowledges c .- l . lee , y .- f . wei , and n. pardillos for their comments . this material is based upon work supported by the national science council , the republic of china , under the contract no . nsc 94 - 2112-m-007 - 023 . most of the simulations were running at the national center for high - performance computing . the author expresses his gratitude to the members and the staffs of the council and the center .

inflation @xcite is the leading paradigm for the very early universe cosmology . inflation has been proposed to explain the horizon , flatness and monopole problems in the standard hot big bang cosmology , and almost all the predictions of the simplest inflation model have now been tested . the observational tests of inflation includes * coherent and nearly scale invariant power spectrum of density perturbations . the power spectrum of the simplest slow roll inflation is @xcite @xmath0 * a small tilt of the scalar power spectrum . @xmath1 where @xmath2 is the slow roll parameter defined from expansion . now @xmath3 is ruled out under the assumptions of simplest inflation models . * nearly gaussian density fluctuations . the non - gaussianities of the density fluctuations are tightly constrained at @xmath4 for the local shape and equilateral shape non - gaussianities respectively . those numbers indicate that , non - gaussian components of the primordial fluctuations , even if exist , have to be at least 3@xmath54 orders of magnitudes smaller than the gaussian component . * gravitational waves . the recent bicep2 experiment reports an over @xmath6 detection of gravitational waves @xcite , with tensor to scalar ratio @xmath7 this corresponds to a gravitational wave fluctuation amplitude @xmath8 despite of the great success of inflation , there are still a few outstanding challenges for theorists and experimentalists . on the theoretical side , large field inflation is now favored . however , large field inflation is hard to construct from the effective field theory , and stringy uv completion points of view . the uv completion of inflation has long suffered from an @xmath9-problem @xcite , in which the mass of the inflaton is theoretically too large to allow enough e - folds of inflation . however , with the current data , a more serious @xmath10-problem emerges the observed energy scale of inflation is too high for an effective field theory or stringy model building to be under control . for single field slow roll inflation , at every e - fold , the inflaton rolls a distance of order @xmath11 . in perturbative string theory , this field motion per e - fold is comparable with , or greater than the string scale @xmath12 . as a result , one can no longer safely globally expand the inflaton field and ignore non - renormalizable terms . more discussions and a local reconstruction of the inflationary potential can be found in @xcite on the observational side , there is yet another ( and maybe the last in the foreseeable future , unless nature is so kind as to imprint other relics on the cmb sky or in the large scale structure ) test for inflation which is possible in light of bicep2 , but not yet achieved the tilt of the tensor power spectrum . the simplest inflation models predict a consistency relation between @xmath13 and @xmath14 as @xmath15 currently the data has not been good enough to test @xmath13 precisely . however , there are a lot of ongoing and upcoming experiments in the near future @xcite , measuring @xmath14 at different scales , which provides a possibility for a precise measurement of @xmath13 . in this paper , we shall explore the possibility of blue @xmath13 . in section [ sec : hints - blue - tensor ] , the bound on @xmath13 is derived the bicep2 @xcite and polarbear @xcite data . the string gas cosmology , null energy condition ( nec ) violating inflation , and general initial condition are addressed in sections [ sec : string - gas - cosmology ] , [ sec : infl - viol - nec ] and [ sec : infl - gener - init ] respectively . in section [ sec : other - possibilities ] , a few other possibilities are discussed , including external sources for tensor modes , modified gravity and matter bounce . we conclude in section [ sec : concl - disc ] . in this section , we show the constraints on @xmath14 and @xmath13 from bicep2 , _ planck _ and polarbear data . the public codes cosmomc @xcite and class @xcite are used in the calculation . here we vary @xmath14 and @xmath13 in the boltzmann codes , and other cosmological parameters @xmath16 are set at the best fit value from planck results @xcite where not explicitly mentioned . the relation between @xmath14 and @xmath13 is illustrated in fig . [ fig : rnt_sample ] . at @xmath17 , the flat tensor spectrum gives a good fit to the data , which is consistent with the result of the bicep2 @xcite . with the fixed ratio @xmath14 at pivot scale @xmath18 , by increasing or decreasing the tilt @xmath13 by @xmath19 , it starts to be ruled out by the current data . if we consider smaller tensor - to - scalar ratio , indicated from the right panel of fig . [ fig : rnt_sample ] , a blue tensor spectrum is required by the bicep2 data . one can observe from the right panel of fig . [ fig : rnt_sample ] that very blue @xmath13 fits all data points within @xmath20 . this is different from the @xmath21 case , where some data points are too high and may be considered as outliers . a more precise correlation of @xmath13 and @xmath14 is illustrated in fig . [ fig : rnt_contour ] , where the pivot scale at @xmath22 is the same as the analysis of bicep2 paper @xcite . power spectrum with different @xmath23 and @xmath13 . left panel : @xmath23 is fixed and variation of @xmath13 is illustrated . right panel : some sets of @xmath23 and @xmath13 are chosen to get good fit against data . in both panels @xmath14 is calculated at pivot scale @xmath24 mpc@xmath25 . ] power spectrum with different @xmath23 and @xmath13 . left panel : @xmath23 is fixed and variation of @xmath13 is illustrated . right panel : some sets of @xmath23 and @xmath13 are chosen to get good fit against data . in both panels @xmath14 is calculated at pivot scale @xmath24 mpc@xmath25 . ] vs @xmath13 to fit against the data . left panel : fit the data of the bicep2 and polarbear . right panel : fit the first five data bins of bicep2 and treat the others as upper bound . note that the strong correlation between @xmath26 and @xmath23 is an artifact from the choice of pivot scale . we have chosen @xmath27 to match with bicep2 conventions . however , the bicep2 experiments measures @xmath28 . thus given a similar tensor spectrum at @xmath28 , the large tilt modifies the @xmath23 value significantly . ] vs @xmath13 to fit against the data . left panel : fit the data of the bicep2 and polarbear . right panel : fit the first five data bins of bicep2 and treat the others as upper bound . note that the strong correlation between @xmath26 and @xmath23 is an artifact from the choice of pivot scale . we have chosen @xmath27 to match with bicep2 conventions . however , the bicep2 experiments measures @xmath28 . thus given a similar tensor spectrum at @xmath28 , the large tilt modifies the @xmath23 value significantly . ] however , as one can observe from fig . [ fig : rnt_contour ] , @xmath14 and @xmath13 are strongly correlated when the pivot scale is chosen at @xmath22 . this makes detailed analysis inconvenient . the strong correlation comes from the fact that the bicep2 experiment is actually not observing the b - mode signals at their pivot scale @xmath22 , but instead the observation is made at about @xmath29 . this does not matter for the bicep2 data analysis itself because @xmath21 is chosen in their data analysis . however , in our case , when @xmath13 is allowed as a free parameter , we had better pay more attention to the pivot scale . in the remainder of this section , we shall choose the pivot scale at @xmath29 and calculate @xmath14 at this pivot scale . in table [ tab : chi2 ] , the best fit parameters and their likelihood are summarized . one can observe that from the bicep2 data only , including @xmath13 improves @xmath30 by 3.161 . considering that one additional parameter is induced , it is useful compute the akaike information criterion ( aic ) or the bayesian information criterion ( bic ) . the bicep2 e - mode polarization , the wmap polarization and the high @xmath31 part of the planck temperature data has a large amount of data points which are not sensitive to the change of @xmath13 . thus we choose aic to compare between models . ] to see how much improvements @xmath13 would bring . aic can be calculated as @xmath32 where @xmath33 is the number of model parameters . thus including @xmath13 results in @xmath34 . the number not very significant , though showing hint that blue @xmath13 would fit data slightly better . to explore the goodness - of - the - fit , we can also look at the relative @xmath30 and p - value of the bicep2 bb correlation . including bicep2 bb correlation and polarbear , the relative @xmath35 , p - value is @xmath36 in the case of @xmath13 as a free parameter and @xmath37 , p - value is @xmath38 for @xmath21 . both of @xmath39 are close to @xmath40 meaning both of them are good fit , but the one with the parameter @xmath13 is slightly better . the improvement mainly comes from the small scale excess of the b - mode power spectrum . keck ( preliminary ) cross correlation @xcite . however , in the bicep2 @xmath41 keck ( preliminary ) cross correlation , the large scale data points are lower than the theoretical prediction , again showing some evidence of blue @xmath13 . ] however , the significance of blue @xmath13 increase dramatically once planck data is considered . after allowing to vary @xmath13 , @xmath30 is improved by @xmath42 for this single parameter . namely , the fitting of @xmath43 part of planck temperature data is improved by @xmath44 . the fitting of @xmath45 part of planck temperature data is improved by @xmath46 . the fitting of bicep2 data is improved by 4.945 . on the other hand , blue @xmath13 actually fit slightly worse for wmap polarization , by @xmath47 . as a result , one can calculate that @xmath48 , showing strong evidence of blue @xmath13 . it is straightforward to understand why inclusion of planck data strongly prefers blue @xmath13 . note that among the likelihoods listed in table [ tab : chi2 ] , the most significant improvement comes from the @xmath43 part of planck temperature data . this is because , there is already an anomalous temperature power deficit in the small @xmath31 part of the planck data , by an amount of @xmath49 , at the statistical significance of @xmath50 . on the other hand , the existence of tensor mode , when assuming @xmath21 , adds to the theoretically predicted temperature power by another @xmath49 , which fits data even worse . by inclusion of @xmath13 , the above tension between planck and bicep is reconciled . as one can compare to the last column of table [ tab : chi2 ] , after adding blue @xmath13 , the likelihood of planck low @xmath31 , planck high @xmath31 and wmap polarization are all comparable with the best fit values before bicep2 . thus the tension between planck and bicep2 can be reconciled by blue @xmath13 . to further get constraints of @xmath14 and @xmath13 , we run mcmc chains to get samples according to the likelihood code of bicep2 , planck and wmap polarization . the @xmath51 and @xmath20 contours for bicep2 only ( 9 bins ) is plotted in fig . [ fig : b9plot ] , and bicep2+planck+wp is plotted in fig . [ fig : b9pwplot ] . again one observes that blue @xmath13 is preferred . the mean value and standard derivation of @xmath14 and @xmath13 from bicep2 ( 9 bins ) is @xmath52 the constraints from bicep2 ( 9 bins ) + planck ( 2013 ) + wmap polarization is @xmath53 on the other hand , it is worth noticing that if one only take the first 5 bins from bicep2 data , there is no evidence of blue @xmath13 . this is plotted in fig . [ fig : b5plot ] for comparison purpose . because as we have discussed , the two major sources of blue @xmath13 are the observed small scale excess of the b - mode power spectrum of bicep2 , and the observed large scale suppression of the temperature spectrum of planck . none of those are present in the first 5 bins of bicep2 data . the constraints from those 5 bins are @xmath54 .[tab : chi2 ] the best fit values for @xmath14 and @xmath13 , likelihood and aic . the 5 columns corresponds to : [ column 1 ] bicep2 only ( 9 bins ) with varying @xmath14 only , [ column 2 ] bicep2 only ( 9 bins ) with varying @xmath14 and @xmath13 , [ column 3 ] bicep2 ( 9 bins ) + planck ( 2013 ) + wmap polarization with varying @xmath14 only , [ column 4 ] bicep2 ( 9 bins ) + planck ( 2013 ) + wmap polarization with varying @xmath14 and @xmath13 , and [ column 5 ] planck ( 2013 ) + wmap polarization with @xmath55 ( for comparison purpose ) , respectively . note that in the first two columns ( bicep2 data only ) the other cosmological parameters @xmath16 are fixed because bicep2 alone does not provide reasonable constraints on those parameters . in the minimization of the last two columns ( combining with planck and wp ) those additional cosmological parameters and nuisance parameters are also varying . [ cols="^,^,^,^,^,^",options="header " , ] the @xmath56 contour ( left panel ) , and the likelihood for @xmath14 ( middle panel ) and @xmath13 ( right panel ) from bicep2 ( 9 bins ) . ] the @xmath56 contour ( left panel ) , and the likelihood for @xmath14 ( middle panel ) and @xmath13 ( right panel ) from bicep2 ( 9 bins ) + planck ( 2013 ) + wmap polarization . ] the @xmath56 contour ( left panel ) , and the likelihood for @xmath14 ( middle panel ) and @xmath13 ( right panel ) from bicep2 ( 5 bins ) . ] note that when @xmath57 , the primordial tensor contribution to the b - mode polarization becomes more important than lensing signal even at small scales ( @xmath58 ) . in this case , the polarbear detection of lensing b - mode " becomes a constraint on blue @xmath13 . the @xmath51 and @xmath20 exclusion curves from polarbear is plotted in fig . [ fig : combined ] . on the other hand , the cross - correlation - detection of b - mode from sptpol does not put constraint on @xmath13 , because the primordial b - mode does not cross correlate with the density perturbation in the simplest inflation models . another rather theoretical constraint for blue @xmath13 is that , very blue tensor tilt can not last long . assuming the running of @xmath13 is not significant , then for @xmath59 , it takes about 23 e - folds to bring the tensor mode to be non - perturbative ( @xmath60 ) . those non - perturbative tensor modes forms primordial black holes ( pbh ) , the induced scalar perturbation also becomes of order one , which forms primordial black holes . ] , which is constrained from current observations . for @xmath61 , it takes about 12 e - folds . thus assuming @xmath62 we get @xmath63 for 60 e - folds of inflation and @xmath64 for 50 e - folds of inflation . those constraints are plotted in fig . [ fig : combined ] . nevertheless , if @xmath13 is large , @xmath65 may be large as well . if @xmath65 is negative @xcite , it relaxes the above constraint . finally , although the current bound on @xmath13 is far from testing the consistency relation of inflation , it is already informative to disfavor scenarios with very blue or red spectrum . for example , a sharp pulse of gravitational waves with rapid decaying tail towards both ends ( @xmath66 or @xmath67 ) are disfavored by the current data . combined constraints on @xmath14 and @xmath13 . the blue contours are the @xmath51 and @xmath20 constraints from bicep2 ( 9 bins ) . the red contours are the constraints from bicep2 + planck ( 2013 ) + wmap polarization . the blue shaded region are the @xmath51 and @xmath20 exclusion curves from polarbear . the gray shaded region is the exclusion curves from pbh constraints , assuming the blue @xmath13 is also present in smaller scales than cmb observations.,scaledwidth=60.0% ] before a survey of inflationary possibilities for the blue tensor spectra , let us mention alternatives to inflation . this is because the first prediction of a slightly blue @xmath13 comes from the string gas cosmology @xcite ( see @xcite for a recent discussion in light of bicep2 data ) .. this is consistent with our @xmath13 bound if all the bicep2 data points are fitted . however , the amplitude of the tensor mode is exponentially suppressed on cosmological scales , which can not be @xmath68 at @xmath69 . there is a similar issue for the bouncing galileons @xcite . also , there is recently another model with blue tensor tilt derived from the hagedorn phase @xcite . ] on the contrary , for the other models that we shall show below , though they have the possibility to tune the tensor spectrum to be blue , the blueness is not a firm prediction . in string gas cosmology , the universe was in a string hagedorn phase before expansion starts . it is conjectured that due to t - duality , and the huge specific heat in the string hagedorn phase , the universe should stay at a nearly constant temperature ( hagedorn temperature @xmath70 ) for a long time , until the string winding modes decays and allow the expansion of the universe ( which is also a possible explanation of three large spatial dimensions ) . string gas cosmology . the universe starts from a string hagedorn phase and a phase transition brings the universe into radiation dominated.,scaledwidth=80.0% ] the density and tensor fluctuations in string gas cosmology are thermal fluctuations . those fluctuations were in causal contact in the string hagedorn phase , and get frozen on super - hubble scales because of the rapid shrinking of the hubble horizon ( fig . [ fig : sgc ] ) . the scalar and tensor fluctuations can be calculated as @xcite @xmath71 where @xmath72 is the newtonian potential . the scalar spectral index is @xmath73 from the string gas picture , one expects that @xmath74 is an decreasing function of @xmath33 and thus string gas cosmology has red scalar spectrum . the tensor power spectrum is @xmath75~.\end{aligned}\ ] ] the tensor to scalar ratio is @xmath76~.\end{aligned}\ ] ] on the other hand , the tensor to scalar ratio measured by bicep2 is between @xmath77 and the comoving curvature perturbation @xmath78 . this results in a conversion factor . on super - hubble scales , @xmath79 where @xmath80 is the comoving hubble parameter and prime denotes derivative with respect to the comoving time . assuming the hagedorn phase is fast enough to the radiation dominated era , the factor can be calculated as @xmath81 thus the tensor to scalar ratio using @xmath78 is @xmath82~.\end{aligned}\ ] ] inserting the cobe normalization @xmath83 and the bicep2 central value @xmath68 , the string scale is derived as @xmath84 . the tensor spectral index is @xmath85 thus string gas cosmology predicts blue tensor spectra . however , note that is suppressed by slow roll parameters . thus it would not explain the hint of order one @xmath13 shown in data . it is also worth to mention that string gas cosmology produces highly gaussian density perturbations ( unless the string scale is near tev scale , which is not preferred from current data ) . the @xmath86 estimator for string gas cosmology can be calculated as @xcite @xmath87 where @xmath88 corresponds to the scale of our present observable universe . with the detection of @xmath14 , the string scale is known in string gas cosmology . as a result , @xmath89 . this is orders - of - magnitude smaller than the observational bound . as we shall see , the inflationary candidates with blue @xmath13 , as far as we consider , produce considerable amount of non - gaussianities . thus in case a blue tensor tilt is detected , non - gaussianity should be the next test ( and maybe already much better constrained than now by the time of @xmath13 measurement ) to distinguish between string gas and inflationary models . in standard einstein - hilbert gravity , the amplitude of the tensor modes is determined by the energy scale of inflation . and thus for the case of blue @xmath13 , super inflation is needed , with null energy condition ( nec ) violation @xcite . it has been debated for a long time that if the cosmological ( inflation , alternatives or dark energy ) background violates nec , the scalar sector of the inflationary perturbations should become a ghost . however , this situation has been changed since galileons are introduced to inflation . here we shall explore the parameter space of the simplest g - inflation scenario , with slight generalization from the explicit model with an exponential potential in @xcite into a general slow roll functional dependence . we start from the action of g - inflation @xmath90 where @xmath91 . here we restrict our attention to the models with @xmath92 a special class of de sitter solution is worked out in @xcite : @xmath93 we shall consider small derivation of the de sitter solution . for this purpose , define small parameters @xmath94 when expanding around the de sitter solution , the @xmath10 , @xmath95 , and @xmath96 parameters should indeed be small . the consistency for the smallness of the @xmath97 parameter shall be checked later . in terms of those slow roll parameters , and near the de sitter solution , the order of magnitude of the following quantities can be estimated : @xmath98 the second order action of the scalar perturbations can be written as @xmath99~,\end{aligned}\ ] ] where with the above slow roll approximation , the @xmath100 and @xmath101 functions can be written as ( for unapproximated definition , see @xcite ) @xmath102 thus the scalar perturbations are stable in general for @xmath103 satisfying the above slow roll conditions . the power spectrum and spectral index of the scalar sector can be calculated as @xmath104 where @xmath105 note that @xmath106 and @xmath107 near the de sitter solution . thus @xmath108 . the gravity sector is not modified . thus the tensor mode has the conventional spectrum @xmath109 the tensor spectral index is @xmath110 and the tensor to scalar ratio in the slow - roll approximation is written as @xmath111 here the tensor tilt can be blue , but the blueness is suppressed by a slow roll parameter . also note that for the tensor mode to be blue , we need @xmath112 . this is indeed possible . however , it is worth noting that from , @xmath113 in other words , near the de sitter solution , the scalar and tensor modes tilt towards the same direction both red spectra or both blue spectra . note that a blue scalar spectrum is not favored by observations . however , in more general cosmological models , for example , with a different number of neutrinos , a blue scalar spectrum is not yet ruled out . on the other hand , we are not sure at this point if the same direction of tilt for scalar and tensor spectra is a general feature of g - inflation ( and generalized galileons @xcite ) , or there exists unexplored models with different tilt of two sectors . also , one has to note the sound speed of the scalar sector of perturbations is @xmath114 this would induce an equilateral non - gaussianity @xmath115 . this is consistent with the current _ planck _ bound , and on the other hand accessible in the future . it is worth mentioning that the small @xmath116 and large @xmath117 are not a coincidence in this particular toy model , but should be a rather model - independent statement ( unless fine tuned ) . to violate nec at the background level while keeping the perturbations stable , the relevant galileons lagrangian at the energy scale of inflation should be highly non - linear . non - linear self - coupling introduces equilateral non - gaussianity . experiments in the near future would reduce the bound for @xmath117 ( or detection ) and examine the possibility of galileons and nec violation . finally , one can check that when @xmath118 , @xmath119 . thus the smallness of @xmath97 is consistent with data . it has been an open question if the inflationary ( scalar and tensor ) perturbations originates from a bunch - davies ( bd ) vacuum , or special care needs to be taken for the non - bd initial conditions . one motivation for choosing non - bd initial conditions , for example , is the transplanckian problem of inflation @xcite ( see @xcite for a recent review ) . inflation requires uv completion . during inflation , the perturbations originates from the uv completion scale , for example , planck scale , before they expand and cross the hubble horizon . note that the planck scale could have been replaced by some lower scales , for example , the string scale . with those energy scales for new physics , we can no longer make sure that the inflationary perturbations were in their lowest energy states before they got stretched by the cosmic expansion . this situation is illustrated in fig . [ fig : trans ] . the transplanckian problem . the inflationary perturbations are initialized on sub - planckian scales ( or above other uv completion scales).,scaledwidth=80.0% ] in the context of tensor perturbations @xcite ( for a related recent study , see also @xcite ) , the action of tensor mode is @xmath120 decomposing the tensor modes by polarization , we have @xmath121~,\end{aligned}\ ] ] as a result , the action contains two copies of modes @xmath122~.\end{aligned}\ ] ] for each mode , one can impose a different initial condition from bd . the @xmath123 ( collectively denote @xmath124 and @xmath125 ) field can be quantized as @xmath126 where @xmath127 the quantization condition requires @xmath128 the lowest energy state has @xmath129 , which corresponds to the bd state . the tensor power spectrum for each polarization mode is @xmath130 @xmath131 ~. \end{aligned}\ ] ] note that the @xmath33-dependence of @xmath132 is determined by high energy physics and is unknown at low energy scales . thus this @xmath33-dependence could in principle tilt the tensor spectrum to the blue end and overwrite the red tilt from decreasing @xmath133 . this requires @xmath134\right\ } > 0~.\end{aligned}\ ] ] and the tensor tilt is @xmath135}{d \ln k}~. \end{aligned}\ ] ] we would like to make a few comments before closing up this section : * the non - bd state of tensor modes also opens up possibilities for small field inflation to be consistent with the observed large tensor to scalar ratio @xmath68 . however , note that a reason is needed for the scalar sector and the tensor sector to achieve different non - bd coefficients . * non - bd initial condition generically introduce non - gaussianities of the folded shape @xcite . the size of the non - gaussianity is proportional to @xmath136 $ ] and peaked at the folded limit @xmath137 . in the squeezed limit @xmath138 , the folded non - gaussianity also blows up more quickly than local shape ( however , in the regime under theoretical control , the squeezed limit remains small @xcite ) . thus the tensor modes generated by non - bd coefficients are non - gaussian . despite of the non - gaussianity in the tensor sector , it is also reasonable to expect a folded non - gaussianity for the scalar sector . because if the non - bd issue exists , it is likely not tensor - only unless a reason is provided there is a huge landscape of inflation models , and also quite a few alternatives to inflation . some of them may also be able to produce blue @xmath13 . we list a few possibilities here . besides gravitational waves from quantum fluctuations during inflation , there is a possibility that the dominant gravitational waves are generated by particles or strings @xcite , or particle states @xcite produced in the period inflation . the kinetic energy of the inflaton @xmath139 dumps to the particles , and it is large enough to produce visible tensor spectrum . the emission rate is related to the square of the coupling constant , i.e , for the particles with energy @xmath140 , @xmath141 . one can derive a bound on @xmath10 from those class of particle production mechanism of sourcing tensor mode . one can estimate the energy density of the classically sourced gravitational waves energy density @xmath142 near horizon crossing : @xmath143 where @xmath144 is the kinetic energy density of the inflaton . on the other hand , near horizon crossing , the physical wave length of the tensor modes is of order hubble scale . thus @xmath145 thus the tensor power spectrum @xmath146 from the observed tensor to scalar ratio @xmath147 we get @xmath148 from cobe normalization , we get @xmath149 below those scales , no classical source could generate primordial tensor modes . in some simple models proposed in @xcite , the produced particles have a time - dependent mass . @xmath150 due to its coupling to the rolling of inflaton . compared with the fact that inflation has the decreasing hubble constant , which predict a red tensor spectrum , the increasing mass will afford an explanation of the blue tensor spectrum . it is also important to note that non - gaussianities are also produced from the particle production process @xcite . the constraints from non - gaussianity of particle production remains to be tested . based on the slow roll expansion @xcite , it is recently pointed out that violation of slow roll can generate blue tensor spectrum @xcite . in @xcite , the tensor spectral index is calculated up to second order in slow roll as @xmath151 thus if the slow roll expansion is extrapolated to @xmath152 , blue tensor spectrum is obtained . for example , the possible blue tensor spectrum in the ultra - slow - roll inflation @xcite is discussed in @xcite . it would be interesting to further examine this possibility without slow roll restrictions , and see if the introduced large running is consistent with data . the tensor modes come from the gravity sector . thus there is no surprise that if gravity is modified , the tensor spectrum could change . among other possibilities , massive gravity is one example to achieve this goal . massive gravity on a time dependent background is a challenge @xcite . nevertheless , there exist viable models @xcite and application to inflation @xcite . we may expect tensor modes with blue spectra in massive gravity : if @xmath153 for the graviton during inflation , the tensor modes which exit the horizon earlier have more time to roll back to the origin of their mass potential , and thus are suppressed more . as a result , a blue tensor tilt can be generated . however , the massive gravity in the late universe does not help for the tensor modes . this is because the mass of the graviton is too small . also , the return - to - horizon effect actually generates a redder spectrum for the gravitational waves @xcite . as another interesting class of possibilities , some inflation models come with space - like condensates of fields . for example , space - like gradient of scalars in solid inflation @xcite and so(3 ) massive gravity @xcite , or space - like vector fields in chromo - natural inflation @xcite . typically , so(3 ) spatial rotational symmetry is imposed to keep isotropy . in this class of models , there can be additional tensor components from decomposition of the so(3 ) group , other than the graviton . those components can also contribute to the b - mode power spectrum . for the additional components , there is no nec type restriction against blue spectra . for example , solid inflation predicts a slightly blue tensor spectrum @xcite . note that the blue tilt in solid inflation is on the one hand suppressed by @xmath10 but on the other hand boosted by sound speed . for @xmath68 one finds @xmath154 . with the assumption of isotropy , scalar modes can not source the tensor modes at the linear level . however , beyond the linear level , tensor modes can be sourced by scalar modes @xcite . those second order effects are typically unobservably small because they are suppressed by the inflationary power spectrum . however , the second order contribution to tensor mode can be boosted by a small sound speed of an isocurvature scalar sector @xcite . the generated gravitational waves are highly non - gaussian . the tilt of tensor spectrum coming from those contributions are not constrained by the null energy condition , thus a blue tilt is possible . as another alternative to inflation scenario , in matter bounce , the universe was in a contracting phase before bouncing back and heating up ( see @xcite for a review ) . in matter bounce , the tensor modes are nearly scale invariant but the amplitudes have been too high . recently , a two - field bounce model is introduced and the tensor modes can be tuned as a parameter @xcite . it is thus interesting to see if this model ( among other possibilities ) fit the current observations and has a potential for blue tensor spectrum . to conclude , we fit the tensor - to - scalar ratio and the tensor spectral tilt with data . the current data is not good enough to test the inflationary consistency relation but nevertheless blue tensor spectra are favored when all 9 bins of bicep2 data , in combination with the polarbear data , are used . from theoretical aspects , string gas cosmology predicts blue tensor spectra . however , the tilt is small , at the same order - of - magnitude of scalar tilt . thus the future experiments targeting to test the inflationary consistency relation can also test this prediction of string gas cosmology . on the other hand , string gas cosmology predicts highly gaussian density and tensor perturbations . this is unlike the other inflationary mechanisms , where a blue tilt also implies non - gaussianities . the simplest model of g - inflation , on the other hand , tilts scalar and tensor power spectra in the same way . thus with a red scalar spectral tilt , the tensor spectra are also red . in the parameter regime where both scale and tensor spectra are blue , the blueness of tensor perturbation is suppressed by slow roll parameter @xmath10 . the equilateral non - gaussianity of g - inflation , at @xmath68 , is about @xmath155 and close to the current observational bound . the non - gaussianities should be a model independent feature for super - inflation type models which generate blue tensor spectrum , because the perturbations have to behave differently from the background to avoid ghosts . generalized initial conditions of inflation is left largely unconstrained . however , the generalized initial conditions are also sources of non - gaussianities . thus non - gaussianities in the tensor sector would be a test of those class of models . inflationary particle production is another possible source of tensor modes . we derived a bound of inflationary hubble scale and energy density for this mechanism to work . non - gaussianities are also present in the case of particle productions . we hope to investigate the particle production mechanism and its relation between the blue tensor spectra in a future work . the current data shows hint of blue @xmath156 . the models with slow roll suppressed @xmath13 would not be enough to explain such a hint . on the other hand , from modified initial conditions , external sources , inflation beyond slow roll , modified gravity and some models of matter bounce , blue and large @xmath13 may be produced . it would be interesting to study those models in more details to see how the fitting of data is improved in those models . * note added : * two related works @xcite appeared on arxiv on the same day as ours . @xcite ( see also @xcite ) overlaps with our section [ sec : hints - blue - tensor ] , and @xcite overlaps with section [ sec : infl - gener - init ] . the data analysis of tensor tilt is also investigated by @xcite , @xcite and @xcite , within a few days before / after our paper . among those papers , @xcite and @xcite reports a nearly zero central value , with small errorbars @xmath157 ( bicep2 only ) and @xmath158 ( bicep2 + planck + wp , with running of scalar spectral index ) . while the following works prefer blue tilt with considerably larger errorbars : @xcite ( bicep2 only ) , our result ( bicep2 , bicep2 + planck + wp , polarbear ) , @xcite ( bicep2 + planck + wp and bicep2 x keck + planck + wp ) and @xcite ( bicep2 + planck + wmap + bao ) . yw is supported by a starting grant of the european research council ( erc stg grant 279617 ) , and the stephen hawking advanced fellowship . this work was undertaken on the cosmos shared memory system at damtp , university of cambridge operated on behalf of the stfc dirac hpc facility . this equipment is funded by bis national e - infrastructure capital grant st / j005673/1 and stfc grants st / h008586/1 , st / k00333x/1 . p. a. r. ade _ et al . _ [ bicep2 collaboration ] , arxiv:1403.3985 [ astro-ph.co ] . m. j. mortonson and u. seljak , arxiv:1405.5857 [ astro-ph.co ] . r. flauger , j. c. hill and d. n. spergel , jcap * 1408 * , 039 ( 2014 ) [ arxiv:1405.7351 [ astro-ph.co ] ] . e. j. copeland , a. r. liddle , d. h. lyth , e. d. stewart and d. wands , phys . d * 49 * , 6410 ( 1994 ) [ astro - ph/9401011 ] . y. -z . ma and y. wang , arxiv:1403.4585 [ astro-ph.co ] . p. a. p. ade et al . , planck 2013 results xxii , arxiv : 1303.5082 a. lewis and s. bridle , phys . d * 66 * , 103511 ( 2002 ) [ astro - ph/0205436 ] . d. blas , j. lesgourgues and t. tram , jcap * 1107 * ( 2011 ) 034 [ arxiv:1104.2933 [ astro-ph.co ] ] . g. hinshaw _ et al . _ [ wmap collaboration ] , astrophys . j. suppl . * 208 * ( 2013 ) 19 [ arxiv:1212.5226 [ astro-ph.co ] ] . a. ashoorioon , k. dimopoulos , m. m. sheikh - jabbari and g. shiu , arxiv:1403.6099 [ hep - th ] . p. a. r. ade _ et al . _ [ planck collaboration ] , arxiv:1303.5075 [ astro-ph.co ] . d. hanson _ et al . _ [ sptpol collaboration ] , phys . lett . * 111 * , 141301 ( 2013 ) [ arxiv:1307.5830 [ astro-ph.co ] ] . y. -f . cai and y. wang , arxiv:1404.6672 [ astro-ph.co ] . r. h. brandenberger and c. vafa , nucl . b * 316 * , 391 ( 1989 ) . a. nayeri , r. h. brandenberger and c. vafa , phys . lett . * 97 * , 021302 ( 2006 ) [ hep - th/0511140 ] . r. h. brandenberger , a. nayeri , s. p. patil and c. vafa , phys . lett . * 98 * , 231302 ( 2007 ) [ hep - th/0604126 ] . r. h. brandenberger , a. nayeri , s. p. patil and c. vafa , int . j. mod . a * 22 * , 3621 ( 2007 ) [ hep - th/0608121 ] . r. h. brandenberger , a. nayeri and s. p. patil , arxiv:1403.4927 [ astro-ph.co ] . j. khoury , b. a. ovrut , p. j. steinhardt and n. turok , phys . d * 64 * , 123522 ( 2001 ) [ hep - th/0103239 ] . t. biswas , t. koivisto and a. mazumdar , jhep * 1408 * , 116 ( 2014 ) [ arxiv:1403.7163 [ hep - th ] ] . l. a. boyle , p. j. steinhardt and n. turok , phys . d * 69 * , 127302 ( 2004 ) [ hep - th/0307170 ] . t. qiu , j. evslin , y. -f . cai , m. li and x. zhang , jcap * 1110 * , 036 ( 2011 ) [ arxiv:1108.0593 [ hep - th ] ] . b. chen , y. wang , w. xue and r. brandenberger , arxiv:0712.2477 [ hep - th ] . y. -s . piao and y. -z . zhang , phys . d * 70 * , 063513 ( 2004 ) [ astro - ph/0401231 ] . t. kobayashi , m. yamaguchi and j. i . yokoyama , phys . lett . * 105 * , 231302 ( 2010 ) [ arxiv:1008.0603 [ hep - th ] ] . c. deffayet , s. deser and g. esposito - farese , phys . d * 80 * , 064015 ( 2009 ) [ arxiv:0906.1967 [ gr - qc ] ] . c. deffayet , x. gao , d. a. steer and g. zahariade , phys . d * 84 * , 064039 ( 2011 ) [ arxiv:1103.3260 [ hep - th ] ] . t. kobayashi , m. yamaguchi and j. i . yokoyama , prog . phys . * 126 * , 511 ( 2011 ) [ arxiv:1105.5723 [ hep - th ] ] . r. h. brandenberger and j. martin , mod . a * 16 * , 999 ( 2001 ) [ astro - ph/0005432 ] . j. martin and r. h. brandenberger , phys . rev . d * 63 * , 123501 ( 2001 ) [ hep - th/0005209 ] . r. h. brandenberger and j. martin , class . grav . * 30 * , 113001 ( 2013 ) [ arxiv:1211.6753 [ astro-ph.co ] ] . t. biswas , r. brandenberger , t. koivisto and a. mazumdar , phys . d * 88 * , no . 2 , 023517 ( 2013 ) [ arxiv:1302.6463 [ astro-ph.co ] ] . a. ashoorioon , k. dimopoulos , m. m. sheikh - jabbari and g. shiu , jcap * 1402 * , 025 ( 2014 ) [ arxiv:1306.4914 [ hep - th ] ] . h. collins , r. holman and t. vardanyan , arxiv:1403.4592 [ hep - th ] . x. chen , m. -x . huang , s. kachru and g. shiu , jcap * 0701 * , 002 ( 2007 ) [ hep - th/0605045 ] . w. xue and b. chen , phys . d * 79 * ( 2009 ) 043518 [ arxiv:0806.4109 [ hep - th ] ] . r. flauger , d. green and r. a. porto , jcap * 1308 * , 032 ( 2013 ) [ arxiv:1303.1430 [ hep - th ] ] . a. aravind , d. lorshbough and s. paban , jhep * 1307 * , 076 ( 2013 ) [ arxiv:1303.1440 [ hep - th ] ] . l. senatore , e. silverstein and m. zaldarriaga , arxiv:1109.0542 [ hep - th ] . j. l. cook and l. sorbo , phys . d * 85 * , 023534 ( 2012 ) [ erratum - ibid . d * 86 * , 069901 ( 2012 ) ] [ arxiv:1109.0022 [ astro-ph.co ] ] . d. carney , w. fischler , e. d. kovetz , d. lorshbough and s. paban , jhep * 1211 * , 042 ( 2012 ) [ arxiv:1209.3848 [ hep - th ] ] . n. barnaby , j. moxon , r. namba , m. peloso , g. shiu and p. zhou , phys . d * 86 * , 103508 ( 2012 ) [ arxiv:1206.6117 [ astro-ph.co ] ] . j. -o . gong , class . grav . * 21 * , 5555 ( 2004 ) [ gr - qc/0408039 ] . j. -o . gong , arxiv:1403.5163 [ astro-ph.co ] . n. c. tsamis and r. p. woodard , phys . d * 69 * , 084005 ( 2004 ) [ astro - ph/0307463 ] . x. chen , h. firouzjahi , m. h. namjoo and m. sasaki , europhys . lett . * 102 * , 59001 ( 2013 ) [ arxiv:1301.5699 [ hep - th ] ] . g. damico , c. de rham , s. dubovsky , g. gabadadze , d. pirtskhalava and a. j. tolley , phys . d * 84 * , 124046 ( 2011 ) [ arxiv:1108.5231 [ hep - th ] ] . a. e. gumrukcuoglu , c. lin and s. mukohyama , jcap * 1111 * , 030 ( 2011 ) [ arxiv:1109.3845 [ hep - th ] ] . a. de felice and s. mukohyama , phys . b * 728 * , 622 ( 2014 ) [ arxiv:1306.5502 [ hep - th ] ] . a. de felice , a. e. gumrukcuoglu and s. mukohyama , phys . d * 88 * , 124006 ( 2013 ) [ arxiv:1309.3162 [ hep - th ] ] . s. dubovsky , r. flauger , a. starobinsky and i. tkachev , phys . d * 81 * , 023523 ( 2010 ) [ arxiv:0907.1658 [ astro-ph.co ] ] . c. lin , arxiv:1307.2574 . a. e. gumrukcuoglu , s. kuroyanagi , c. lin , s. mukohyama and n. tanahashi , class . * 29 * , 235026 ( 2012 ) [ arxiv:1208.5975 [ hep - th ] ] . s. endlich , a. nicolis and j. wang , jcap * 1310 * , 011 ( 2013 ) [ arxiv:1210.0569 [ hep - th ] ] . m. akhshik , r. emami , h. firouzjahi and y. wang , arxiv:1405.4179 [ astro-ph.co ] . p. adshead and m. wyman , phys . lett . * 108 * , 261302 ( 2012 ) [ arxiv:1202.2366 [ hep - th ] ] . x. chen , r. emami , h. firouzjahi and y. wang , jcap * 1408 * , 027 ( 2014 ) [ arxiv:1404.4083 [ astro-ph.co ] ] . s. matarrese , s. mollerach and m. bruni , phys . d * 58 * , 043504 ( 1998 ) [ astro - ph/9707278 ] . m. biagetti , m. fasiello and a. riotto , phys . rev . d * 88 * , no . 10 , 103518 ( 2013 ) [ arxiv:1305.7241 [ astro-ph.co ] ] . r. h. brandenberger , int . j. mod . 01 * , 67 ( 2011 ) [ arxiv:0902.4731 [ hep - th ] ] . y. -f . cai , e. mcdonough , f. duplessis and r. h. brandenberger , jcap * 1310 * , 024 ( 2013 ) [ arxiv:1305.5259 [ hep - th ] ] . m. gerbino , a. marchini , l. pagano , l. salvati , e. di valentino and a. melchiorri , arxiv:1403.5732 [ astro-ph.co ] . antony lewis , http://cosmocoffee.info/viewtopic.php?t=2302 .

diluted magnetic semiconductors ( dms ) @xcite with transition temperatures as high as 150k ( gamnas at @xmath3 mn @xcite ) or even above room temperature ( gamnn , gamnp@xcite ) are attracting much attention lately in part because of possible ` spintronic ' applications @xcite . the prototypical dms material is ga@xmath4mn@xmath5as with the mn ions substitutionally replacing ga at the cation sites . it is widely believed that the ferromagnetism in this material is carrier induced with holes donated by mn ions mediating a ferromagnetic interaction between the randomly localized mn spins @xcite . the coupling of carrier spins ( holes in gamnas ) and localized moments ( mn impurities ) gives rise to the unique magnetic and transport ( as well as optical ) properties in dms . dms transport properties are influenced by the exchange interaction between the carriers and the localized moments as spin fluctuation scattering contributes to the resistivity . the experimentally measured dc - resistivity in dms materials @xcite shows interesting behavior strongly depending on the concentration of the magnetic impurity and temperature . in in@xmath4mn@xmath5as @xcite and ga@xmath4mn@xmath5as @xcite with low mn concentration ( @xmath6 ) only an insulating behavior has been observed in transport measurements . however , near optimal doping @xmath7 , where the highest value of @xmath2 is reported , the non - monotonic behavior ( insulator - metal - insulator ) of the resistivity as a function of temperature is observed . a resistivity peak appears near the critical temperature ( @xmath2 ) , and the resistivity shows metallic behavior ( @xmath8 ) below @xmath2 and insulating behavior ( @xmath9 ) at higher temperatures . the peak has been understood as the critical scattering effects at @xmath2 of spin fluctuations @xcite . in this paper we present theoretical calculations for dms transport properties and study the role of the carrier - spin coupling which are crucial for ferromagnetic properties . transport measurements have proven useful in understanding the physics of the colossal magnetoresistance ( cmr ) manganites where carrier - spin coupling is also crucial @xcite . in our calculation we use a recently developed non - perturbative method , the `` dynamical mean - field theory '' ( dmft),@xcite for calculating the dms transport properties . a non - perturbative method is needed in dms materials because the crucial physics involves bound - state formation and other aspects of intermediate carrier - spin couplings not accessible to perturbative methods . ( we note that the most interesting phenomena in dms involve intermediate couplings and intermediate temperatures . this regime is very difficult to treat by standard analytical or numerical methods . ) the dmft has been recently applied to the dms system to calculate the magnetic transition temperature and the optical conductivity @xcite . dmft is essentially a lattice quantum version of the weiss mean field theory where the appropriate density of states ( including impurity band formation ) along with temporal fluctuations are incorporated within an effective local field theory . an important ingredient of dmft @xcite is that it reduces to the standard rkky physics in the weak - coupling regime and the double exchange physics in the strong coupling regime . our dmft results show interesting dependence of resistivity on carrier - spin coupling ( @xmath10 ) , carrier density ( @xmath11 ) , doping of the magnetic material ( @xmath12 ) , and temperature ( @xmath13 ) revealing key features of the underlying physics . we find that our results show many similarities to colossal magnetoresistance ( cmr ) manganites @xcite , especially , for a large carrier - spin coupling and near half filling of the impurity band ( i.e. in the double exchange regime ) since the spin - disorder scattering dominates in this parameter range . however , the dms dc - resistivity exhibits novel features not found in the cmr in the weak coupling rkky limit ( @xmath14 , where @xmath15 is the band width of the carrier ) . these novel features arise mainly from the bound state formation of the carrier - local spins and carrier occupation of the minority spin band . the formation of the bound state gives rise to an insulating behavior even if the carrier localization effects are not taken into account . experimental observation of our predictions should lead to crucial information about bound state formation and impurity band physics in this problem . even though we focus on iii - v compound based dms such as ga@xmath4mn@xmath5as , the results presented in this paper are general for all dms materials . an important limitation of dmft is that it is a non - perturbative _ local _ theory that can not really incorporate _ spatial _ fluctuations , and therefore resistive charged impurity scattering with its strong momentum dependence is essentially impossible to handle in dmft . we therefore consider the boltzmann transport theory to calculate the resistivity of the dms systems @xcite . the boltzmann theory is used to calculate the charged impurity disorder limited dms resistivity ( whereas the dmft is used for obtaining the spin disorder limited dms resistivity ) . charged impurity disorder arises here both from the ionized mn acceptor in gamnas as well as other charged defects / impurities invariably present in a semiconductor . using relaxation time approximation we assume that the boltzmann resistivity is due to ionized impurities . the carriers ( holes ) are scattered by the screened coulomb potential , which we calculate using the linearized thomas - fermi ( tf ) approximation and the random phase approximation ( rpa ) . we find that the dominant temperature dependence of the resistivity comes from the change in the screening length in the high density metallic samples . in the metallic gamnas dms samples the change in the measured resistivity , when the temperature goes from @xmath2 to zero , is about 20% in good agreement with our calculation . in the ferromagnetic state ( @xmath1 ) , as the temperature decreases from @xmath16 , we find strong temperature dependence of the resistivity arising from the low temperature screening function and the thermal change of the carrier densities in each spin - split subband . when all the carriers are polarized ( this happens in the low density limit ) the screening function is almost independent of temperature and the resistivity is also temperature independent . inclusion of both spin disorder and ionized impurity disorder in dms transport is the important ingredient of our theory . this paper is organized as follows . in sec . ii , we describe our model and our theoretical approach based on the dynamical mean - field theory ( dmft ) . in sec . iii , we study , in detail , the density of states and describe the formation of the spin polarized impurity band . in sec . iv , we provide the results of calculated dms resistivity within dmft . in sec . v we calculate the transport resistivity within the boltzmann transport theory for ionized impurities . in sec . vi , we summarize our qualitative findings and providing a critical discussion of the applicability of our results to dms systems . a brief conclusion is given in sec . our basic model of dms systems is that of magnetic dopants ( `` impurities '' ) interacting through a local exchange coupling with carriers in the host semiconductor material . the generally accepted hamiltonian of the system is given by @xmath17 where @xmath18 describes carrier propagation in the host semiconductor band . for simplicity we consider a host material with a single non - degenerate band . we therefore write @xmath19 \psi _ { \alpha } ( x),\ ] ] where @xmath20 is the spin index and @xmath21 is a random potential arising from non - magnetic defects in the material ( e.g. , as antisite defects , unintentional background charged impurities , etc . ) . the second ( magnetic ) term in eq . ( [ ham ] ) , @xmath22 , describes coupling of the carriers to an array of impurity ( e.g. mn ) spins at positions @xmath23 , @xmath24 \psi_{\beta } ( r_i),\ ] ] where @xmath10 is the local exchange coupling between the spin of the magnetic impurity and the the spins of the semiconductor carriers , @xmath25 is the ( coulombic ) potential arising from the magnetic dopant , @xmath23 are the positions of the magnetic dopants , and @xmath26 is the pauli matrix . here we absorb the magnitude of the impurity spin into the coupling @xmath10 ( which we take to be positive ) , and represent the spin direction by the unit vector @xmath27 . the third term in eq . ( [ ham ] ) , @xmath28 , is the direct mn@xmath29mn short - range antiferromagnetic exchange interaction @xmath30 where @xmath31 is a direct antiferromagnetic exchange coupling between impurity spins . in this section we approximate our model by neglecting the nonmagnetic random potential , @xmath21 , and the direct antiferromagnetic exchange interaction , @xmath31 . lattice defects may be playing an important role in determining magnetic and transport properties of the samples , but we assume here that these defects enter our theory only in determining the basic parameters of the model , namely , the density of magnetically active dopants @xmath32 , the hole density @xmath33 , and perhaps the local effective exchange coupling @xmath10 between the holes and the magnetic impurities , and do not include any defects into our model explicitly . we include charged impurity scattering through the boltzmann equation in section iv of this paper . we believe that the effects of the antiferromagnetic coupling between magnetic impurities are either negligibly small or incorporated into the effective parameters of the model . actually , in the parameter range of interest to us ( @xmath34 ) , where dms ferromagnetism typically occurs , the magnetic impurities are separated from each other by non - magnetic atoms , and this short - range antiferromagnetic interaction , which rapidly decays with the distance , should be negligible . these approximations are nonessential and are done in the spirit of identifying the minimal dms magnetic model of interest . both of these effects , which may be of quantitative importance in some situations , can be included in the theory by adjusting the parameters of the model or perhaps at the cost of introducing more unknown parameters characterizing these interactions . recently several theories for dms ferromagnetism explicitly including spatial disorder effects and antiferromagnetic coupling have been developed @xcite . in our dmft dms model there are two sources of coupling between the carrier and the impurity magnetic moment : a spin - spin coupling ( @xmath10 ) and a potential scattering ( @xmath25 ) . the crucial physical issues are revealed by the consideration of a ferromagnetic state in which all impurity spins @xmath35 are aligned , say , in the @xmath36 direction . then the carriers with spin parallel to @xmath35 feel a potential @xmath37 on each magnetic impurity site and anti - parallel carriers feel a potential @xmath38 . these potentials self - consistently rearrange the electronic structure . the spin - dependent part of this rearrangement provides the energy gain which stabilizes the ferromagnetic state . the key physics issue is , evidently , whether the potential @xmath39 is weak ( so its effect on carriers near the lower band edge is simply a scattering phase shift ) or strong ( so only majority spin or perhaps both species of carriers are confined into spin - polarized impurity bands ) . recent density functional supercell calculations @xcite suggest that in gamnas @xmath37 is close to the critical value for bound state formation for the majority spin systems . unfortunately , the precise effective values of @xmath10 and @xmath25 are typically unknown in a dms system , and may have to be extracted experimentally . we assume that magnetic impurities under consideration enter substitutionally at the cation sites ( _ e . g. _ mn impurities at ga sites ) and the iii@xmath4mn@xmath5v system as a lattice of sites , which are randomly nonmagnetic ( with probability @xmath40 ) or magnetic ( with probability @xmath12 ) , where @xmath12 now indicates the relative concentration ( _ i.e. _ per ga site ) of active mn local moments in iiimnv . if more complete information about mn locations on the gaas lattice becomes available it will be straightforward to incorporate that in the dmft formalism . we now introduce the dmft for the hamiltonian given by eq . ( [ ham ] ) . within the general scheme of the dmft , the local ( momentum independent ) self energy of the system , @xmath41 , can be obtained from the time dependent mean field function . then the single particle green function is approximated by @xmath42 where @xmath43 is the chemical potential . with the local self energy all of the relevant physics may be determined from the local ( momentum - integrated ) green function defined by @xmath44 where @xmath45 is the density of states ( dos ) for the noninteracting system . the information of the lattice geometry is included through the noninteracting dos . in our model @xmath46 is a matrix in spin index and depends on whether one is considering a magnetic ( @xmath47 or non - magnetic ( @xmath48 site . since @xmath46 is a local function , it is the solution of a local problem specified by a mean - field function @xmath49 , which is related to the partition function @xmath50 with action @xmath51 c_{\beta } ( \tau ^{\prime } ) \nonumber \\ & + & \int d\tau \sum_{\alpha,\beta } c_{\alpha } ^{+}(\tau ) \left [ j \hat { \textbf{s } } \cdot \bm{\sigma}_{\alpha \beta } + w \delta_{\alpha,\beta } \right ] c_{\beta } ( \tau)\end{aligned}\ ] ] on the @xmath52 ( magnetic ) site and @xmath53 on the non - magnetic ( @xmath54 ) site . here @xmath55 ( @xmath56 ) is the destruction ( creation ) operator of a fermion in the spin state @xmath20 and at time @xmath57 . @xmath58 plays the role of the weiss mean field ( bare green function for the local effective action @xmath59 ) and is a function of time . @xmath60 depends only on frequency and is therefore the solution of a single - site problem . the local green function @xmath60 of the effective single - site problem is solely determined by the partition function , @xmath61 , namely , @xmath62 which is identical to the local green function computed by performing the momentum integral using the same self energy . then the self - energy is defined by @xmath63 the @xmath52-site mean - field function @xmath64 can be written as @xmath65 with @xmath66 the magnetization direction and @xmath67 vanishing in the paramagnetic state ( @xmath0 ) . since the spin axis is chosen parallel to @xmath68 @xmath69 becomes a diagonal matrix with components parallel ( @xmath70 ) and antiparallel ( @xmath71 ) to @xmath68 . the form of the dispersion given in full hamiltonian eq . ( 1 ) applies only near the band edges . it is necessary for the method to impose a momentum cutoff , arising physically from the carrier band - width . we impose the cutoff by assuming a semicircular density of states @xmath72 with @xmath73 . the parameter @xmath15 is chosen to correctly reproduce the band edge density of states . other choices of upper cutoff would lead to numerically similar results . this choice of cutoff corresponds to a bethe lattice in infinite dimensions . other ( perhaps more realistic ) choices for the density of states would give results qualitatively similar to our results since the band edge density of states has the correct physical behavior in our model . for this @xmath74 the self consistent equation for @xmath75 obeys the equation @xmath76^{-1 } \right \rangle,\end{aligned}\ ] ] where the angular brackets denote averages performed in the ensemble defined by the appropriate @xmath61 , i.e. , @xmath77 , with @xmath78 . with these mean field functions the local green function can be written as @xmath79 the self energies are evaluated using eq . ( [ self ] ) and the full green function from eq . ( [ green ] ) . physical observables can be obtained from the full green function @xmath80 . in particular , the mean field function @xmath81 can be easily calculated at @xmath82 @xmath83 and @xmath84 @xmath85 the density of states plays crucial roles in determining the physical properties of the dms system . especially , the formation of the impurity band arising from the impurity doping gives rise to many different aspects from the continuum ( i.e. virtual crystal approximation ) semiconductor band model . in this section , we calculate the dos for different parameters ( @xmath10 , @xmath12 , @xmath25 , and @xmath13 ) and show how the impurity bands are formed and separated from the main band . we describe the dos of dynamical mean - field calculations applying to simple semi - circle models . the dos is given by the imaginary part of the green function @xmath86 in our model for a strong magnetic coupling @xmath87 two spin polarized impurity bands appear at the bottom ( majority spin ) and at the top ( minority spin ) of the main band . each isolated impurity band has the weight @xmath12 . however , if the coupling is not strong ( @xmath88 ) the impurity band is not completely separated from the main bend . all dms samples show that the carrier density is much smaller than the impurity concentration ( @xmath89 ) due to the heavy compensation . as the system is the partially compensated , the chemical potential @xmath43 is located in the lower impurity band ( if the impurity bands are formed ) or the lower band edge ( if the bands are not formed ) . thus all physical properties are determined in the lower energy band edge . throughout this paper we only show the dos near the lower energy band . temperature dependence of the dos for @xmath90 and for a fixed coupling ( a ) @xmath91 and ( b ) @xmath92 . the evolutions of majority ( minority ) spin dos show in top ( bottom ) panels for various temperature @xmath930.0 , 0.5 , 0.7 , 0.9 , and 1.0 . , title="fig:",width=249 ] temperature dependence of the dos for @xmath90 and for a fixed coupling ( a ) @xmath91 and ( b ) @xmath92 . the evolutions of majority ( minority ) spin dos show in top ( bottom ) panels for various temperature @xmath930.0 , 0.5 , 0.7 , 0.9 , and 1.0 . , title="fig:",width=249 ] the calculated majority spin dos at @xmath82 ( upper half ) and @xmath16 ( lower half ) is shown for various doping @xmath94 0.001 , 0.01 , 0.05 , 0.1 and a fixed coupling ( a ) @xmath91 , and ( b ) @xmath95 . , title="fig:",width=249 ] the calculated majority spin dos at @xmath82 ( upper half ) and @xmath16 ( lower half ) is shown for various doping @xmath94 0.001 , 0.01 , 0.05 , 0.1 and a fixed coupling ( a ) @xmath91 , and ( b ) @xmath95 . , title="fig:",width=249 ] in fig . [ fig_1 ] we show the calculated dos for various temperatures as a function of energy . the evolutions of majority ( minority ) spin dos are shown in top ( bottom ) panels . in fig . [ fig_1](a ) the strength of the coupling constant ( @xmath91 ) is not strong enough to form the impurity band . note that this value of coupling constant ( @xmath91 ) is the critical value for impurity band formation as @xmath96 . at @xmath82 the majority ( minority ) spin band is shifted to lower ( higher ) energy compared with the noninteracting band which has a band edge at @xmath97 . thus all carriers are fully polarized when the carrier density is low . as the carrier density increases they start to occupy the minority band if the chemical potential crosses into the minority spin band . recently we showed that the optical conductivity of the system is dramatically changed with the occupation of the minority band @xcite . we show in this paper that the calculated dc transport properties are also very sensitive to the minority band occupation . as the temperature increases , the minority band occupation grows and the carriers with minority spin increase due to thermal fluctuations . at @xmath16 both spins are equally populated and the bands becomes symmetric . as expected we have a separated impurity band from main band for a strong coupling @xmath92 shown in fig . [ fig_1](b ) . when @xmath98 the impurity band is fully occupied , no low energy hopping processes to main band are allowed and the system becomes a band insulator . if the impurity band is partially occupied the delocalization energy increases . at the half filling of the impurity band the system has the highest @xmath2 @xcite . as the temperature increases spin disorder grows and the band becomes symmetric . but the impurity band width of the paramagnetic state ( @xmath99 ) is smaller than that of the ferromagnetic state . this band shrinking occurs because the neighboring spins in the paramagnetic state are uncorrelated . the calculated majority spin dos at ( a ) @xmath82 and ( b ) @xmath16 is shown for a fixed @xmath90 and for various coupling constant @xmath100 0.5 , 1.0 , 1.25 , 1.5 , 2.0 . , width=268 ] in fig . [ fig_2 ] we show the calculated majority spin dos corresponding to the disordered spin state ( at @xmath16 , bottom panels ) and ferromagnetic state ( at @xmath82 , top panels ) . the evolutions of the energy ( @xmath101 ) dependent dos are shown for different doping parameter @xmath940.001 , 0.01 , 0.05 , 0.1 and for fixed coupling constant ( a ) @xmath91 and ( b ) @xmath95 . in our model the impurity level ( acceptor energy level ) and the formation of an impurity band depend on the ferromagnetic coupling @xmath10 . if @xmath102 the impurity level is not isolated from the main band , but if @xmath103 we find an isolated impurity level below the main band . the small dot in fig . [ fig_2](b ) indicates the isolated impurity level in the dilute limit ( @xmath96 ) . as @xmath12 increases for @xmath103 an impurity band centered around the impurity level is formed below the main band . for @xmath88 the impurities give rise to band tailing in the main band edge instead of forming impurity band . the band width of the impurity band for @xmath103 increases with @xmath12 since the number of states of the impurity band increases with @xmath12 . if @xmath12 is bigger than @xmath104 the impurity band merges into the main band . for @xmath95 we have @xmath105 at @xmath82k and @xmath106 at @xmath16 . in fig . [ fig_3 ] we show the majority spin dos at ( a ) @xmath82 and ( b ) @xmath16 for a fixed @xmath90 and for various coupling constant @xmath100 @xmath107 , @xmath108 , @xmath109 , 1.5 t , 2.0 t . for @xmath88 we see the expected band shift proportional to @xmath110 . for @xmath111 an impurity band centered at impurity level and containing @xmath12 states is seen to split off from the main band . due to the widening of the impurity band with @xmath12 we find the separated impurity band when @xmath112 at @xmath82 and @xmath113 at @xmath16 for @xmath90 . thus the separated impurity band is expected with small @xmath12 and large values of @xmath10 . the majority ( top panel ) and minority ( bottom panel ) dos at @xmath82 are shown for a fixed value of @xmath95 and @xmath90 for various potential scattering @xmath114 , @xmath115 , 0.0 , 0.25 , 0.5 . , width=268 ] in fig . [ fig_4 ] we show the calculated dos at @xmath82 for a fixed value of @xmath95 and @xmath90 , and for various values of potential scattering @xmath116@xmath117 , @xmath118 , 0.0 , @xmath119 , @xmath107 . the dos of majority ( minority ) spin is shown in the top ( bottom ) panel . when we include the potential scattering ( @xmath25 ) in addition to the spin - spin coupling ( @xmath10 ) , the carriers with spin parallel to local impurity spin feel a potential @xmath37 on each magnetic impurity site and anti - parallel carriers feel a potential @xmath38 . thus the formation of the impurity band and the corresponding physical properties of the system depend on the combined coupling @xmath120 . [ fig_4 ] shows that while the band edge of the minority spins is slightly dependent on the potential scattering , the majority spin dos is strongly affected by the potential scattering . even weak potential scattering can change the extended majority spin band into the spin - polarized impurity bands or the well formed impurity band into the band tail of the main band . in the following section we show that our calculated resistivities depend strongly on whether the carriers are within the impurity band or in the band tail of the extended main band . the conductivity is calculated from the usual kubo formula . the kubo formula for the conductivity @xmath26 involves the two particle current - current response function . since the irreducible vertex in the response function is purely local in our approximation of dmft there is no vertex correction @xcite . thus , within this approximation only the simple bubble survives and the real part of the finite frequency conductivity is given by @xmath121 } { \omega } \nonumber \\ & \times & a_{\sigma } ( \varepsilon , \omega ) a_{\sigma } ( \varepsilon , \omega + \omega ) , \label{sigxx}\end{aligned}\ ] ] where @xmath122 is the spectral function , @xmath123 is the current vertex for the bethe lattice @xcite , @xmath124 is the fermi distribution function . for the hypercubic lattice @xcite @xmath125 has been used in eq . ( [ sigxx ] ) , but for the bethe lattice the explicit form is derived in ref . . the dc - resistivity @xmath126 is then found from eq . ( [ sigxx ] ) in the limit @xmath127 with @xmath128 , which is given by @xmath129 the density dependent resistivity at @xmath16 for a fixed value of @xmath90 and for various coupling constant @xmath130 , 1.0 , 1.2 , 1.4 , 1.5 , 2.0 . insets show the dos with the corresponding parameters . , width=288 ] in fig . [ fig_5 ] we show the calculated resistivity at @xmath16 as a function of carrier density for a fixed value of @xmath90 and various coupling constants @xmath131=0.7 , 1.0 , 1.2 , 1.4 , 1.5 and 2.0 . insets show the density of states near the band edge corresponding to the disordered spin state ( @xmath16 ) . all calculated results show that @xmath132 diverges as @xmath133 due to the absence of carriers . as density increases we find two different behaviors depending on the formation of the impurity band . ( the critical coupling constant which gives rise to the formation of the well separated impurity band below the main band for @xmath90 , @xmath16 is @xmath134 . ) in the weak coupling limit @xmath135 , where the impurity band formation dose not happen , the resistivity decreases as the density increases monotonically . however , in the strong coupling limit @xmath136 , where the impurity band is formed , the resistivity diverges again when the impurity bands are fully filled ( i.e. , for @xmath98 ) and the system becomes a band insulator . in the intermediate coupling regime ( @xmath137 ) the resistivity shows non - monotonic behavior . the calculated resistivity for ( a ) a strong coupling @xmath92 with densities @xmath138 0.01 , 0.015 , 0.02 , 0.025 , 0.03 0.035 , 0.04 , 0.045 ( top to bottom)and ( b ) a weak coupling @xmath92 with densities @xmath138 0.008 , 0.01 , 0.015 , 0.02 , 0.025 , 0.035 , 0.04 , 0.045 ( top to bottom ) . dots indicate @xmath2 for given densities . , title="fig:",width=240 ] the calculated resistivity for ( a ) a strong coupling @xmath92 with densities @xmath138 0.01 , 0.015 , 0.02 , 0.025 , 0.03 0.035 , 0.04 , 0.045 ( top to bottom)and ( b ) a weak coupling @xmath92 with densities @xmath138 0.008 , 0.01 , 0.015 , 0.02 , 0.025 , 0.035 , 0.04 , 0.045 ( top to bottom ) . dots indicate @xmath2 for given densities . , title="fig:",width=240 ] in fig . [ fig_6 ] we show the calculated resistivity as a function of temperature for various density . for the strong coupling limit ( @xmath92 ) , in fig . [ fig_6](a ) , we find a crossover ( resistivity peak ) separating a good metal at low @xmath13 from a semiconductor at higher @xmath13 . the resistivity peaks are proportional to the energy separation between the chemical potential and the band edge of the minority band . in the high temperature regime above the resistivity peak the decreasing resistance with increasing temperature , characteristic of a semiconductor or insulator , is due to thermal excitation of the carriers from impurity band to the upper minority spin band . the resistivity decreases with density because the carriers in the main band are scattering off the impurities . but at low temperatures most carriers in the impurity band contribute to the scattering and give non - monotonic density dependence of the resistivity . the metallic behavior at low temperature can be understood by the disappearance of the coherent central quasiparticle peak in the dos . for the weak coupling limit ( @xmath91 ) , in fig . [ fig_6](b ) , the resistivity can be explained by the scattering effects in the main band except the behavior in the ferromagnetic state ( @xmath1 ) . for @xmath1 the resistivity changes from insulating at low densities to metallic at high density . details of this behavior are given in fig . [ fig_7 ] . we see that the resistivity crossover takes place only at low densities since the minority band is occupied by the carriers at high density . the low temperature dc - resistivity as a function of temperature for a fixed @xmath90 and for various densities . in ( a ) the results for a strong coupling constant ( @xmath92 ) and densities @xmath139 , 0.012 , 0.015 , 0.025 , 0.3 , 0.35 , 0.38 , 0.40 are shown . in ( b ) we use @xmath91 and @xmath140 , 0.016 , 0.0163 , 0.0165 , 0.18 , 0.2 , 0.25 , 0.3 ( from top to bottom ) . dots indicate the @xmath2 for given parameters . , title="fig:",width=240 ] the low temperature dc - resistivity as a function of temperature for a fixed @xmath90 and for various densities . in ( a ) the results for a strong coupling constant ( @xmath92 ) and densities @xmath139 , 0.012 , 0.015 , 0.025 , 0.3 , 0.35 , 0.38 , 0.40 are shown . in ( b ) we use @xmath91 and @xmath140 , 0.016 , 0.0163 , 0.0165 , 0.18 , 0.2 , 0.25 , 0.3 ( from top to bottom ) . dots indicate the @xmath2 for given parameters . , title="fig:",width=240 ] in fig . [ fig_7 ] we show the dc - resistivity as a function of temperature for different densities in the low temperature regime ( @xmath141 ) . in fig . [ fig_7](a ) we use the parameters @xmath90 and a strong coupling @xmath92 . in this case all carriers ( if @xmath142 ) occupy the impurity band and stay mainly at mn sites . thus , the carriers in the impurity band follow the fluctuation of the localized mn spin . at half filling of the spin polarized impurity band ( @xmath143 ) the resistivity has the lowest value and its behavior corresponds to that of the double exchange ( de ) model ( see fig . [ fig_10 ] ) , that is , the resistivity decreases below the critical temperature because of the spin disorder scattering as the temperature decreases @xcite . in the low density limit ( @xmath144 , less than half filling ) the resistivity is dominated by the ` impurity band ' contribution and the resistivity increases as density decreases due to the lack of mobile carriers . as the carrier density is increased above the half filling ( @xmath145 ) the resistivity increases due to the filling of the band . we also find a very different temperature behavior of the resistivity from the low density case . as the temperature decreases the resistivity increases just below @xmath2 , then decreases at very low temperatures . the counter - intuitive increase of resistivity just below @xmath2 ( as @xmath13 is decreased ) arises because , as the carrier spins are aligned to the impurity spins , the binding of the carriers to the impurity spins increases corresponding to an increase in the basic scattering rate . in the very low temperature range , however , the de - like mechanism dominates , which gives rise to decrease of the resistivity . ( when the impurity band is spin - polarized , carriers which are bound to impurity site must have spins parallel to impurity spin . thus , as the spins order ferromagnetically , the basic ability of carriers to move in the impurity band increases . ) the overall temperature dependence of resistivity is very weak in the strong coupling limit . this weak @xmath13-dependence of the dc resistivity below @xmath2 occurs because the increase in scattering rate due to the binding is compensated by de - like mechanism . in fig . [ fig_7](b ) we show the results for a weak coupling limit , @xmath91 . in this case the impurities contribute to form the band tail of the main band and the chemical potential lies in the main band . in the low density limit ( @xmath146 , where @xmath147 is the density above which the minority band starts to fill ) the resistivity below @xmath2 increases as the temperature decreases . the increase in resistivity is due to increased carrier - spin coupling as mentioned above . but , in the high density limit ( @xmath148 , where at @xmath82 the minority spin - band is occupied ) we find the resistivity decreasing as the temperature decreases . this metallic behavior in high density regime and for a low coupling constant can be understood by the small scattering rate of the carriers in the minority band . if the carriers are near the edge of the majority spin band , the carriers form a spin - polarized bound state , so the effective scattering rate strongly increases , which gives rise to the insulating behavior as shown in the low density results . on the other hand , if the carriers are in the minority spin band the carriers form an anti - bound state at the top of the band , so that at the physically relevant lower band edge , the effective scattering rate decreases . in addition in three dimensions the vanishing of the density of states at the band edge further decreases the scattering . these effects are quite large and dominate as the minority band is occupied , which gives rise to the metallic behavior in the high density limits . in the paramagnetic state the resistivity is almost temperature independent . in the limit @xmath0 the local green function @xmath149 is temperature independent and the resistivity depends on temperature only weakly through the fermi function ( thermal smearing around the chemical potential ) . the energy evolution of optical conductivity for ( a ) a intermediate coupling @xmath91 and ( b ) a strong coupling @xmath92 . solid ( dashed ) lines correspond to the results for @xmath82 ( @xmath16 ) . insets show the dc conductivities with corresponding parameters . arrows indicate @xmath2 . , width=288 ] in fig . [ fig_8 ] we show the relation between the optical conductivity and dc - resistivity . the main panels of fig . [ fig_8 ] show the evolution of the conductivity for two couplings ; weak ( @xmath91 , fig . [ fig_8](a ) , where the impurity band formation is not accomplished ) , and strong ( @xmath92 , fig . [ fig_8](b ) , where the impurity band is well formed ) ; the insets show the dc resistivity with the same parameters . the solid ( dashed ) lines indicate the optical conductivity for @xmath82 ( @xmath16 ) . when the impurity band is not formed ( for @xmath91 ) we find approximately the drude form for optical conductivity expected for carriers scattering off random impurities ( a closer examination reveals minor differences due to density of states variations near the band edge ) . in this case ( and for @xmath150 ) the dc conductivity shows insulating behavior due to the formation of the bound state . since the carrier density is low enough not to fill the minority band , the formation of anti - bound state dose not take place , which reduce the scattering rate . in the @xmath92 case the density of states plot shows the formation of an impurity band and the corresponding conductivity has two structures : a low - frequency quasi - drude peak corresponding to motion within the impurity band and a higher frequency peak corresponding to excitations from the impurity band to the main band . in this case the dc resistivity shows metallic behavior because the reduction of spin - disorder scattering dominates over the bound - state formation . in in@xmath4mn@xmath5as @xcite we find the drude - like conductivity is correlated with the insulating behavior , but in ga@xmath4mn@xmath5as @xcite the mid infrared peak in the optical conductivity is closely related to the decrease of the dc - resistivity below the critical temperature . the calculated dc - resistivity as a function of temperature for a fixed @xmath90 , @xmath139 , @xmath91 , and for various potential scattering @xmath114 , @xmath151 , 0.0 , 0.05 , 0.2 . arrows indicate the critical temperature @xmath2 . , width=288 ] fig . [ fig_9 ] shows the sensitivity of the predicted behavior to potential scattering . in this figure we use the parameters : @xmath90 , @xmath139 , @xmath91 , and for various potential scatterings @xmath152@xmath153 , @xmath151 , 0.0 , 0.05 , 0.2 . at zero scalar potential ( @xmath154 , middle panel ) the impurity band is not formed , and the resistivity shows insulating behavior below @xmath2 due to the bound state formation of the carrier spins with impurity spins . as the potential is made more attractive ( negative ) , the impurity band features become pronounced and the spin - disorder scattering decreases ( the carriers move easily in the impurity band ) . as the potential is made more repulsive ( positive ) , the impurity band rapidly rejoins the main band . this reduces the energy gap between the fermi energy and the band edge of the minority band , and for large enough repulsive potential the minority band starts to be filled by the carriers . thus , the anti - bound state formation dominates and it reduces the scattering rate below @xmath2 and shows a metallic behavior . the dms transport properties are sensitive to the combined coupling @xmath120 , and not solely to the exchange coupling @xmath10 . resistivity @xmath155 for various magnetic field is given as a function of temperature for ( a ) @xmath156 , @xmath157 , @xmath158 , and ( b ) @xmath156 , @xmath90 , @xmath143 . insets show the normalized magnetization @xmath159 , where @xmath160 is the saturation magnetization . , title="fig:",width=211 ] resistivity @xmath155 for various magnetic field is given as a function of temperature for ( a ) @xmath156 , @xmath157 , @xmath158 , and ( b ) @xmath156 , @xmath90 , @xmath143 . insets show the normalized magnetization @xmath159 , where @xmath160 is the saturation magnetization . , title="fig:",width=211 ] now we compare the transport properties of dms to those of another system with strong carrier - spin couplings , namely cmr manganites @xcite . in cmr , instead of being dilute random impurities as in dms , the mn ions form an ordered lattice . they possess large local moment , to which mobile carriers are very strongly coupled . thus instead of a spin - polarized impurity band , there is a spin - polarized conduction band , sufficiently well separated from other spin bands . the periodic arrangement of the mn sites means that ( in the absence of other physics ) the scattering rate decreases as t is lowered . cmr materials can be understood well by the double exchange ( de ) model @xcite . in our model this corresponds to @xmath161 , in which all magnetic ions replace ga at the cation sites . in this case ( @xmath161 ) the fully polarized spin band is well separated from the other bands instead of forming impurity band . the temperature dependences of the resistivity @xmath155 and magnetization @xmath162 are given in fig . [ fig_10](a ) for @xmath157 with @xmath156 , @xmath158 , and various magnetic fields . above @xmath2 the resistivity has a small temperature dependence since the local spin fluctuation is saturated above @xmath2 . below @xmath2 resistivity decrease as magnetization increases . the origin of the resistivity dependence on the magnetization is spin disorder scattering , which gives rise to the scaled behavior of the resistivity , @xmath163 , where @xmath164 is a temperature / field independent constant @xcite . the origin of the resistivity is qualitatively explained by the carrier scattering due to the thermally fluctuating spin configurations , or the spin - disorder scattering . as the spontaneous or the induced magnetic moment is developed , the amplitude of the spin fluctuation decreases so that the resistivity also decreases . in fig . [ fig_10](b ) we show the resistivity for dms system with @xmath90 . the overall features of temperature and magnetic field dependence look similar to the de model . however , we find that the negative magnetoresistence at @xmath2 is very weak and the resistivity above @xmath2 is not saturated . the fast drop of the resistivity just below @xmath2 can be explained by spin - disorder scattering . these ideas of de model have limited applicability to the dms systems , i.e. only for strong couplings and near half filling of the spin polarized impurity band . as shown in previous figures these ideas can not explain the resistivity behavior in low coupling and high density regimes . note that an important ingredient of dms transport properties is missing from our dmft theory which , while accounting well for the non - perturbative effects of spin disorder ( @xmath10 ) and local potential ( @xmath25 ) scattering by the magnetic impurities , leaves out all ionized impurity disorder that may very well be important . when the dominant scattering mechanism is the scattering by charged impurities the boltzmann transport theory may be used to calculate the electrical resistivity of the carriers since dmft is not well - suited for treating long - range disorder . due to the band splitting in the ferromagnetic state the carrier densities @xmath165 for spin up / down are not equal . note that the total density @xmath166 . in this situation the total conductivity can be expressed as a sum of spin up / down contributions @xmath167 where @xmath168 is the conductivity of the ( @xmath169 ) spin subband . the conductivities @xmath168 are given by @xmath170 where @xmath171 is the carrier effective mass , and the energy averaged transport relaxation time @xmath172 for the ( @xmath169 ) subbands are given in the boltzmann theory by @xmath173 } { \int d\varepsilon \varepsilon \left [ -\frac{\partial f^{\pm}(\varepsilon)}{\partial \varepsilon } \right ] } , \label{tau}\ ] ] where @xmath174 is the energy dependent relaxation time for the ( @xmath169 ) subbands , and @xmath175 is the carrier ( fermi ) distribution function @xmath176/k_bt}},\ ] ] where @xmath177 is the chemical potential at finite temperature . the energy dependent relaxation time is given in the born approximation by @xmath178^{-1 } = \frac{2\pi}{\hbar } \int \frac{d^3k'}{(2\pi)^3}n_i|v^{sc}_{\bf k - k'}|^2 ( 1-\cos\theta_{\bf kk ' } ) \delta(\varepsilon_{\bf k } - \varepsilon_{\bf k'}),\ ] ] where @xmath179 is the charged impurity concentration , and @xmath180 is the screened carrier - impurity coulomb interaction , which can be expressed as @xmath181 ^ 2},\ ] ] where @xmath182 is the charge of impurities , @xmath183 the background ( gaas ) lattice dielectric constant , and @xmath184 the temperature dependent screening function . wave vector dependence of the effective screening function ( polarizability function @xmath185 normalized its paramagnetic state value @xmath186 ) for various polarizations @xmath187 . , width=211 ] in this paper we consider two screening approaches : thomas - fermi ( tf ) approximation and random phase approximation ( rpa ) . within tf screening we have @xmath188,\ ] ] where @xmath189 with @xmath190 being the fermi wave vector of the spin unpolarized state and @xmath191 the effective bohr radius . @xmath192 is the fermi wave vector of the each spin - split subband . note that tf screening wave vector , being a long wavelength approximation , is dependent only on the spin polarization , but not on the temperature explicitly . it is therefore temperature independent above the critical temperature . within tf screening approximation we have the energy dependent relaxation time ( by integrating eq . 21 ) @xmath178^{-1 } = \frac{2\pi}{3}\frac{n_i}{n}\frac{e_{f}^0}{\hbar } \frac{k k_f^0 } { \tilde{k}_f^2 } f(4k^2/q_{s}^2),\ ] ] where @xmath193 and @xmath194 $ ] . when a system is fully polarized we have @xmath195 and @xmath196 . for the random phase approximation ( rpa ) the temperature dependent screen function is given by @xmath197,\ ] ] where @xmath198 is the temperature dependent static lindhard function for each spin subband . in fig . [ fig_11 ] we show the screening function @xmath199 , where @xmath186 , for various polarization @xmath187 . for an unpolarized state ( @xmath200 ) there is an inflection point at @xmath201 , which is the usual kohn anomaly . for partially polarized state @xmath202 we have two inflection points at @xmath203 and the value of @xmath204 decreases as @xmath205 increases . when the system is fully polarized @xmath206 , and @xmath185 has an inflection point at @xmath207 , and @xmath208 . calculated resistivity as a function of @xmath209 for various value of @xmath210 1.0 , 2.0 , 5.0 , 10.0 ( top to bottom ) . solid ( dashed ) lines show the results calculated with rpa ( tf ) screening function.,width=240 ] at @xmath82 , @xmath211 where @xmath212 is the fermi energy for the ( @xmath169 ) subbands , and then @xmath213 giving the familiar result @xmath214 . within tf screening the ratio of the resistivity of the fully polarized state to that of the unpolarized state becomes @xmath215 , where @xmath216 . ( note @xmath217 . ) in fig . [ fig_12 ] we show the calculated resistivity as a function of @xmath209 for several value of @xmath218 . solid ( dashed ) lines show the results calculated with rpa ( tf ) screening . in general , for small values of @xmath218 , @xmath219 , but for large @xmath218 , @xmath220 . at @xmath221 we have @xmath222 . for gaas @xmath223 corresponds to the hole density @xmath224 . in relatively high density limits ( large @xmath218 ) the two approximations agree very well , which indicates that the @xmath225 scattering mostly contributes to the scattering time . however , in the low density regime ( small @xmath218 ) we that find the two screening theories give very different results for the spin polarized state . noting that the tf approximation is just the long wavelength limit of rpa , we emphasize that rpa is obviously the better theory . calculated resistivity , @xmath226 , as a function of temperature . in ( a ) the rpa results are shown with hole densities @xmath227 0.05 , 0.1 , 0.2 , 0.5 . in ( b ) the tf screening results are shown with hole densities @xmath227 0.01 , 0.05 , 0.1 , 0.2 . here the impurity density @xmath228 is used . , title="fig:",width=211 ] calculated resistivity , @xmath226 , as a function of temperature . in ( a ) the rpa results are shown with hole densities @xmath227 0.05 , 0.1 , 0.2 , 0.5 . in ( b ) the tf screening results are shown with hole densities @xmath227 0.01 , 0.05 , 0.1 , 0.2 . here the impurity density @xmath228 is used . , title="fig:",width=211 ] calculated resistivity as a function of temperature for a hole density ( a ) @xmath229 0.1 and ( b ) @xmath2300.2 with finite temperature rpa screening function . the top insets show the temperature dependence of the mn ions and hole magnetization . the bottom insets show the calculated temperature dependence of the tf screening wave vector @xmath231 . here the impurity density @xmath228 is used . , title="fig:",width=211 ] calculated resistivity as a function of temperature for a hole density ( a ) @xmath229 0.1 and ( b ) @xmath2300.2 with finite temperature rpa screening function . the top insets show the temperature dependence of the mn ions and hole magnetization . the bottom insets show the calculated temperature dependence of the tf screening wave vector @xmath231 . here the impurity density @xmath228 is used . , title="fig:",width=211 ] in figs . [ fig_13 ] and [ fig_14 ] we show our calculated resistivity for gamnas samples . we use the parameters corresponding to gamnas : dielectric constant @xmath232 , hole effective mass @xmath233 , impurity density @xmath234 , and the magnetic coupling constant @xmath235120 mev nm@xmath236 . [ fig_13 ] shows normalized resistivity @xmath226 as a function of temperature @xmath237 for several values of hole density . in fig . [ fig_13 ] there are two independent sources of temperature dependence in our calculated resistivity one source is the energy averaging defined in eq . ( [ tau ] ) and the other is the explicit temperature dependence of the finite temperature rpa screening function @xmath238 . since the fermi temperature is much higher than the critical temperature ( @xmath239 ) the screening function is weakly temperature dependent above the critical temperature ( unpolarized state ) . thus the decrease of the resistivity above @xmath2 with increasing temperature arises from the thermal energy averaging . this is a well - known high - t effect of ionized impurity scattering in semiconductors : @xmath240 decreases with increasing @xmath13 due to the thermal averaging . as density decreases this effect is enhanced . in the ferromagnetic polarized state ( @xmath1 ) , as the temperature decreases from @xmath16 , the screening length increases until all the holes are polarized . in this temperature range we find a strong temperature dependent resistivity . this strong temperature dependence arises from the low temperature screening function and the change of the carrier densities of the each subband . when all the carriers are polarized ( i.e. at low density and temperature ) the screening function is almost independent of temperature and the resistivity is also therefore temperature independent . the interplay between the screening length and the down spin ( ) carrier density gives rise to the minimum of the resistivity in the low density regime . the decrease of the scattering time due to the fast increase of the down spin density overwhelms the increase of the scattering time due to the decrease of the screening length , producing the non - monotonic behavior for @xmath241 and @xmath242 in fig . 13 . but in the high density regime , where the spins are not fully polarized even at low @xmath13 , screening is the dominant effect on the temperature dependent resistivity . as the density increases the relative low temperature resistivity , @xmath226 , decreases until the holes are partially polarized . the change of screening wave vector is larger in this case , leading to the monotonically increasing @xmath240 in the @xmath243 regime . in fig . [ fig_14 ] we show our calculated resistivity as a function of temperature for two hole densities ( a ) @xmath227 0.1 and ( b ) @xmath2300.2 with finite temperature rpa screening function . the temperature dependence of the impurity ( i.e. mn ) moment magnetization as well as the hole spin polarization is given in the insets , which is calculated using the weiss molecular mean - field theory for delocalized carriers @xcite . note that the hole gas is almost fully spin polarized upto @xmath244 at low density , but at higher density the holes are partially polarized even at @xmath82 . the bottom insets show the calculated temperature dependence of the finite temperature screening wave vector @xmath231 . at high density the screening wave vector changes by 6% when the temperature increases from zero to @xmath2 due to the partial polarization of the holes at @xmath82 . but at low density the decrease of the screening wave vector is about 20% . in the metallic gamnas samples the change of the resistivity when the temperature goes from @xmath2 to zero is about 20% in good agreement with our calculation . similarly the observed decrease of @xmath240 for @xmath0 also arises naturally in our theory as a consequence of thermal averaging . thus , the temperature dependence of the resistivity in the metallic dms samples may be arising almost entirely from the temperature dependence of screening and thermal averaging in the charged impurity scattering . we have developed in this work two complementary theories for understanding dms transport properties . our work establishes that dms transport , even in ideal intrinsic circumstances , is rather complex , and depends sensitively on many system parameters including the exchange coupling , the magnetic impurity density , the carrier density , the temperature , the band parameters of the parent semiconductor ( e.g. effective mass , etc . ) , and the details of the charged impurity distribution ( and therefore compensation ) . given such a complex parameter space , it is not meaningful to try to develop quantitative transport theories at this early stage of the subject since all the intrinsic parameters may not be known . we therefore focus in this work on developing a comprehensive qualitative theoretical description which emphasizes general broad features of how various parameters affect dms transport behavior . as such we have concentrated in this work on understanding temperature and carrier density dependence of dc resistivity in a model dms system , keeping primarily the extensively studied ga@xmath4mn@xmath5as system in mind . even for gamnas , the transport data for various values of mn concentration ( @xmath245 ) cover much too broad a range of behavior for a unified and coherent theoretical description . for example , low ( and sometimes large ) mn concentrations ( @xmath246 and sometimes @xmath247 ) are known to lead to insulating transport behavior usually attributed to a disorder - driven metal - insulator ( anderson ) localization . we neglect _ all _ localization effects in our theory . the localized gamnas regime in all likelihood requires its own characteristic theoretical description such as the bound magnetic polaron percolation theory @xcite whose transport properties @xcite have recently been theoretically analyzed . even without the disorder - induced strong localization complications , neglected completely in this work , we face the formidable difficulty of using the semiconductor valence / conduction band ( the valence band for gamnas , where the carriers mediating the ferromagnetic interaction are holes ) or the magnetic impurity induced impurity band ( i.e. mn induced d - band in the fundamental band gap of gaas ) picture for describing the carrier dynamics . the precise nature ( i.e. valence band versus impurity band ) of the dms carriers is still a controversial issue although it is likely that at the very high doping densities ( e.g. mn density @xmath248 in gamnas ) of dms interest the impurity band overlaps strongly with the valence band ( i.e. forms the tail of the valence band ) , and therefore , the distinction between the valence and the impurity band picture is not a real qualitative difference . our dmft theory , presented in sections ii - iv of this paper , clearly shows that in the strong exchange coupling ( @xmath249 ) regime the impurity band picture applies whereas in the weak - coupling regime ( @xmath250 ) , the valence band picture applies . it is possible that gamnas belongs to the intermediate coupling regime ( @xmath251 ) , where it may be more appropriate to think of the holes to be residing in the extended tail of the valence band , presumably with an enhanced effective mass compared with the gaas valence band hole mass . such a coupled impurity - valence band picture of gamnas is consistent with recent optical spectroscopy measurements @xcite , but more experimental work is needed to settle this question . the theoretical strength of our dmft description is that , being a nonperturbative technique , it can handle both strong - coupling and weak - coupling regimes , and our results presented in figs . 6 10 of this paper show qualitative difference between the strong - coupling regime ( @xmath252 ) with an impurity band well - separated from the semiconductor band and the intermediate - coupling regime ( @xmath253 ) with only band - tailing and no separate impurity band formation . temperature , carrier density , and impurity concentration all play qualitatively important roles in determining the dc transport properties within dmft , and sorting out the details with respect to experimental results may be extremely difficult . the weakness of dmft is that it can only include spin - disorder scattering ( controlled by @xmath10 ) and mn impurity induced local potential scattering ( controlled by @xmath25 ) effects on transport properties . as such it leaves out the most important scattering mechanism which may be operational in real samples , namely scattering by charged impurities which is often the most important resistive scattering process in heavily doped semiconductos below the room temperature ( or the optical phonon scattering regime which may well be above the room temperature ) . the reason dmft is unable to account for charged impurity scattering is that the coulombic impurity potential is long ranged , and dmft by construction is a local theory . thus , rewriting our starting hamiltonian ( eq . 1 ) more completely we have @xmath254,\ ] ] where @xmath255 is the carrier - charged impurity interaction and @xmath256 is the carrier - carrier ( i.e. hole - hole in gamnas ) interaction . in principle , the terms ( i.e. @xmath255 , @xmath256 ) within the square bracket are parts of the @xmath18 , but it is important to appreciate their considerable ( perhaps even dominant ) importance in determining the dc transport properties . to include the charged impurity scattering effects on transport , we use the highly successful and robust semiclassical boltzmann transport theory to dms systems assuming a mean - field approach where the long - range coulomb impurity potential arising from @xmath255 ( assuming random impurity scattering ) is screened by the polarization bubble diagrams arising from @xmath256 . this type of boltzmann transport theory is extremely successful in describing semiconductor transport properties @xcite . we note that our dmft and boltzmann transport theories are complementary dmft treats the spin disorder and the local potential scattering associated with mn dopants and boltzmann theory treats the scattering by screened charged impurity scattering . the magnetic dms properties enter the boltzmann theory only indirectly through the carrier spin polarization calculations . spin disorder scattering is _ not _ explicitly included in the boltzmann theory although it is straightforward to do so . our boltzmann theory manifests nontrivial interplay among temperature dependent screening , temperature dependent spin polarization ( i.e. spin up - down carrier densities ) , and thermal energy averaging , leading to temperature dependent resistivity ( figs . 13 and 14 ) that are rather similar to experimental observations @xcite in gamnas . based on this qualitative similarity we conclude that much of gamnas dms transport is dominated by screened charged impurity scattering with spin disorder scattering playing only a rather minor quantitative role . recently , lopez - sancho and brey @xcite have come to a very similar conclusion for gamnas transport properties . schematic diagram for the experimental gamnas dc - resistivity for decreasing hole density ( from bottom to top ) ( see ref . [ 3 ] ) . , width=211 ] before concluding we now discuss our theoretical results in light of the existing dms transport data . although there are some experimental transport data in other dms systems ( most notably inmnas with qualitatively similar behavior to gamnas ) , truly extensive and reliable transport data@xcite are available only for ga@xmath4mn@xmath5as ( in the @xmath257 regime ) dms samples . even for this well - studied system , experimental transport results are problematic , and have considerable spread in the sense that nominally `` identical '' gamnas samples ( i.e. same nominal carrier density and mn concentration made in the same growth run ) may have different @xmath2 and transport properties . this situation is improving as sample quality and processing ( e.g. annealing ) techniques improve , but experimental dms transport properties are still not robust in a quantitative way . this is of course very understandable given the very large parameter space ( i.e. exchange coupling , hole density , mn concentration , defects and impurities , compensation , band structure parameters , etc ) that dms transport depends on . with these serious caveats we show in fig . 15 a schematic depiction of the generic experimental observation for @xmath240 in gamnas for various hole densities . at high hole density the system shows `` metallic '' behavior for @xmath1 with @xmath240 increasing somewhat with @xmath13 upto @xmath2 , and then decreasing slowly for @xmath0 . even the optimally doped most `` metallic '' gamnas is , however , at best a bad metal with mobilities of the order of 10 @xmath258 with @xmath259 where @xmath260 is the transport mean free path . it is important to realize that dms transport is always extremely highly resistive due to the very large amount of impurities and defects invariably present in the low temperature mbe process needed for producing homogeneous gamnas samples . thus , from the perspective of heavily doped semiconductors , although the dms systems may be above the mott limit ( i.e. the carrier density in the mott metallic range ) , they are close to being anderson insulators due to strong disorder effects . as hole density decreases the system eventually becomes an insulator at low enough carrier densities ( the top curve in fig . 15 ) , with @xmath240 decreasing monotonically increasing @xmath13 . two generic features of fig . 15 are : ( a ) a peak ( or a kink ) in @xmath240 near @xmath261 , and ( 2 ) the slow decrease of @xmath240 for @xmath0 . both of these generic features of @xmath240 as well as the `` metallic '' high - density behavior are qualitatively ( perhaps even semi - quantitatively ) well - explained by our boltzmann theory approach including only scattering by ionized impurity scattering . this is apparent by comparing fig . 15 with figs . 14 and 13 where our boltzmann theory results are shown . physically , the increasing @xmath240 with @xmath13 ( @xmath262 ) arises from the decreasing strength of screening due to the interplay of two independent and competing physical effects : temperature induced suppression of screening and spin polarization ( i.e. the spin polarization decreasing with increasing temperature ) induced enhancement of screening with increasing temperature . as explained in sec . v the competition between these two effects depend on the carrier density , leading to some weak non - monotonicity in @xmath240 for @xmath1 . in this screening picture , the resistivity peak or cusp at @xmath263 arises from the ferromagnetic to paramagnetic transition which affects screening as the carrier density of states ( which is inversely proportional to the screening length ) changes from one in the fully spin polarized state to two in the fully paramagnetic phase . thus , even without any spin disorder scattering effects , just screened ionized impurity scattering by itself will give rise to the peak or the kink in @xmath240 at @xmath16 . we believe that spin disorder scattering , which also produces a kink at @xmath263 ( see , for example , figs . 7 and 9 ) , play only a minor role in the resistivity `` maximum '' at @xmath2 in optimally metallic gamnas with most of the peak structure arising from the screening properties of ionized impurity scattering . the second generic experimental feature in fig . 15 , the slow decrease of @xmath240 for @xmath0 , can not be explained _ at all _ by spin disorder scattering since spin disorder should remain large in the paramagnetic system ( @xmath99 ) and certainly should not decrease with increasing @xmath13 . our boltzmann theory provides a natural explanation ( figs . 13 and 14 ) for the decreasing @xmath264 as arising from the energy averaging of the relaxation time ( eq . 19 ) , i.e. , @xmath240 decreases with increasing @xmath13 simply because the holes move `` faster '' at higher temperatures ( i.e. increasing kinetic energy with increasing @xmath13 ) . this decreasing @xmath240 with increasing @xmath13 ( @xmath265 ) also shows that our neglect of phonon scattering in the transport theory is a valid approximation since phonon effects will always increase @xmath240 with increasing @xmath13 . @xmath240 will increase again when phonon scattering starts dominating over ionized impurity scattering at much higher temperatures . the relative lack of importance of phonon scattering in dms systems arises from their very strong charged impurity resistive scattering effects as reflected in very small sample mobilities . in this paper we investigate the transport properties of the diluted magnetic semiconductors using dynamical mean field theory and boltzmann transport theory . we have shown that the resistivity depends strongly on the system parameters , i.e. , exchange coupling , carrier density , doping , and temperature . the resistivity drop with decreasing temperature in the ferromagnetic state can be partially explained by the screening theory for metallic samples . the parameter dependence of the resistivity contains important information about the physics of diluted magnetic semiconductors . we find that in the strong exchange coupling regime the spin disorder scattering and the formation of the bound state in the impurity band compete to produce an unusual behavior in the temperature dependent resistivity . we also show that in the weak coupling regime the occupation of the minority spin band is critical to the scattering mechanisms , and substantially reduces the resistivity because the repulsive interaction between local moments and `` wrong - spin '' carriers suppresses the carrier amplitude at the impurity site , reducing the effective carrier - spin coupling . our boltzmann transport theory for charged impurity scattering is good qualitative agreement with the existing dms experimental data , showing that transport in dms gamnas system may very well be dominated primarily by screened ionized impurity scattering ( with spin disorder scattering playing only a minor secondary role ) , at least for the optimally doped metallic gamnas samples . we have completely neglected detailed band structure complications ( e.g. spin - orbit coupling in the valence band ) in our theory . these effects are certainly very important , but our interest in this paper is the development of a conceptually coherent qualitative theory for dms transport identifying the main transport mechanisms , and as such we have neglected all nonessential complications . we thank darpa and us - onr for support . f. matsukura _ et al . _ , , r2037(1998 ) ; s.j . potashnik , k.c . chun , j.j . berry , n. samarth , and p. schiffer , appl . . lett . * 79 * , 1495 ( 2001 ) ; sh.u . yuldashev , h. i m , v.sh . yalishev , c.s . park , t.w . kang , s. lee , y. sasaki , x. liu , and j.k . furdina , appl . phys . lett . * 82 * , 1206 ( 2003 ) . et al . _ , lett . * 79 * , 3473 ( 2001 ) ; n. theodoropoulou , a. f. hebard , m. e. overberg , c. r. abernathy , s. j. pearton , s. n. g. chu , and r. g. wilson phys . rev . lett . * 89 * , 107203 ( 2002 )

@xmath0department of electrical and computer engineering , university of illinois , urbana , il 61801 @xmath1department of chemical and biomolecular engineering , university of illinois , urbana , il 61801 @xmath2department of bioengineering , university of illinois , urbana , il 61801 @xmath3institute for genomic biology , university of illinois , urbana , il 61801 @xmath4these authors contributed equally to the work . @xmath5to whom correspondences should be addressed : olgica milenkovic , e - mail : milenkov@illinois.edu addressing the emerging demands for massive data repositories , and building upon the rapid development of technologies for dna synthesis and sequencing , a number of laboratories have recently outlined architectures for archival dna - based storage @xcite . the architecture in @xcite achieved a storage density of @xmath6 tb / gram , while the system described in @xcite raised the density to @xmath7 pb / gram . the success of the latter method may be largely attributed to three _ classical coding schemes : _ huffman coding , differential coding , and single parity - check coding @xcite . huffman coding was used for data compression , while differential coding was used for eliminating homopolymers ( i.e. , repeated consecutive bases ) in the dna strings . parity - checks were used to add controlled redundancy , which in conjunction with four - fold coverage allows for mitigating assembly errors . due to dynamic changes in biotechnological systems , none of the three coding schemes represents a suitable solution from the perspective of current dna sequencer designs : huffman codes are fixed - to - variable length compressors that can lead to catastrophic error propagation in the presence of sequencing noise ; the same is true of differential codes . homopolymers do not represent a significant source of errors in illumina sequencing platforms @xcite , while single parity redundancy or rs codes and differential encoding are inadequate for combating error - inducing sequence patterns such as long substrings with high gc content @xcite . as a result , assembly errors are likely , and were observed during the readout process described in @xcite . an even more important issue that prohibits the practical wide - spread use of the schemes described in @xcite is that accurate partial and random access to data is impossible , as one has to reconstruct the whole text in order to read or retrieve the information encoded even in a few bases . furthermore , all current designs support read - only storage . the first limitation represents a significant drawback , as one usually needs to accommodate access to specific data sections ; the second limitation prevents the use of current dna storage methods in architectures that call for moderate data editing , for storing frequently updated information and memorizing the history of edits . moving from a read - only to a rewritable dna storage system requires a major implementation paradigm shift , as : * 1 . * editing in the compressive domain may require rewriting almost the whole information content ; * 2 . * rewriting is complicated by the current data dna storage format that involves reads of length @xmath8 bps shifted by @xmath9 bps so as to ensure four - fold coverage of the sequence ( see figure [ fig : system ] ( a ) for an illustration and description of the data format used in @xcite ) . in order to rewrite one base , one needs to selectively access and modify four `` consecutive '' reads ; * 3 . * addressing methods used in @xcite only allow for determining the position of a read in a file , but can not ensure precise selection of reads of interest , as undesired cross - hybridization between the primers and parts of the information blocks may occur . to overcome the aforementioned issues , we developed a new , random - access and rewritable dna - based storage architecture based on dna sequences endowed with specialized address strings that may be used for selective information access and encoding with inherent error - correction capabilities . the addresses are designed to be _ mutually uncorrelated _ and to satisfy the _ error - control running digital sum constraint _ given the address sequences , encoding is performed by stringing together properly terminated prefixes of the addresses as dictated by the information sequence . this encoding method represents a special form of _ prefix - synchronized coding _ given that the addresses are chosen to be uncorrelated and at large hamming distance from each other , it is highly unlikely for one address to be confused with another address or with another section of the encoded blocks . furthermore , selection of the blocks to be rewritten is made possible by the prefix encoding format , while rewriting is performed via two dna editing techniques , the gblock and oe - pcr ( overlap - extension polymerase chain reaction ) methods @xcite . with the latter method , rewriting is done in several steps by using short and cheap primers . the first method is more efficient , but requires synthesizing longer and hence more expensive primers . both methods were tested on dna encoded wikipedia entries of size @xmath10 kb , corresponding to six universities , where information in one , two and three blocks was rewritten in the dna encoded domain . the rewritten blocks were selected , amplified and sanger sequenced @xcite to verify that selection and rewriting are performed with @xmath11 accuracy . the main feature of our storage architecture that enables highly sensitive random access and accurate rewriting is _ addressing_. the rational behind the proposed approach is that each block in a random access system must be equipped with an address that will allow for unique selection and amplification via dna sequence primers . instead of storing blocks mimicking the structure and length of reads generated during high - throughput sequencing , we synthesized blocks of length @xmath12 bps tagged at both ends by specially designed address sequences . adding addresses to short blocks of length @xmath8 bps would incur a large storage overhead , while synthesizing blocks longer than @xmath12 bps using current technologies is prohibitively costly . more precisely , each data block of length @xmath12 bps was flanked at both ends by two unique , yet different , address blocks of length @xmath13 bps . these addresses are used to provide specificity of access ( see figure [ fig : system ] ( b ) and the supplementary information for details ) . the remaining @xmath14 bases in a block are divided into @xmath15 sub - blocks of length @xmath16 bps , with each block encoding six words of the text . the `` word - encoding '' process may be seen as a specialized compaction scheme suitable for rewriting , and it operates as follows . first , different words in the text are counted and tabulated in a dictionary . each word in the dictionary is converted into a binary sequence of length sufficiently long to allow for encoding of the dictionary . for our current implementation and texts of choice , described in the supplementary information section , this length was set to @xmath17 . encodings of six consecutive words are subsequently grouped into binary sequences of length @xmath18 . the two - bit @xmath19 is appended as a word marker to the left hand side of each binary sequence of length @xmath18 , resulting in sequences of length @xmath20 bits . the binary sequences are subsequently translated into dna blocks of length @xmath16 bps using a new family of dna prefix - synchronized codes described in the methods section . our choice for the number of jointly encoded words is governed by the goal to make rewrites as straightforward as possible and to avoid error propagation due to variable codelengths . furthermore , as most rewrites include words , rather than individual symbols , the word encoding method represents an efficient means for content update . details regarding the counting and grouping procedure may be found in the supplementary information . for three selected access queries , the @xmath12 bps blocks containing the desired information were identified via primers corresponding to their unique addresses , pcr amplified , sanger sequenced , and subsequently decoded . two methods were used for content rewriting . if the region to be rewritten had length exceeding several hundreds , new sequences with unique primers were synthesized as this solution represents a less costly alternative to rewriting . for the case that a relatively short substring of the encoded string had to be modified , the corresponding @xmath12 bps block hosting the string was identified and the changes were generated via dna editing . both the random access and rewriting protocols were tested experimentally on two jointly stored text files . one text file , of size @xmath21 kb , contained the history of university of illinois , urbana - champaign ( uiuc ) based on its wikipedia entry retrieved on @xmath22 . the other text file , of size @xmath23 kb , contained the introductory wikipedia entries of berkeley , harvard , mit , princeton , and stanford , retrieved on @xmath24 . encoded information was converted into dna blocks of length @xmath12 bps synthesized by idt ( integrated dna technologies ) , at a cost of @xmath25 per @xmath12 bps ( see http://www.idtdna.com/pages/products/genes/gblocks-gene-fragments ) . the rewriting experiments encompassed : * 1 . * _ pcr selection and amplification of one @xmath12 bps sequence and simultaneous selection and amplification of three @xmath12 bps sequences in the pool . _ all @xmath26 linear @xmath12 bps fragments were mixed , and the mixture was used as a template for pcr amplification and selection . the results of amplification were verified by confirming sequence lengths of @xmath12 bps banks via gel electrophoresis ( figure [ fig : electro ] ( a ) ) and by randomly sampling @xmath27-@xmath28 sequences from the pools and sanger sequencing them ( figure [ fig : electro ] ( b ) ) . * 2 . * _ experimental content rewriting via synthesis of edits located at various positions in the @xmath12 bps blocks . _ for simplicity of notation , we refer to the blocks in the pool on which we performed selection and editing as b1 , b2 , and b3 . two primers were synthesized for each rewrite in the blocks , for the forward and reverse direction . in addition , two different editing / mutation techniques were used , gblock and overlap - extension ( oe ) pcr . gblocks are double - stranded genomic fragments used as primers or for the purpose of genome editing , while oe - pcr is a variant of pcr used for specific dna sequence editing via point editing / mutations or splicing . to demonstrate the plausibility of a cost efficient method for editing , oe - pcr was implemented with general primers ( @xmath29 bps ) only . note that for edits shorter than @xmath30 bps , the mutation sequences were designed as overhangs in primers . then , the three pcr products were used as templates for the final pcr reaction involving the entire @xmath12 bps rewrite . figure [ fig : system ] ( c ) illustrates the described rewriting process . in addition , a summary of the experiments performed is provided in table [ tab : summary ] . .[tab : summary ] selection , rewriting and sequencing results . each rewritten @xmath12 bps sequence was ligated to a linearized pcrtm - blunt vector using the zero blunt pcr cloning kit and was transformed into _ _ the _ e. coli _ strains with correct plasmids were sequenced at acgt , inc . sequencing was performed using two universal primers : m13f_20 ( in the reverse direction ) and m13r ( in the forward direction ) to ensure that the entire block of @xmath12 bps is covered . [ cols="^,^,^,^",options="header " , ] to encode information on dna media , we employed a two - step procedure . first , we designed address sequences of short length which satisfy a number of constraints that makes them suitable for highly selective random access @xcite . _ constrained coding _ ensures that dna patterns prone to sequencing errors are avoided and that dna blocks are accurately accessed , amplified and selected without perturbing or accidentally selecting other blocks in the dna pool . the coding constraints apply to address primer design , but also indirectly govern the properties of the fully encoded dna information blocks . the design procedure used is semi - analytical , in so far that it combines combinatorial methods with computer search techniques . we required the address sequences to satisfy the following constraints : * ( c1 ) constant gc content ( close to @xmath31 ) of all their prefixes of sufficiently long length . dna strands with @xmath31 gc content are more stable than dna strands with lower or higher gc content and have better coverage during sequencing . since encoding user information is accomplished via prefix - synchronization , it is important to impose the gc content constraint on the addresses as well as their prefixes , as the latter requirement also ensures that all fragments of encoded data blocks have balanced gc content . * ( c2 ) large mutual hamming distance , as it reduces the probability of erroneous address selection . recall that the hamming distance between two strings of equal length equals the number of positions at which the corresponding symbols disagree . an appropriate choice for the minimum hamming distance is equal to half of the address sequence length ( @xmath32 bps in our current implementation which uses length @xmath13 address primers ) . * ( c3 ) uncorrelatedness of the addresses , which imposes the restriction that prefixes of one address do not appear as suffixes of the same or another address and vice versa . the motivation for this new constraint comes from the fact that addresses are used to provide unique identities for the blocks , and that their substrings should therefore not appear in `` similar form '' within other addresses . here , `` similarity '' is assessed in terms of hybridization affinity . furthermore , long undesired prefix - suffix matches may lead to read assembly errors in blocks during joint informational retrieval and sequencing . * ( c4 ) absence of secondary ( folding ) structures , as such structures may cause errors in the process of pcr amplification and fragment rewriting . addresses satisfying constraints c1-c2 may be constructed via error - correcting codes with small running digital sum @xcite adapted for the new storage system . properties of these codes are discussed in section [ sub : balanced ] . the novel notion of _ mutually uncorrelated sequences _ is introduced in [ sub : correlation ] . constructing addresses that simultaneously satisfy the constraints c1-c4 and determining bounds on the largest number of such sequences is prohibitively complex @xcite . to mitigate this problem , we resort to a _ semi - constructive _ address design approach , in which balanced error - correcting codes are designed independently , and subsequently expurgated so as to identify a large set of mutually uncorrelated sequences . the resulting sequences are subsequently tested for secondary structure using _ mfold _ and _ vienna _ we conjecture that the number of sequences satisfying c1-c4 grows exponentially with their length : proofs towards establishing this claim include results on the exponential size of codes under each constraint individually . given two uncorrelated sequences as flanking addresses of one block , one of the sequences is selected to encode user information via a new implementation of _ prefix - synchronized encoding _ @xcite , described in [ sub : prefix ] . the asymptotic rate of an optimal single sequence prefix - free codes is one . hence , there is no asymptotic coding loss for avoiding prefixes of one sequence ; we only observe a minor coding loss for each finite - length block . for multiple sequences of arbitrary structure , the problem of determining the optimal code rate is significantly more complicated and the rates have to be evaluated numerically , by solving systems of linear equations @xcite as described in [ sub : prefix ] and the supplementary information . this system of equations leads to a particularly simple form for the generating function of mutually uncorrelated sequences , as explained in the supplementary information . an important criteria for selecting block addresses is to ensure that the corresponding dna primer sequences have prefixes with a gc content approximately equal to @xmath31 , and that the sequences are at large pairwise hamming distance . due to their applications in optical storage , codes that address related issues have been studied in a different form under the name of _ bounded running digital sum _ ( brds ) codes @xcite . a detailed overview of this coding technique may be found in @xcite . consider a sequence @xmath33 over the alphabet @xmath34 . we refer to @xmath35 as the running digital sum ( rds ) of the sequence @xmath36 up to length @xmath37 , @xmath38 . let @xmath39 denote the largest value of the running digital sum of the sequence @xmath36 . for some predetermined value @xmath40 , a set of sequences @xmath41 is termed a brds code with parameter @xmath42 if @xmath43 for all @xmath44 . note that one can define non - binary brds codes in an equivalent manner , with the alphabet usually assumed to be symmetric , @xmath45 , and where @xmath46 . a set of dna sequences over @xmath47 may be constructed in a straightforward manner by mapping each @xmath48 symbol into one of the bases @xmath49 and @xmath50 into one of the bases @xmath51 , or vice versa . alternatively , one can use brds over an alphabet of size four directly . to address the constraints c1-c2 , one needs to construct a large set of brds codewords at sufficiently large hamming distance from each other . via the mapping described above , these codewords may be subsequently translated to dna sequences with a gc content approximately equal to @xmath31 for all sequence prefixes , and at the same hamming distance as the original sequences . let @xmath52 be the parameters of a brds error - correcting code , where @xmath53 denotes the number of codewords of length @xmath54 , @xmath55 denotes the minimum distance of the code , while @xmath56 equals the code rate . for @xmath57 and @xmath58 , the best known brds - code has parameters @xmath59 , while for @xmath60 and @xmath61 , codes with parameters @xmath62 exist . for @xmath60 and @xmath58 , the best known brds code has parameters @xmath63 @xcite . note that each of these codes has an exponentially large number of codewords , among which a ( sufficiently ) large number of sequences satisfy the required correlation property c3 , discussed next , and the folding property c4 . codewords satisfying constraints c3-c4 were found by expurgating the brds codes via computer search . we describe next the notion of autocorrelation of a sequence and introduce the related notion of mutual correlation of sequences . it was shown in @xcite that the autocorrelation function is the crucial mathematical concept for studying sequences avoiding forbidden strings and substrings . in the storage context , forbidden strings correspond to the addresses of the blocks in the pool . in order to accommodate the need for selective retrieval of a dna block without accidentally selecting any undesirable blocks , we find it necessary to also introduce the notion of mutually uncorrelated sequences . let @xmath64 and @xmath65 be two words , possibly of different lengths , over some alphabet of size @xmath66 . the correlation of @xmath64 and @xmath65 , denoted by @xmath67 , is a binary string of the same length as @xmath64 . the @xmath68-th bit ( from the left ) of @xmath67 is determined by placing @xmath65 under @xmath64 so that the leftmost character of @xmath65 is under the @xmath68-th character ( from the left ) of @xmath64 , and checking whether the characters in the overlapping segments of @xmath64 and @xmath65 are identical . if they are identical , the @xmath68-th bit of @xmath67 is set to @xmath69 , otherwise , it is set to @xmath70 . for example , for @xmath71 and @xmath72 , @xmath73 , as depicted below . note that in general , @xmath74 , and that the two correlation vectors may be of different lengths . in the example above , we have @xmath75 . the autocorrelation of a word @xmath64 equals @xmath76 . in the example below , @xmath77 . @xmath78 _ a sequence x is self - uncorrelated if _ @xmath79 . a set of sequences @xmath80 is termed mutually uncorrelated if each sequence is self - uncorrelated and if all pairs of distinct sequences satisfy @xmath81 and @xmath82 . intuitively , correlation captures the extent to which prefixes of sequences overlap with suffixes of the same or other sequences . furthermore , the notion of mutual uncorrelatedness may be relaxed by requiring that only sufficiently long prefixes do not match sufficiently long suffixes of other sequences . sequences with this property , and at sufficiently large hamming distance , eliminate undesired address cross - hybridization during selection and cross - sequence assembly errors . we proved the following bound on the size of the largest mutually uncorrelated set of sequences of length @xmath54 over an alphabet of size @xmath83 . the bounds show that there exist exponentially many mutually uncorrelated sequences for any choice of @xmath54 , and the lower bound is constructive . furthermore , the construction used in the bound `` preserves '' the hamming distance ( see the supplementary information ) . [ thm : bounds ] suppose that @xmath84 is a set of @xmath85 pairwise mutually uncorrelated sequences of length @xmath54 . let @xmath86 denote the largest possible value of @xmath85 for a given @xmath54 . then [ lem : upper_bound ] @xmath87 as an illustration , for @xmath88 , the _ lower bound _ equals @xmath89 . the proof of the theorem is give in the supplementary information . it remains an open problem to determine the largest number of address sequences that jointly satisfy the constraints c1-c4 . we conjecture that the number of such sequences is exponential in @xmath54 , as the numbers of words that satisfy c1-c2 , c3 and c4 @xcite are exponential . exponentially large families of address sequences are important indicators of the scalability of the system and they also influence the rate of information encoding in dna . using a casting of the address sequence design problem in terms of a simple and efficient greedy search procedure , we were able to identify @xmath90 sequences for length @xmath88 that satisfy constraints c1-c4 , out of which @xmath26 pairs were used for block addressing . another means to generate large sets of sequences satisfying the constraints is via approximate solvers for the _ largest independent set problem _ @xcite . examples of sequences constructed in the aforementioned manner and used in our experiments are listed in the supplementary information . in the previous sections , we described how to construct address sequences that can serve as unique identifiers of the blocks they are associated with . we also pointed out that once such address sequences are identified , user information has to be encoded in order to _ avoid _ the appearance of any of the addresses , sufficiently long substrings of the addresses , or substrings similar to the addresses in the resulting dna codeword blocks . for this purpose , we developed new prefix - synchronized encoding schemes based on @xcite . to address the problem at hand , we start by introducing comma free and prefix - synchronized codes which allow for constructing codewords that avoid address patterns . a block code @xmath91 comprising a set of codewords of length @xmath92 over an alphabet of size @xmath93 is called _ comma free _ if and only if for any pair of not necessarily distinct codewords @xmath94 and @xmath95 in @xmath91 , the @xmath92 concatenations @xmath96 are not in @xmath91 @xcite . comma free codes enable efficient synchronization protocols , as one is able to determine the starting positions of codewords without ambiguity . a major drawback of comma free codes is the need to implement an exhaustive search procedure over sequence sets to decide whether or not a given string of length @xmath54 should be used as a codeword or not . this difficulty can be overcome by using a special family of comma free codes , introduced by gilbert @xcite under the name _ prefix - synchronized codes_. prefix - synchronized codes have the property that every codeword starts with a prefix @xmath97 , which is followed by a constrained sequence @xmath98 . moreover , for any codeword @xmath99 of length @xmath100 , the prefix @xmath101 _ does not appear _ as a substring of @xmath102 more precisely , the constrained sequences of prefix - synchronized codes avoid the pattern @xmath101 which is used as the address . due to the choice of mutually uncorrelated addresses at _ large hamming distance _ , we encode each information block by _ avoiding only one of the address sequences _ , used for that particular block . to explain how to perform encoding , assume that @xmath103 is a self - uncorrelated sequence . this guarantees that @xmath104 without loss of generality , let @xmath105 and @xmath106 and define @xmath107 for all @xmath108 . in addition , assume that the elements of @xmath109 are arranged in increasing order , say using the lexicographical ordering @xmath110 . we subsequently use @xmath111 to denote the @xmath112-th smallest element in @xmath109 , for @xmath113 for example , if @xmath114 then @xmath115 and @xmath116 next , we define a sequence of integers @xmath117 that satisfies the following recursive formula @xmath118 for an integer @xmath119 and @xmath120 , let @xmath121 be a length-@xmath122 ternary representation of @xmath123 . conversely , for each @xmath124 , let @xmath125 be the integer @xmath123 such that @xmath126 every integer @xmath127 can be mapped into a sequence of @xmath128 symbols @xmath129 via an encoding algorithm that consists of two parts : @xmath130 and @xmath131 . algorithm @xmath130 calls @xmath131 and returns the concatenation of @xmath101 and @xmath131 . the steps of the encoding procedure are listed in algorithm 1 , where @xmath132 , and where @xmath54 denotes the length of the sequence @xmath101 . the decoding steps are described in the same chart . @rrrrrrr & & & & & & & & & & & & [ table : alg ] the following theorems are proved in the supplementary information . [ thm : encoding ] @xmath133 is a prefix - synchronized codeword . [ thm : decoding ] [ thm : decodable]the algorithm outputs a uniquely decodable string , for any @xmath134 a simple example describing the encoding and decoding procedure for the short address string @xmath135 , which can easily be verified to be self - uncorrelated , is provided in the supplementary information . the previously described @xmath130 algorithm imposes no limitations on the length of a prefix used for encoding . this feature may lead to unwanted cross hybridization between address primers used for selection and the prefixes of addresses encoding the information . one approach to mitigate this problem is to perturb long prefixes in the encoded information in a controlled manner . for small - scale random access / rewriting experiments , the recommended approach is to first select all prefixes of length greater than some predefined threshold . afterwards , the first and last quarter of the bases of these long prefixes are used unchanged while the central portion of the prefix string is cyclically shifted by half of its length . for example , for the address primer @xmath136 , the suffix @xmath137 produced by @xmath130 maps to @xmath138 . the process of shifting applied to this string is illustrated below : @xmath139 for an arbitrary choice of the addresses , this scheme may not allow for unique decoding @xmath130 . however , there exist simple conditions that can be checked to eliminate primers that do not allow this transform to be `` unique '' . given the address primers created for our random access / rewriting experiments , we were able to uniquely map each modified prefix to its original prefix and therefore uniquely decode the readouts . as a final remark , we would like to point out that prefix - synchronized coding also supports error - detection and limited error - correction . error - correction is achieved by checking if each substring of the sequence represents a prefix or `` shifted '' prefix of the given address sequence and making proper changes when needed . we described a new dna based storage architecture that enables accurate random access and cost - efficient rewriting . the key component of our implementation is a new collection of coding schemes and the adaptation of random - access enabling codes from classical storage systems . in particular , we encoded information within blocks with unique addresses that are prohibited to appear anywhere else in the encoded information , thereby removing any undesirable cross - hybridization problems during the process of selection and amplification . we also performed four access and rewriting experiments without readout errors , as confirmed by post - selection and rewriting sanger sequencing . the current drawback of our scheme is high cost , as synthesizing long dna blocks is expensive . cost considerations also limited the scope of our experiments and the size of the prototype , as we aimed to stay within a budget comparable to that used for other existing architectures . nevertheless , the benefits of random access and other unique features of the proposed system compensate for this high cost , which we predict will decrease rapidly in the very near future . this work was partially supported by the strategic research initiative of university of illinois , urbana - champaign , and the nsf stc on science of information , purdue university . a provisional patent for rewritable , random - access dna - based storage was filed with the university of illinois in november 2014 . 10 [ 1]#1 url@samestyle [ 2]#2 [ 2]l@#1=l@#1#2 c. bancroft , t. bowler , b. bloom , and c. t. clelland , `` long - term storage of information in dna . '' _ science ( new york , ny ) _ , vol . 293 , no . 5536 , pp . 17631765 , 2001 . j. davis , `` microvenus , '' _ art journal _ , vol . 55 , no . 1 , pp . 7074 , 1996 . g. m. church , y. gao , and s. kosuri , `` next - generation digital information storage in dna , '' _ science _ , vol . 6102 , pp . 16281628 , 2012 . n. goldman , p. bertone , s. chen , c. dessimoz , e. m. leproust , b. sipos , and e. birney , `` towards practical , high - capacity , low - maintenance information storage in synthesized dna , '' _ nature _ , 2013 . r. n. grass , r. heckel , m. puddu , d. paunescu , and w. j. stark , `` robust chemical preservation of digital information on dna in silica with error - correcting codes , '' _ angewandte chemie international edition _ , vol . 54 , no . 8 , pp . 25522555 , 2015 . m. g. ross , c. russ , m. costello , a. hollinger , n. j. lennon , r. hegarty , c. nusbaum , and d. b. jaffe , `` characterizing and measuring bias in sequence data , '' _ genome biol _ 14 , no . 5 , p. r51 , 2013 . g. d. cohen and s. litsyn , `` dc - constrained error - correcting codes with small running digital sum , '' _ information theory , ieee transactions on _ , vol . 37 , no . 3 , pp . 949955 , 1991 . m. blaum , s. litsyn , v. buskens , and h. c. van tilborg , `` error - correcting codes with bounded running digital sum , '' _ ieee transactions on information theory _ , vol . 39 , no . 1 , pp . 216227 , 1993 . e. gilbert , `` synchronization of binary messages , '' _ information theory , ire transactions on _ , vol . 6 , no . 4 , pp . 470477 , 1960 . h. packer , `` gblocks gene fragments , related decoded articles , '' 2014 . a. v. bryksin and i. matsumura , `` overlap extension pcr cloning : a simple and reliable way to create recombinant plasmids , '' _ biotechniques _ , vol . 48 , no . 6 , p. 463 , 2010 . s. c. schuster , `` next - generation sequencing transforms today s biology , '' _ nature methods _ , vol . 5 , no . 1 , pp . 1618 , 2008 . k. a. s. immink , _ codes for mass data storage systems_.1em plus 0.5em minus 0.4emshannon foundation publisher , 2004 . h. morita , a. j. van wijngaarden , and a. han vinck , `` on the construction of maximal prefix - synchronized codes , '' _ information theory , ieee transactions on _ , vol . 42 , no . 6 , pp . 21582166 , 1996 . o. milenkovic and n. kashyap , `` on the design of codes for dna computing , '' in _ coding and cryptography_.1em plus 0.5em minus 0.4em springer , 2006 , pp . 100119 . rouillard , m. zuker , and e. gulari , `` oligoarray 2.0 : design of oligonucleotide probes for dna microarrays using a thermodynamic approach , '' _ nucleic acids research _ , vol . 31 , no . 12 , pp . 30573062 , 2003 . l. j. guibas and a. m. odlyzko , `` maximal prefix - synchronized codes , '' _ siam journal on applied mathematics _ , vol . 35 , no . 2 , pp . 401418 , 1978 . berman and m. frer , `` approximating maximum independent set in bounded degree graphs . '' in _ soda _ , vol . 94 , 1994 , pp . 365371 . r. g. gallager , `` low - density parity - check codes , '' _ information theory , ire transactions on _ , vol . 8 , no . 1 , pp . 2128 , 1962 . 1 . encoding wikipedia entries a working example ( section [ sec : working ] ) . 2 . proofs of theorems ( section [ sec : proofs ] ) . 3 . address sequences ( section [ sec : addresses ] ) . 4 . example of encoding and decoding procedure ( section [ sec : encdec ] ) . experimental synthesis , access and rewrite of dna storage sequences ( section [ sec : experiments ] ) . 6 . hybrid dna - based and classical storage ( section [ sec : hybrid ] ) . in this section we describe the data format used for encoding two files of size @xmath10 kb containing the introductory sections of wikipedia pages of six universities : berkeley , harvard , mit , princeton , stanford , and university of illinois urbana - champaign . there were @xmath140 words in the text , out of which @xmath141 were distinct . note that in our context , words are elements of the text separated by a space . for example , university and university . are counted as two different words , while urbana - champaign is counted as a single word . these @xmath140 words were mapped to @xmath142 dna blocks of length @xmath12 bps , as we grouped six words into fragments , and combined @xmath15 fragments for prefix - synchronized encoding . table [ tab:1_comp ] provides the word counts in the files and encoding lengths ( in bits ) of the of the outlined procedure . assume that instead of using a prefix - synchronized code , we used classical ascii encoding without compression to encode the same wikipedia pages . the total number of characters in the text equals @xmath143 , and each character is mapped to a binary string of length @xmath144 . hence , one would need @xmath145 bits to represent the data , which is equivalent to @xmath146 dna blocks of length @xmath12 bps if we set aside two unique address flags for the blocks . as one can see , prefix - synchronized codes offer an almost @xmath147-fold improvement in description length compared to ascii encoding . this comes at the cost of storing a larger dictionary , as one encodes words rather than symbols of the alphabet . for the working example , one would require roughly @xmath148-times larger dictionaries , as there are @xmath149 words with an average of @xmath150 symbols per word . this increased in the dictionary is not a significant problem , as only one copy of the dictionary is ever needed . bits / distinct symbol & # bits + characters & @xmath151 & @xmath152 & @xmath153 & @xmath154 + words & @xmath149 & @xmath141 & @xmath15 & @xmath155 + _ proof of theorem [ thm : bounds]_. the proof consists of two parts . first , we prove the upper bound on @xmath86 in lemma 1 , and then proceed to prove a lower bound in lemma 2 . recall that @xmath156 denotes the largest possible size for a set of mutually uncorrelated words of length @xmath54 . * lemma 1 . * [ lem : upper_bound ] _ let @xmath156 the largest set of distinct mutually uncorrelated sequences of length @xmath54 . then @xmath157 _ _ proof : _ to prove the lemma , let us introduce some terminology . let @xmath158 stand for the hamming distance between two words , and define the hamming ball of radius @xmath55 around a point @xmath159 in @xmath160 as @xmath161 furthermore , let @xmath162 denote the set of sequences correlated with @xmath159 that are also at most at hamming distance @xmath55 from @xmath159 . we claim that for @xmath163 , one has @xmath164 to prove the result , assume without loss of generality that @xmath159 starts with the symbol @xmath165 , i.e. , @xmath166 . next , consider two scenarios regarding the structure of @xmath166 : * @xmath167 in this case , any word @xmath168 in @xmath169 that starts with @xmath170 or ends with @xmath165 is an element of @xmath171 + let @xmath172 and @xmath173 + clearly , @xmath174 and @xmath175 therefore , @xmath176 * @xmath177 in this case , any word @xmath168 in @xmath169 which starts or ends with @xmath165 is also an element of @xmath178 . using an argument similar to the one described for the previous scenario , one can show that @xmath179 + moreover , it is straightforward to see that @xmath180 for any mutually uncorrelated set @xmath84 of size @xmath85 , we have @xmath181 , for @xmath182 this implies that @xmath183 at the same time , the previous claim suggests that @xmath184 therefore , @xmath185 , which completes the proof . * lemma 2 . * _ let @xmath156 the largest set of distinct mutually uncorrelated sequences of length @xmath54 . then @xmath186 _ _ proof : _ for simplicity , assume that @xmath85 is even . given a mutually uncorrelated set @xmath187 with words of length @xmath54 and over the alphabet @xmath188 , partition @xmath84 into two arbitrary sets @xmath189 and @xmath190 of equal size , say @xmath191 and @xmath192 we argue that @xmath193 is a mutually uncorrelated set with words of length @xmath194 . * first , we show that the elements in @xmath53 are self - uncorrelated : for an arbitrary element @xmath195 , we have @xmath196 since the two sequences @xmath197 are mutually uncorrelated , one can easily verify that @xmath198 for @xmath199 moreover , since @xmath200 , it holds that @xmath201 this establishes the claim . * next , we argue that any two distinct elements in @xmath53 are uncorrelated : for any two distinct elements @xmath202 and @xmath203 in @xmath53 , one can show that @xmath204 , for @xmath205 . in addition , @xmath206 implies that @xmath207 this completes the proof . as a result , given a mutually uncorrelated set @xmath208 , where @xmath209 one can construct another mutually uncorrelated set @xmath210 , where @xmath211 . therefore , @xmath212 observing that for @xmath213 it is possible to construct the following set of @xmath15 mutually uncorrelated sequences @xmath214 establishes the base of a recursive procedure which gives @xmath215 note that this bound is constructive , and the concatenation procedure preserves normalized minimum hamming distances . we now turn our attention to prefix - synchronized coding , and describe a number of results relevant for our subsequent discussion . [ @xcite ] given a positive integer @xmath92 , chose the unique integer @xmath216 so that @xmath217 satisfies @xmath218 then , the maximal prefix - synchronized code of length @xmath92 has cardinality @xmath219 for a prefix of the form @xmath220 . note that the above results indicate that codes avoiding one address sequence represent an exponentially large family of binary sequences . we prove a similar result for the case of @xmath21-ary sequences that avoid a set of @xmath85 mutually uncorrelated sequences . to establish the claim , we need the following definitions . let @xmath221 be an integer sequence over a finite alphabet . define the generating function of the sequence @xmath222 [ thm : generating ] suppose that @xmath84 is a set of mutually uncorrelated sequences of length @xmath54 over the alphabet @xmath223 . let @xmath224 , with @xmath225 , be the number of strings of length @xmath92 over @xmath129 that do not contain substrings in @xmath208 . then @xmath226 where @xmath227 is the generating function of the sequence @xmath228 . _ proof of theorem [ thm : generating]_. the result is a direct consequence of theorem 4.1 of @xcite . for @xmath229 let @xmath230 denote the number of strings of length @xmath54 over @xmath129 that contain no element of @xmath208 , except for a single copy of @xmath231 at the right - hand side of the string . let @xmath232 be the generating function of @xmath230 . then , we have the following system of equations that holds for the two sets of aforementioned functions : @xmath233 by using the fact that @xmath234 , for @xmath235 , and @xmath236 , for @xmath237 , one can show that @xmath238 the result follows by replacing ( [ eq:1 ] ) into the first line of ( [ eq:2 ] ) . as the dominant pole of the generating function is close to @xmath21 , the number of sequences avoiding a set of mutually uncorrelated sequences grows roughly as @xmath239 . _ proof of theorem [ thm : encoding]_. since @xmath101 is self - uncorrelated , we need to show that this string is not contained in the output of @xmath131 , where the output of @xmath131 equals @xmath240 for some input @xmath241 and @xmath242 . consequently , if @xmath101 is a substring of the output of @xmath131 , then the last symbol of @xmath101 ( recall that we assumed this symbol to be @xmath243 ) has to appear in one of the following three positions : * the symbol appears in @xmath244 for a unique @xmath245 in this case , there exists a suffix of @xmath101 appearing as a prefix of @xmath246 this contradicts our assumption that @xmath101 is self uncorrelated . * the symbol appears in @xmath247 for a unique @xmath245 this contradicts our assumption that @xmath248 * the symbol appears in @xmath249 this contradicts our assumption that @xmath243 does not appear in @xmath250 . therefore , the string @xmath101 does not appear as a substring in the output of @xmath251 , which completes the proof . _ proof of theorem [ thm : decoding]_. it suffices to show that the output of @xmath131 is uniquely decodable . we use induction arguments to establish this result . for the basis step , by the definition of the output of @xmath252 , it is straightforward to show that @xmath131 returns the encoding @xmath253 , which represents a one - to - one mapping from @xmath254 to @xmath255 whenever @xmath256 . for the inductive step , we assume that the result is true for all @xmath257 , as well as for all @xmath258 , and show that it is consequently true for @xmath259 . for @xmath259 , @xmath131 returns @xmath260 for some integer values @xmath261 and for some @xmath262 where @xmath263 therefore @xmath264 is uniquely decodable if and only if @xmath265 and @xmath266 are unique . since sequences of the form @xmath267 are prefix - free one can uniquely identify both @xmath268 and @xmath269 moreover @xmath270 , hence by the induction hypothesis it follows that @xmath266 is also uniquely decodable from @xmath271 . hence , @xmath264 can be uniquely decoded . b1-forward & @xmath272 + b1-reverse & @xmath273 + gblock - b1-reverse & @xmath274 + b1-su1-reverse & @xmath275 + b1-sd1-forward & @xmath276 + b2-forward & @xmath277 + b2-reverse & @xmath278 + gblock - b2-forward & @xmath279 + b2-su1-reverse & @xmath280 + b2-sd1-forward & @xmath281 + b3-forward & @xmath282 + b3-reverse & @xmath283 + b3-su1-reverse & @xmath284 + b3-su2-forward & @xmath285 + b3-sd2-reverse & @xmath286 + b3-sd2-forward & @xmath287 + consider the following set of strings of length @xmath13 , @xmath288 with gc content equal to @xmath31 , i.e. , @xmath32 gc bases . the sequences are mutually uncorrelated and at hamming distance exactly @xmath32 from each other . the sequences do not exhibit secondary structures at room temperature , as verified by the mfold and vienna packages . we used these addresses for a very small - scale , proof - of - concept random access / rewriting experiment of a @xmath21 kb file . in the large scale random access / rewriting experiment described in section [ sec : experiments ] , we used different address sequences for the two flanking ends of the @xmath12 bps blocks . the sequences we synthesized include : @xmath289 the pairs of sequences were used to flank the two ends of the data blocks . only the addresses on the left were used for subsequent prefix - synchronized coding . the sequences on the left - hand side of the pairing have `` interleaved '' @xmath290 and @xmath291 bases for example , they all start with @xmath292 . this ensures a gc balancing property for the prefixes of the addresses . in this section , we illustrate the encoding and decoding procedure for the short address string @xmath135 , which can easily be verified to be self - uncorrelated . more precisely , we explain how to compute a sequence of integers @xmath293 , described in the main body of the paper . as before , @xmath54 denotes the length of the address string , which in this case equals five . one has @xmath294 the algorithm @xmath295 produces : & 550=0g_5,7 + 550 & + & ( p,8,550)=(p,7,550 ) & + & 550=0g_5,6 + 550 & + & ( p,7,550)=(p,6,550 ) & + & 550=2g_5,5 + 0g_5,4 + 16 & + & ( p,6,550)=(p,4,16 ) , & + & 16=03 ^ 3 + 13 ^ 2 + 23 ^ 1 + 13 ^ 0 & + & ( p,4,16)= , & + & ( p,8,550)= & + when running @xmath296 on the encoded output @xmath297 , the following steps are executed : & ( p,)=0g_5,7 & + & + ( p , ) & + & ( p,)=0g_5,6 & + & + ( p , ) , & + & ( p,)=2g_5,5 + 0g_5,4 & + & + ( p , ) & + & ( p,)=16 & + & ( p , ) = 2g_5,5 + 16=550 & a total of @xmath298 sequences of length @xmath12 bps each were designed to encode information retrieved from the berkeley , harvard , mit , princeton , stanford , and uiuc wikipedia page in 2014 . except for sequence # 4 , which was rejected due to the complexity of its secondary structure , all sequences were synthesized by idt ( integrated dna technologies ) . in addition , @xmath298 corresponding address primers were synthesized by the same company . the address sequences of the blocks are listed in section [ sec : addresses ] . as a proof of concept , we performed a number of selection and editing experiments . these include selecting individual blocks and rewriting one of its sections , selecting three blocks and rewriting three sections in each , two close to the flanking ends , and one in the middle . the edits involved information about the budget of the institutions at a given year of operation . detailed information about the original sequences and their rewritten forms is given in the following sections . we denoted the blocks on which we performed selection and editing by b1 , b2 , and b3 . the primers used for performing the edits in the blocks are listed in table [ tab:2 ] . note that two primers were synthesized for each rewrite , for the forward and reverse direction . in addition , two different editing ( mutation ) techniques were used , gblock and overlap - extension ( oe ) pcr ; gblocks are double - stranded genomic fragments that are frequently used as primers , for gene construction or for mediated genome editing . an illustration of editing via gblocks is shown in fig . [ fig:2 ] . on the other hand , oe - pcr is a variant of pcr used for specific dna sequence editing via point mutations or splicing . an illustration of the procedure is given in fig . [ fig:2 ] . to demonstrate the plausibility of a cost efficient method for editing , oe - pcr was used with general primers ( @xmath29 bps ) only . for edits shorter than @xmath30 bps , the mutation sequences were designed as overhangs in primers . then , the three pcr products were used as templates for the final pcr reaction involving the entire @xmath12 bps rewrite . all @xmath298 linear @xmath12 bps fragments were mixed , and the mixture was used as a template for pcr amplification and selection of the b1 , b2 and b3 sequences . the results of selection are shown in fig [ fig:1 ] , where three banks of size @xmath12 bps are depicted . these banks indicate that sequences of the correct length were isolated . subsequent sequencing confirmed that the sequences were indeed the user requested b1 , b2 and b3 strands . a summary of the experiments performed is provided in table [ tab : summary ] . sequence samples & length of the edited region ( in bps ) & selection accuracy / readout error percentage & description of editing method + b1-m - gblock & @xmath28 & @xmath13 & @xmath299 & gblock method + b1-m - pcr & @xmath28 & @xmath13 & @xmath299 & oe - pcr method + b2-m - gblock & @xmath28 & @xmath300 & @xmath299 & gblock method + b2-m - pcr & @xmath28 & @xmath300 & @xmath299 & oe - pcr method + b3-m - gblock & @xmath28 & @xmath301 & @xmath299 & gblock method + b3-m - pcr & @xmath28 & @xmath301 & @xmath299 & oe - pcr method + a ) schematic depiction of the editing method using gblocks . b ) detailed description of the generation of the mutation . four sequences ( ranging in length from @xmath302 to @xmath303 bps ) containing the entire edit region were gblock synthesized from idt . the remaining parts of the @xmath12 bps sequences were pcr amplified . a homology in at least 30 bps between the flanking end sequence of the blocks and the corresponding end of the gblock fragment was created . by one oe - pcr , the desired edits were generated in a one - pot matter . ] the unedited b1_original ( b1 ) sequence is of the form : @xmath304 where the bases written in red represent the regions we edited . pcr of 1000 bps sequences - b1 , b2 , b3 from a mixture of 26 sequences . ] the edited b1_mutation ( b1_m ) sequence reads as : @xmath305 with rewrites listed in red . since a gblock of length longer than @xmath306 bps was needed , it was more costly to synthesize the gblock and perform rewriting than to directly re - synthesizing the whole block . hence , the gblock method was not used in this case . illustration of the process of generating the b1 edit / mutation using general primers . ] one pair of primers was designed to pcr amplify the first portion of the sequence b1-m . for the forward direction , the primer was @xmath307 while for the reverse direction , the primer was @xmath308 the second part of the sequence was pcr amplified by using the forward direction primer @xmath309 and reverse direction primer @xmath310 both pcr reactions used the sequence b1 as template . two such pcr products are shown in fig . [ fig:4 ] , indicating that the correct length products were isolated in each reaction . oe - pcr was performed in a @xmath311 ul reaction volume containing the two aforementioned pcr products without primers for the first @xmath28 cycles and the products with primers ( b1 primers in table [ tab:2 ] ) for the later @xmath312 cycles . a single bank with correct size of @xmath12 bps was obtained ( see fig . [ fig:4 ] ) . the unedited b2_original ( b2 ) sequence is of the form : @xmath313 a schematic depiction of the process of generating the b2 mutation using standard 60 bps primers . ] the edited b2_mutation ( b2_m ) sequence is : @xmath314 where , as before , red letters were used to indicate the rewritten region . a @xmath302 bps sequence , containing the entire edited region and the b2 string , was gblock synthesized by idt . another part of b2 was pcr amplified using the forward primer @xmath315 and reverse primer @xmath316 the b2 sequence served as a template . see fig . [ fig:4 ] for an illustration . over extension pcr ( oe - pcr ) was performed in a @xmath311 ul reaction volume containing the above @xmath302 bps gblock product and pcr products without primers for the first @xmath28 cycles and with b2 forward and reverse primers listed in table [ tab:2 ] for the subsequent @xmath312 cycles . the pcr product was deposited on a gel substrate and the correct @xmath12 bps band was obtained as shown in fig . [ fig:5 ] . one pair of primers was designed to pcr amplify the first part of the sequence b2-m , with forward primer @xmath317 and reverse primer @xmath318 the second part was pcr amplified by the forward primer @xmath319 and reverse primer @xmath320 both pcrs used b2 as a template . two pcr products are shown in fig . [ fig:5 ] . pcr products of b1 and b2 . ] pcr products of b3 . ] the unedited original b3 sequence equals : @xmath321 scheme for generating the b3 edits using standard 60 bps primers . ] the edited sequence b3_m mutation sequence is : @xmath322 two sequences , the @xmath323 bps sequence containing the first mutation region and the second @xmath323 bps sequence containing the second mutation region , were gblock synthesized by idt . there was a @xmath324 bps overlap between the two gblocks . oe - pcr was performed in a @xmath311 ul reaction volume containing the above two @xmath323 bps gblock products without primers for the first @xmath28 cycles and additional b3 forward and reverse primers listed in table [ tab:2 ] for the subsequent @xmath312 cycles . the pcr product was deposited on a gel substrate and the correct @xmath12 bps band was obtained . one pair of primers was designed to pcr amplify the first part of the sequence b2-m , using @xmath325 in the forward direction and @xmath326 in the reverse direction . the second part was pcr amplified in the forward direction by using the primer @xmath327 and in the reverse direction by using the primer @xmath328 the third part was pcr amplified by the forward direction primer @xmath329 and reverse direction primer @xmath330 all three pcrs used the sequence b3 as the template . all three pcr products are shown in fig . [ fig:8 ] . oe - pcr was performed in a @xmath311 ul reaction volume containing the above three pcr products without primers for the first @xmath28 cycles and with b3 primers listed in table [ tab:2 ] for the subsequent @xmath312 cycles . a single bank of correct size @xmath12 bps was obtained ( see fig . [ fig:9 ] ) . the generated pcr products of @xmath12 bps edits from the gblock method , involving b1-gblock , b2-gblock and b3-gblock . ] the generated pcr products of 1000bps sequence editing for the oe - pcr based method , and sequences b1-pcr , b2-pcr and b3-pcr . ] correctness of the synthesized edited regions was confirmed via dna sanger sequencing as follows . the pcr products of the gblock method and the oe - pcr method were named b1-m - gblock , b2-m - gblock , b3-m - gblock and b1-m - pcr , b2-m - pcr , b3-m - pcr , respectively . all final mutations / edits of pcr products were purified using the qiagen gel purification kit . the purified @xmath12 bps edited sequences were blunt - ligated to the vector named @xmath331-blunt ( fig . [ fig:10 ] ) using the zero blunt pcr cloning kit and following the manufacturers protocol . five colonies of each pcr - blunt - mutation were sent to actg , int . sequencing was performed using two universal primers : m13f_20 ( for the reverse direction ) and m13r ( for the forward direction ) . bi - directional sequencing was performed in order to ensure that the entire @xmath12 bps block was completely covered . map and features of pcr - blunt vector ( life technologies ) . ] in our small - scale experiments , sanger sequencing produced two erroneous symbols in one strand which we were able to correct using prefix matching . one possible problem that may arise in large scale dna - storage systems involving millions of blocks is erroneous sequencing which may not be corrected via prefix matching . in current high throughput sequencing technologies , such as illumina hiseq or miseq , the dominant sources of errors are substitutions . due to our word grouping scheme , such substitution errors can not cause catastrophic error propagation , but may nevertheless accumulate as the number of rewrite cycles increases . in this case , prefix matching may not suffice to correct the errors and more sophisticated coding schemes need to be used . unfortunately , adding additional parity - check symbols into the prefix - encoded data stream may cause problems as the parities may violate the prefix properties and dis - balance the gc content . furthermore , every time rewriting is performed , the parity - checks will need to be updated , which incurs additional cost for maintaining the system . a simple solution to this problem is a hybrid scheme , in which the bulk of the information is stored in dna media , while only parity - checks are stored on a classical device , such as flash memory . given that the current error - rate of short - read sequencing technologies roughly equals @xmath332 , the most suitable codes for performing this type of coding are low - density parity - check codes @xcite . these codes offer excellent performance in the presence of a large number of errors and are decodable in linear time .

outflows are ubiquitous among quasars . they manifest themselves as blue - shifted broad and narrow absorption lines in uv and x - rays ( e.g. , boksenberg et al . 1978 ; turnshek et al . 1980 ; weyman et al . 1991 ) . the broad absorption lines ( bals ) with widths of around a few thousands to a few ten thousands km s@xmath0 appear in 15 - 25% of all quasars ( e.g.,weymann et al . 1991 ; knigge et al . 2008 ) , and intrinsic narrow absorption lines ( nals ) with widths around hundreds to one thousand km s@xmath0 in 40 - 50% of quasars ( nester et al . almost all known bal qsos show broad high ionization absorption lines , such as c iv , n v ( weymann et al . 1991 ; cf , mrk 231 , veilleux et al . 2013 ) , while only a small fraction ( about 25% ) of bal qsos display also low ionization broad absorption lines ( lobals , weymann et al . 1991 ; zhang et al . 2010 ; c.f . urrutia et al . 2009 ) , commonly mg ii . a yet smaller subset of lobal qsos ( @xmath1% of all quasars ) exhibit broad fe ii absorption lines ( hereafter , felobals ; trump et al . 2006 ; c.f . , dai et al . 2012 ) . there are two different scenarios for bal and non - bal quasars or for a subclass of bal qsos . in the first scenario , the presence or the absence of bals , or more generally , the manifestation of different subclasses of bals , are attributed to the different viewing angles of the same intrinsic population ( weymann et al . 1991 ) . while in the second scenario , lobal and non lobal qsos are at different evolutionary stages ( boroson & meyers 1992 ; voit et al . 1993 ; hall et al . in other words , only a small fraction of quasars possess low - ionization outflows . early studies showed that the emission lines and broad - band spectral energy distributions ( sed ) are very similar for hibal and non bal quasars ( weymann et al . 1991 ) , suggesting the unification scheme . hamann , korista & morris ( 1993 ) constrained the global covering factor to less 20% from the fraction of resonant scattering light . however , systematic trends in the sed and emission lines have been revealed with a large sample of bal , lobal in particular , and a non - bal quasar sample from sdss ( e.g. , zhang et al . 2010 ; baskin , laor & hamann 2013 ; farrah et al . 2010 ; wang et al . 2013 ; and reference therein ; c.f . , lazarova et al . 2012 ) , suggesting that the global covering factor and properties of lobals depend on the intrinsic properties of the system . feedback through these outflows is now widely believed to be a key physical process in the formation of massive galaxies ( fabian 2012 ) . however , the mass loss rate and kinetic power , as well as the size , of outflows are still poorly constrained for the majority of bal quasars ( lucy et al . 2014 ) . to estimate these parameters from absorption lines , one needs to know the global covering factor , column density and the gas density or distance . the gas column density can be inferred from the column densities of a set of ions through photoionization modeling . however , the gas density diagnostics for absorption lines are available for only a rare subset of bal qsos with measured absorption lines from low - lying excitation levels of heavy elements ( de kool et al . 2001 ; korista et al . 2008 ; dunn et al . 2012 ) . previous measurements of density in a dozen of bals , mostly felobals , quasars , indicate that bal region ( balr ) is extended from 100 parsec to several kpc from quasars ( de kool et al . 2001 ; korista et al . 2008 ; borguet et al . 2013 ; chamberlain et al . the large size implies a very large mass outflow rate ( @xmath2 m@xmath3 yr@xmath0 ) and large kinetic power ( a few to 10 percent of the total luminosity ) if the outflows cover a substantial fraction of the sky . however , absorption lines can only probe the gas on the line of sight , so the global covering factor of absorbing gas is virtually unknown in an individual object . because these quasars are so rare , the average global covering factor must be very small , according to the unification scheme , but if they are at a special evolution stage , on the other hand , the covering factor can be substantial . interestingly , liu et al . ( 2015 ) detected extended outflow components in two low z agns on the physical sizes consistent with those inferred from ultraviolet absorption lines . in this paper , we report four quasars , including one that was noted already by ross et al . ( 2015 ) , with peculiar ultraviolet fe ii spectra . they were discovered during the course of studying a sample of heavily obscured quasars in wise : @xmath4 . we perform an analysis of the emission lines and the broad band sed , and present evidence that the fe ii emission lines in these objects are dominated by continuum fluorescence pumping in the extended outflows , so they can be used to constrain the covering factor and size of balr of lofebal . the paper is organized as follows . in section 2 , we present an analysis of the optical spectra and broad band sed . in section 3 , we discuss the nature of fe ii emission and continuum emission , and a brief conclusion is given in section 4 . spectral models of fluorescent fe ii emission are presented in an appendix . the optical spectra are shown in figure 1 . one of these quasars , sdss j153542.4 + 090341 has already been noted in ross et al . ( 2015 ) ; and another one ( sdss j163246.6 + 340526 ) appeared on the list of unusual quasars in meusinger et al . the most striking features of these objects are weak continua , strong ultraviolet fe ii spikes and mg ii emission lines , and a relatively narrow mg ii line . individual objects are described and analyzed as follows . * sdss j153542.4 + 090341 : the boss spectrum shows three prominent fe ii spikes due to uv1,2,3 and 62,63 , and mg ii , al iii and c iv lines . he ii 1640+o iii1663 and c iii ] are weak with respective to these lines . we measure an extreme fe ii spike / gap ratio of 8.5 . a narrow fe ii 78 is also visible . in fact , the equivalent widths of c iii ] and he ii are consistent with those of i zw 1 ( see figure 1 , also laor et al . 1997 ) , while those of fe ii , mg ii , al iii and c iv are a factor of 10 - 20 times larger . this indicates that the uv spectrum consists of two components , a weak i zw 1 component and another component dominated by strong emission lines , fe ii , c iv , al iii and mg ii . c iv shows two components , a narrow component at the systematic redshift , derived from [ o ii ] , and a blue - shifted component . alternatively , c iv may be split into two components due to a blue shifted narrow absorption line . the mg ii doublet can be well modeled with two gaussians . the best fit converges to a width of @xmath5 km s@xmath0 and no shift in velocity ( 2@xmath6500 km s@xmath0 ) . [ oii ] appears fairly narrow overall , but with weak broader wings . [ ne iii]@xmath73869 looks similar with a narrow core and extended wings . however , both lines are severely contaminated by the subtraction of sky lines . * sdss j100552.6 + 493448 : in addition to strong ultraviolet fe ii spikes , this object displays very strong high - ionization forbidden and permitted lines : [ o iii]@xmath74363 , [ neiii]@xmath83869,3987 , [ nev]@xmath83346,3426 , [ neiv]@xmath72422 , [ ciii]@xmath71906,1909 , as well as low ionization lines @xmath9 , @xmath10 , [ oii]@xmath83726,3729 , cii]@xmath72326 , mg ii , and al iii@xmath81855 , 1863 . in fact , the spectrum can be reproduced by a supposition of broadened sdss j1535 + 0903 emission lines and the narrow lines mentioned above ( see figure 1 ) . the al iii strength can be well accounted for by the sdss j1535 + 0903 component . to first order , all but [ o ii ] , have the same width , around fwhm@xmath113150 km s@xmath0 , using a single gaussian fit . [ o ii ] is considerably narrower than other lines , probably from an extended star - forming region . there is also a weak broad wing in [ o ii ] , which may be similar to other lines . the weak line to the red side of mg ii can be either fe ii or mg i. [ o iii]@xmath74363 is stronger than h@xmath12 according to the single gaussian fit with their centers tied to other lines . unfortunately , the boss spectrum does not cover [ o iii]@xmath84959,5007 , which should impose important constraints on the gas density and gas temperature in combination with [ o iii]@xmath74363 . another constraint on the density comes from the fit to 1900 blending . we find that in the single gaussian model for each line the fit is significantly worse when the lines are assumed to be [ si iii ] and [ c iii ] with a @xmath13 with respect to models assuming they are si iii ] and c iii ] . a further fit with a mixture of c iii ] and [ c iii ] yields a [ c iii]/c iii]=0.11@xmath60.02 with a @xmath14 lower than pure c iii ] model . this gives a gas density @xmath15 @xmath16 ( osterbrock & ferland 2006 ) . the outflow has a considerably lower density than that in q 1321 + 058 ( wang et al . detailed photoionization modeling of the outflow component will be presented elsewhere . * sdss j163246.6 + 340526 : the spectrum is dominated by narrow ultraviolet fe ii lines . the fe ii spikes / gap ratio ( 3.2 ) is less extreme than in j1535 + 0903 . we adopt the redshift of narrow [ o ii ] as the systemic redshift . fe ii and mg ii appears fairly narrow . we fit the mg ii multiplet using two gaussian and obtain a best fit with an @xmath17 km s@xmath0 and a doublet ratio of @xmath18 with no systematic redshift . mg ii is slightly narrower than mg ii in i zw 1 ( laor et al . 1997 ) . interestingly , the observed spectrum in the 2100 - 3100 range can be reproduced approximately by the addition of a scaled i zw 1 spectrum with four - times stronger emission lines plus the fe ii spectrum of sdss j1535 + 0903 ( see figure 1c ) . [ o iii ] @xmath75007 is visible in the gap of the sky line forest , and has the same redshift as [ o ii ] . unfortunately , h@xmath19 is seriously affected by residuals of sky lines . * sdss j212546.9 + 004455 : unlike other objects , the continuum emission in this object is relatively strong . the narrow mg ii line has the same redshift as [ o ii ] , which is considered as the systematic redshift . a comparison with the fe ii template of i zw 1 and the spectrum of sdss j1535 + 0903 suggests the presence of two blue - shifted ultraviolet fe ii components . a broadened sdss j1535 + 0903-like component is shifted by 5700 km s@xmath0 and a i zw 1-like ultraviolet fe ii component is blueshifted by @xmath202850 km s@xmath0 . the i zw 1-like component has weaker mg ii , c iii , si iii and c iv lines but broader line widths than i zw 1 . we measure the fwhm of mg ii using the above two - component fit ( see also figure 1 ) . there appear narrow absorption lines of c iv at velocities of 900 and 2300 km s@xmath0 . but the signal to noise ratio is low . table 1 presents the emission - line parameters , including the width of mg ii and the equivalent widths of fe ii spikes , mg ii and al iii . to summarize , sdss j1535 + 0903 represents a prototype peculiar fe ii spectrum , while the spectra in other objects can be modeled as a combination of this and a i zw 1-like fe ii component . broad band seds of these quasars are shown in figure 2 . they are characterized by strong and steep mid - infrared emission , and a relatively weak ultraviolet and optical continuum . the photometry is strongly affected by the ultraviolet fe ii and mg ii . we corrected the emission - line flux using the sdss spectrum . we fit these seds using a model consisting of a clumpy torus plus a galaxy template , accounting for the host galaxy light . we use all galaxy templates , excluding agns , that are in the swire library ( polletta et al . the purpose of this fit is to give an estimate how much of the infrared emission comes from the agn heated dust and how this dust is distributed . the seds of clumpy tori were calculated for a wide range of parameters by hnig & kishimoto ( 2010 ) . in their models , the torus consists of optically thick clouds with a small total filling factor . the number density of clouds decreases outward in a power law form , and has a gaussian distribution in the vertical direction . the parameters of the models are the number of clouds along the line of sight on the equator ( @xmath21 ) , the radial dust - cloud distribution index ( @xmath22 ) , the subtending angle of the torus ( @xmath23 ) , the optical depth of individual clouds , and the inclination ( refer to hnig & kishimoto 2010 for details ) . the inner radius is fixed to the dust sublimation radius and the outer radius of the torus to 300 dust sublimation radii . the broad - band sed can be fitted fairly well with such a model . the best fits are shown in figure [ sedfit ] and the parameters are listed in table [ emissionline ] . we also compute the distribution of the agn fraction in the rest frame 3 - 15 @xmath24 m using all models weighted by the @xmath25 probability of each model and by the solid angle of the viewing angle , i.e. , @xmath26 . the results suggest that the infrared emission is dominated by a torus component , and that the galaxy component contributes less than 5% in all cases . the power - law index for the dust distribution is quite broad from @xmath27 . thus the steepness in the mid - infrared spectrum is due to weak host galaxy light in the near - infrared , indicating a high eddington ratio of the system , and an edge - on viewing angle . this result is probably a color selection effect , where we only choose objects with a very red infrared color . second , the half opening angle of the torus is small ( 45 to 30 degree ) , and the system is viewed at a relatively large inclination angle ( 65 - 75 degree ) , consistent with the large infrared to optical ratio . finally , torus models predict a uv flux higher than the observed ones if there is no internal extinction . the reddening to the reflected light can be due to diffuse interstellar dust further out in the host galaxy , which may be partially covering . note that our model does not include other optical or uv continuum sources , such as direct or electron scattered accretion disk light . if such components are present , then the contribution from the host galaxy should be even lower and the optical and uv continuum from the dusty torus should be highly extinguished . the large infrared to optical / uv ratio and the steep near to mid - infrared ratio indicates that the optical to uv , and even the near - infrared , emission from the accretion disk is obscured . assuming the mid - infrared emission , which is dominated by the dusty torus emission as suggested by sed fitting , is nearly isotropic , the continuum emission at rest frame 3000 is a factor of 30 - 300 weaker than in normal quasars ( refer to figure 1 ) . if the 3000 flux is attributed to the transmitted light , the dust extinction would be @xmath28 for an smc - like extinction curve . applying such dust extinction to the average uv spectrum of quasars , we will get a slope @xmath29 ( @xmath30 ) between 2000 and 3000 , which is much steeper than the observed one . adopting an average galactic extinction curve will yield only slightly smaller slope . this indicates that the observed uv continuum is reflected / scattered light . in fact , the residual ultraviolet flux can be well accounted for by reflected light from the clumpy dust torus itself as in the sed fitting . alternatively , the continuum can be due to electron scattering , which is known to be important in seyfert 2 galaxies ( e.g. , antonucci & miller 1985 ) . this suggests that the optical depth to the continuum emission region is much larger than indicated by the @xmath31 derived above . in either case , the continuum should be polarized , and future polarization observation could verify this . in the sed fit , we found that the dusty torus has a large subtending angle , i.e. , covering 71 to 87% of the sky . the large inclination angle suggests that our line of sight to the continuum source is blocked by the torus , consistent with weak optical and uv emission . in this case , the hot dust in the illuminated face of inner dust torus is also blocked , leading to a steep rise from the near to mid - infrared . so the dusty torus can naturally explain the obscuration and infrared sed . the sed fitting also suggests that the contribution of stellar light to the near and mid - infrared is small , which implies a high eddington ratio . once the direct continuum is obscured , the large line equivalent width can be understood if the emission - line region is more extended than the dusty torus . in a few nearby seyfert galaxies , the warm few hundred k dust which is responsible for mid - infrared emission is resolved into a thick structure of a few parsecs with a bolometric luminosity around @xmath32 erg s@xmath0 ( tristam et al . 2007 ; lpez - gonzaga et al . 2014 ; tristram et al . 2014 ) . note that this is not the outer radius of the torus , which might extend to several tens of parsecs . since the size of the torus is proportional to @xmath33 , with an infrared luminosity of @xmath34 erg s@xmath0 for our quasars , the true size will be at least several tens of parsecs . in either case , the emission line region should be larger than this . before making a detailed discussion the size of the obscurer , let us consider the constraint on the size of emission line region first . the emission line width ( fwhm ) , as fitted with mg ii , is in the range 900 - 8000 km s@xmath0 , which is broader than narrow lines in seyfert galaxies , but is comparable with or narrower than the broad lines . naively , this can be interpreted as lines that come from an intermediate region between the traditional blr and nlr ( hu et al . 2008 ; li et al . however , note that sdss j1005 + 4934 also displays strong forbidden lines of similar width ( @xmath203000 km s@xmath0 ) , different from traditional intermediate broad lines . thus , we consider it more likely that these emission lines come from the inner region of nlr , and that the large width is due to outflows ( e.g. , komossa et al . 2008 ; zhang et al . 2011 ) . in this picture the traditional blr is completely obscured . while the dusty torus provides the most naturally explanation for the obscuration , it is worth considering other alternatives . the obscurer can be a single dusty cloud in the clumpy torus model or the distant interstellar medium . the obscuration of the blr requires a minimum size cloud to be @xmath35 parsecs ( bentz et al . 2013 ) . using the infrared luminosity as an approximation for the bolometric luminosity and a bolometric correction to @xmath36 , we estimate blr sizes in the range of 0.1 - 0.3 parsecs . assuming a galactic dust to gas ratio ( @xmath37 @xmath38 , bohlin et al . 1978 ; cox et al . 2006 ) , the gas column density of the obscurer should be much greater than @xmath39 @xmath40 as discussed above . note that in the clumpy torus , the gas density should be of the order of a few @xmath41 @xmath16 ( krolik & begelman 1988 ) , so a blr - sized cloud would have a column density two orders of magnitude larger than the above minimum value , or @xmath42 mag for a galactic dust to gas ratio . on the other hand , assuming a gas density more typical of the interstellar medium @xmath43 @xmath16 , the minimum size of cloud would be several tens of parsecs to meet the minimum column density . in the case of the single - cloud obscurer , one would expect to see long - term very large amplitude variability as the cloud moves in and out of the line of sight . the typical time scale would be the size of the cloud divided by the transverse speed , which is of order the local keplerian velocity . in a typical galaxy , the orbit velocity is order of 500 - 1000 km s@xmath0 , depending on the distance to the center , and a minimum size of a cloud should be blr sized , of order of 0.1 to 1 parsecs . this would give a variability time scale @xmath44 so it is difficult to observe the cloud transit event fully . however , the continuum source is much smaller ( a hundred to thousand times smaller ) , so it is possible to observe transient obscuration of the continuum source with small changes in broad lines on time scales of years if the cloud has a sharp edge . there can also be cases where the cloud blocks the continuum source fully but only partially blocks the broad line region as the obscuring cloud moves in and out . these sources will show very large equivalent widths of broad lines . the probability of detecting such a source depends on the relative size of the obscuring cloud relative to the blr size . in the case of comparable sizes of the two , one would expect that in most sources only the continuum is fully covered , while if the cloud is much larger than the blr , most likely both continuum and blr are obscured . similarly , we will observe many more such partially obscured sources in the clumpy torus than in the smooth dust torus . in reality , the cloud may have a smooth edge so the contrast between emission line and continuum should be much moderate . we note that a single cloud occultation event is likely to give a smaller polarization degree than the dusty torus obscuration due to different symmetries ( marin et al . 2015 ) . the weakness of the optical to uv continuum and the strong mid - infrared emission may be due to a temporarily shut - off of the central engine . a comparison of infrared , optical , and uv fluxes requires a drop in the uv flux from the inner disk by a factor of 30 - 300 in these sources , compared to typical agn ( see figure 2 ) . the size of the mid - infrared emission region is of order of tens to hundreds of dust sublimation radii at the infrared luminosity according to our model of the sed ( hnig et al . 2010 ) . at the infrared luminosity of these sources , the light - crossing time of the mid - infrared emission region is of order ten to a hundred years , so the central engine would have to have shutdown in less than this time . a comparison of sdss photometric data with palomar observatory sky survey ( poss ) on a time scale of 50 years in the observer frame shows a small fraction of quasars with a variability amplitude of 3 mags in @xmath45-band ( de vries et al . although these amplitudes are still smaller than that is required , since we do not have light curves of agn on such long time scales , it is impossible to tell whether such a large amplitude variability happened on time scale of a hundred years and how frequent they are . thus this remains a possibility . there are several tests that could distinguish among these possibilities . first , scattered light should be highly polarized , while direct light from a much less active accretion disk is likely not . the observed scattered light in seyfert 2 galaxies is polarized by a few to tens percent , while the polarization of direct emission from quasars is usually less than 1% ( marin et al . 2015 ) . it should be noted that if a quasar resides in the gas - rich environment , there is inevitable scattered light from surrounding gas . as the central continuum source dims quickly , the fraction of scattered light increases , and so does the polarization degree . so a more critical test would be to observe the transmitted hard x - rays . second , direct emission from accretion disk is likely to be variable on relatively short time scales , while the scattered light is not . thus , the detection of variability on time scales of less than a year would support the intrinsic weak continuum emission model . we now turn to the properties of the emission region . previous studies suggested that quasar emission lines consist of distinct kinematic components : broad , intermediate and narrow components ( brotherton et al . 1994 ; marziani et al . ( 2008 ) found that optical fe ii lines are formed in the intermediate region , that is redshifted with respective to the system . the fe ii spectrum of our sources does not look like those in normal quasars or from most theoretical simulations aimed at reproducing observed fe ii in i zw 1 ( baldwin et al . 2004 ; sigut & pradhan 2003 ; ferland et al . 2009 ; bruhweiler & verner 2008 ) . we will mainly focus on the fe ii emission in sdss j1535 + 0903 first , because in other sources the spectra can be reproduced with a combination of sdss j1535 + 0903 and i zw 1-like components . the uv fe ii emission is dominated by a few bumps , namely multiplets uv 1 , uv 2,3 , uv 62,63 , with weak emission in between and very weak or no emission at wavelengths longer than 2900 . in sdss j1005 + 4934 , the sdss spectrum extends to 4400 in the quasar rest frame , and optical fe ii lines appear relatively weak . for thermal excitation , the extreme spike / gap ratio of 8.5 in sdss j1535 + 0903 may indicate very high densities ( @xmath46 @xmath16 ) and very low ionization parameters ( @xmath47 ) , as shown in figure 3 of baldwin et al . but their models are only for the broad line region , and the ionizing photon flux is too large for our case . however , in their figure , the curves of iso - spike / gap ratio run diagonally , so the dimensionless ionization parameter may be more crucial than the density . unfortunately their models do not extend to the low density and very low ionization regime , so it is not clear whether extreme spike / gap ratios can be reproduced there . but if this is indeed the case , the fe ii emission region must be far from the quasar so that it is not obscured by the dusty torus . at very low values of the ionization parameter , some other lines , such as c ii and ni ii , might become strong . detailed photo - ionization modeling would be required . also the presence of strong c iv and al iii , and the lack of c iii and si iii can not be explained by this model . alternatively , the extended emission may come from resonant scattering in low - ionization quasar outflows . previous observations suggest that felobals and hibals , at least in some quasars , are formed on the scale of 100 pc to a few kpc ( see references in 1 ) , so it is likely that outflows are not obscured by the putative dusty torus that blocks the line of sight to the central continuum in type 2 quasars . the outflows may produce fe ii via collisional excitation , as well as via continuum fluorescence . the contribution of fluorescence relative to collisional excitation depends on the optical depth of fe ii as well as the density and temperature of the gas . in the case of felobal outflows , the density is usually low ( order of @xmath48 @xmath16 , de kool et al . 2001 ; lucy et al . 2014 ; c.f . , zhang et al . 2015 ) , and the optical depth of most fe ii lines is usually not large ( order of a few to tens for strong lines ) . thus , continuum fluorescence may be the dominant process in the outflows for fe ii lines . some other emission lines may be also dominated by resonant scattering from such outflows . consequently , these lines should appear strong in comparison with non - resonant lines . the presence of strong al iii , but very weak or missing similar component of c iii ] and si iii ] , which are usually much stronger , in sdss j1535 + 0903 , indicates that resonant scattering is the dominant excitation mechanism in this object . al iii can be produced by either resonant scattering or collisional excitation , while c iii ] and si iii are only collisionally excited . although the relative strength of al iii increases with gas metallicity , this does not explain the fact that al iii is much stronger than c iii ] . in table 2 , the al iii equivalent width is a factor of 3.7 larger than the c iii ] equivalent width . as discussed in [ 3.1 ] , the observed uv continuum is likely scattered light , and , as seen in figure [ specfit ] , the equivalent widths of c iii ] and si iii is similar to that of i zw 1 . so c iii ] emission may be fully accounted for by scattered component and the contribution from the outflow component may be negligible . if fe ii is produced by continuum fluorescence in outflows , a simple idea is that its spectrum may resemble those of the fe ii absorption line spectra in felobal qso sdss j080957.39 + 181804.42 . we note that there are usually very weak fe ii absorption lines above 2800 , and no optical fe ii absorption lines , consistent with the observed fe ii in these spectra . to test this idea , we extract the absorption line spectrum of a felobal qso . this spectrum is shown in figure 1 , and is similar , to first order . it should be noted that the spectrum of fluorescent emission is not exactly the same as the fe ii absorption spectrum , because an fe ii ion in an excited level , after absorbing a photon , may decay to a different lower level , thus emitting a photon with an energy different from the absorbed one , especially when resonance lines become optically thick . detailed calculations are presented in [ model ] . in appendix a , we demonstrate that a pure fluorescent model can indeed reproduce the observed fe ii band ratios in sdss j1535 + 0903 very well . in order to make fe ii the dominant ion of fe , we let the incident ionizing continuum be filtered by an intervening absorbing gas of column density 10@xmath49 @xmath40 . we adopt a gas density @xmath50 @xmath16 , consistent with those in the felobal qsos , and a turbulent velocity of 1000 km s@xmath0 to mimic the outflow velocity . we find that at the optical depth of the resonant fe ii line at @xmath51 is 900 , and that the model produces uv feii band ratios that are consistent with the observed ones . as far as we know , this is the only model that gives such an extreme fe ii ratio at a low ionizing continuum flux , which is consistent with its large line formation region . furthermore , the model gives weak fe ii lines at wavelength longer than 2800 , which is also consistent with the observation . the best matched model predicts an equivalent width of fe ii at 2355 band of 99 , which is about one 6th of the observed value for sdss j1535 + 0903 , for a covering factor of one . however , as we discussed in [ 3.1 ] , the continuum incident to the gas is obscured and the observed continuum is attenuated by two orders of magnitude , so only a few percentages covering factor would be sufficiently to explain the observed fe ii strength . this covering factor is consistent with the estimate in next subsection . the gas producing the fe ii emission is cool , as shown in the appendix , so collisional excitation is unimportant . this accounts for the weak optical fe ii emission . note that this type of model will also predict a number of other low ionization lines , such as ti ii , ni ii and si ii , but their atomic data are very sparse . so a full treatment of these lines has to wait for a major update in the atomic data . our models in appendix a do not produce al iii and c iv , because ionizing photons above 13.6 ev are required to produce these ions while in our scenario these photons are intentionally filtered out to prevent destruction of fe@xmath52 ions . both al iii and c iv are observed in absorption in a felobal qso , and their average emission spectrum should not be different from the absorption line spectrum in a felobal qso because they are scattered resonantly . so our purpose is only to demonstrate the effect of fe ii radiation transfer after fe ii resonant absorption . note also that the observed fe ii ratios from different fe ii bands depends on the reddening , because felobal qsos often show a large internal dust extinction . we do not apply such a reddening correction because we do not know the location of the dust . if dust lies outside of the balr , then an extinction correction should be applied . on the other hand , if dust is within the balr , no correction is needed . anyway , to first order the absorbed line spectrum of felobal quasars are similar to the quasars observed in the sample . in sdss j1005 + 4934 , we also detect strong high ionization forbidden or semi - forbidden lines that are excited by collisional process , while the fe ii spectrum as well as fe ii /al iii ratio , are nearly the same as sdss j1535 + 0903 . we also obtained an electron density of ( 3 - 4)@xmath53 @xmath16 from c iii and si iii . these lines are rather broad with widths of 3000 km s@xmath0 . the presence of such lines does not necessarily contradict our interpretation of the strong fe ii , al iii and mg ii lines as resonantly scattering from an outflow . according the the recent model of faucher - gigure et al . ( 2012 ) , the felobal may be produced in the shocked cooling gas during interaction of outflows with the cold gas . the gas producing strong high ionization forbidden lines may be the warm phase of such shocked gas , and the large width of the line may be explain in this way as well . blue shifted fe ii and mg ii lines are seen in the sdss j1632 + 3405 , while emission lines in other quasars do not display systematic blueshift . the blue shifted lines can be easily explained as from an outflow since one expects that the far side of the outflow is subject to large dust extinction and would not be seen . however , emission lines from outflows are not necessarily systematically blueshifted if the farside of the outflows is not particularly obscured or the system is seen nearly edge - on . in the continuum fit , we found that sdss j1632 + 3405 has the smallest inclination angle among these objects . if the fe ii in these objects is due to fluorescent excitation , we can constrain the global covering factor of the outflows in these objects . to first order , we can assume an isotropic angular redistribution function , so the scattered fe ii flux will be @xmath54f_c(\lambda)*c\ ] ] where @xmath55 is the un - absorbed continuum flux . assuming a maximum @xmath56 at certain @xmath7 , one can then estimate a minimum covering factor @xmath57_{max}$ ] . to estimate @xmath55 , we assume that the intrinsic sed is similar to the mean quasar sed , and scale the latter to match the wise luminosity at 22@xmath24 m , the peak wavelength of sed . by doing so , we obtain covering factors of 0.05 , 0.20 , 0.01 , and 0.06 for sdss j2125 + 0044 , sdss j1632 + 3405 , sdss j1005 + 4934 and j1535 + 0903 . note that if mg ii is dominated by resonant scattering in the balr , then the covering factor of mg ii balr would be a factor of 2 - 3 times larger than the above value . there are several uncertainties in this estimate . first , due to large number of energy levels , fe ii can not be rigidly considered as resonant scattering , but rather as the fluorescent process . the absorbed photons may emerge at other wavelength , so the strong resonant lines are usually weaker than in the case of resonant scattering . second , it is possible that infrared emission of these quasars is enhanced because we selected objects with very red infrared colors . in this case , the unabsorbed continuum is over - estimated . since sed modeling suggested that infrared emission is dominated by the agn component , while it was shown that , on average , thermal dust emission accounts for 35 percent of the total power of quasars ( e.g. , ma & wang 2013 ) , thus at most , the intrinsic continuum is over - estimated by a factor of 2 - 3 . third , the scattered photons may be destroyed by dust within or outside the balr , so we may underestimate the covering factor . it has been known for quite a long time that felobal qsos are reddened . but the location of the dust is still unknown . recently , there are arguments that dust may be mixed within the balr because near - infrared emission is correlated with the outflow velocity ( zhang et al . 2014 ) , and the variability of bal is correlated with the uv spectral slope ( he et al . further , dunn et al . ( 2015 ) showed that reddening must be caused by large scale dust because both the continuum and narrow lines are reddened . if dust is outside of the balr and the scattered light suffers the same extinction as the continuum source , and adopting a median value for the continuum extinction for lobal @xmath58 ( dunn et al . 2015 ) , the correction will raise the covering factor by about 40% . despite these uncertainties , we note that in 3 of 4 objects , the uncorrected covering factors are already larger than 0.05 , while according to statistics , felobal qsos account for no more than 2% of total quasars ( trump et al . 2006 ; dai et al . this suggests strongly that felobal qsos must be in a special phase of quasar evolution , rather than representing a very small covering factor . we presented an analysis of four quasars with peculiar uv fe ii spectra . they had extremely strong uv1,3,4,62,63 and very weak fe ii at wavelengths longer than 2800 . their seds rise very steeply from the optical to the mid - infrared , suggesting they are obscured . the obscurer is most likely the dusty torus , since that also provides a self - consistent explanation for the infrared emission although other alternatives can not be ruled out . the characteristics of the fe ii spectra can be understood as due to resonant scattering in an extended outflow since the size of the scattering region must be outside the obscuring torus and its fe ii spectrum similar to absorption spectrum of felobals . our fluorescent models can successfully produce the observed uv fe ii band ratios and the weakness of fe ii lines at wavelengths longer than 2800with a low ionizing photon fluxes . this is because the strong uv lines are excited by continuum pumping within gas that is too cool for collisional excitation to be significant . strong al iii and weak c iii ] and si iii emission may be explained as resonant scattering , as is observed in felobal qsos . the obscuration of the direct continuum also explains the very large equivalent widths of fe ii , mg ii and al iii lines . these objects are misaligned felobals . by considering the amount of scattered light we set a lower limit on the covering factor of felobals to be 5% to 20% in three objects , supporting the scenarios that felobal qsos are a special stage of quasar evolution . we thank the referee for constructive comments that help to improve the presentation of the paper . we acknowledge the financial support by the strategic priority research program `` the emergence of cosmological structures '' of the chinese academy of sciences ( xdb09000000 ) , nsfc ( nsfc-11233002 , nsfc-11421303 , u1431229 ) and national basic research program of china ( grant no . 2015cb857005 ) . funding for sdss - iii has been provided by the alfred p. sloan foundation , the participating institutions , the national science foundation , and the u.s . department of energy office of science . the sdss - iii web site is http://www.sdss3.org/. sdss - iii is managed by the astrophysical research consortium for the participating institutions of the sdss - iii collaboration . gjf acknowledges support by nsf ( 1108928 , 1109061 , and 1412155 ) , nasa ( 10-atp10 - 0053 , 10-adap10 - 0073 , nnx12ah73 g , and atp13 - 0153 ) , and stsci ( hst - ar- 13245 , go-12560 , hst - go-12309 , go-13310.002-a , hst - ar-13914 , and hst - ar-14286.001 ) . 40 antonucci , r. r. j. , & miller , j. s. 1985 , , 297 , 621 baldwin , j. a. , ferland , g. j. , korista , k. t. , hamann , f. , & lacluyz , a. 2004 , , 615 , 610 baskin , a. , laor , a. , & hamann , f. 2013 , , 432 , 1525 bentz , m. c. , denney , k. d. , grier , c. j. , et al . 2013 , , 767 , 149 bohlin , r. c. , savage , b. d. , & drake , j. f. 1978 , , 224 , 132 boksenberg , a. , carswell , r. f. , smith , m. g. , & whelan , j. a. j. 1978 , , 184 , 773 borguet , b. c. j. , arav , n. , edmonds , d. , chamberlain , c. , & benn , c. 2013 , , 762 , 49 boroson , t. a. , & meyers , k. a. 1992 , , 397 , 442 brotherton , m. s. , wills , b. j. , francis , p. j. , & steidel , c. c. 1994 , , 430 , 495 bruhweiler , f. , & verner , e. 2008 , , 675 , 83 cicone , c. , maiolino , r. , sturm , e. , et al . 2014 , , 562 , a21 chamberlain , c. , arav , n. , & benn , c. 2015 , , 450 , 1085 cox , n. l. j. , cordiner , m. a. , cami , j. , et al . 2006 , , 447 , 991 dai , x. , shankar , f. , & sivakoff , g. r. 2012 , , 757 , 180 de kool , m. , arav , n. , becker , r. h. , et al . 2001 , , 548 , 609 dunn , j. p. , arav , n. , aoki , k. , et al . 2012 , , 750 , 143 dunn , j. p. , wasik , b. , holtzclaw , c. l. , et al . 2015 , , 808 , 94 farrah , d. , urrutia , t. , lacy , m. , et al . 2010 , , 717 , 868 fabian , a. c. 2012 , , 50 , 455 faucher - gigure , c .- a . , quataert , e. , & murray , n. 2012 , , 420 , 1347 feruglio , c. , maiolino , r. , piconcelli , e. , et al . 2010 , , 518 , l155 ferland , g. j. , peterson , b. m. , horne , k. , welsh , w. f. , & nahar , s. n. 1992 , , 387 , 95 ferland , g. j. , hu , c. , wang , j .- , et al . 2009 , , 707 , l82 ferland , g. j. , porter , r. l. , van hoof , p. a. m. , et al . 2013 , , 49 , 137 hall , p. b. , anderson , s. f. , strauss , m. a. , et al . 2002 , , 141 , 267 hamann , f. , korista , k. t. , & morris , s. l. 1993 , , 415 , 541 he , z .- c . , bian , w .- h . , jiang , x .- , & wang , y .- f . 2015 , , 454,3962 hnig , s. f. , & kishimoto , m. 2010 , , 523 , a27 hu , c. , wang , j .- m . , ho , l. c. , et al . 2008 , , 683 , l115 jenkins , e. b. 2009 , , 700 , 1299 knigge , c. , scaringi , s. , goad , m. r. , & cottis , c. e. 2008 , , 386 , 1426 krolik , j. h. , & begelman , m. c. 1988 , , 329 , 702 komossa , s. , xu , d. , zhou , h. , storchi - bergmann , t. , & binette , l. 2008 , , 680 , 926 korista , k. , bautista , m.a.,arav , n. et al . 2008 , , 688 , 108 laor , a. , jannuzi , b. t. , green , r. f. , & boroson , t. a. 1997 , , 489 , 656 lazarova , m. s. , canalizo , g. , lacy , m. , & sajina , a. 2012 , , 755 , 29 li , z. , zhou , h. , hao , l. , et al . 2015 , , 812 , 99 liu , g. , arav , n. , & rupke , d. s. n. 2015 , , 221 , 9 lpez - gonzaga , n. , jaffe , w. , burtscher , l. , tristram , k. r. w. , & meisenheimer , k. 2014 , , 565 , a71 lucy , a. b. , leighly , k. m. , terndrup , d. m. , dietrich , m. , & gallagher , s. c. 2014 , , 783 , 58 ma , x. , & wang , t.g . 2013 , , 430 , 3445 marin , f. , goosmann , r. w. , & gaskell , c. m. 2015 , , 577 , a66 marziani , p. , sulentic , j. w. , dultzin - hacyan , d. , calvani , m. , & moles , m. 1996 , , 104 , 37 meusinger , h. , schalldach , p. , scholz , r .- d . , et al . 2012 , , 541 , a77 nestor , d. , hamann , f. , & rodriguez hidalgo , p. 2008 , , 386 , 2055 osterbrock , d. e. , & ferland , g. j. 2006 , astrophysics of gaseous nebulae and active galactic nuclei , 2nd . ed . by d.e . osterbrock and g.j . ferland . sausalito , ca : university science books , 2006 , 126 polletta , m. , tajer , m. , maraschi , l. , et al . 2007 , , 663 , 81 richards , g. t. , lacy , m. , storrie - lombardi , l. j. , et al . 2006 , , 166 , 470 rllig , m. , abel , n. p. , bell , t. , et al . 2007 , , 467 , 187 ross , n. p. , hamann , f. , zakamska , n. l. , et al . 2015 , , 453 , 3932 sigut , t. a. a. , & pradhan , a. k. 2003 , , 145 , 15 tristram , k. r. w. , burtscher , l. , jaffe , w. , et al . 2014 , , 563 , a82 tristram , k. r. w. , meisenheimer , k. , jaffe , w. , et al . 2007 , , 474 , 837 o trump , j. r. , hall , p. b. , reichard , t. a. , et al . 2006 , , 165 , 1 turnshek , d. a. , weymann , r. j. , liebert , j. w. , williams , r. e. , & strittmatter , p. a. 1980 , , 238 , 488 urrutia , t. , becker , r. h. , white , r. l. , et al . 2009 , , 698 , 1095 verner , e. m. , verner , d. a. , korista , k. t. , et al . 1999 , , 120 , 101 voit , g. m. , weymann , r. j. , & korista , k. t. 1993 , , 413 , 95 veilleux , s. , trippe , m. , hamann , f. , et al . 2013 , , 764 , 15 de vries , w. h. , becker , r. h. , white , r. l. , & loomis , c. 2005 , , 129 , 615 wang , t. , zhou , h. , yuan , w. , et al . 2009 , , 702 , 851 wang , h. , xing , f. , zhang , k. , et al . 2013 , , 776 , l15 wang , t.g . , yang , c.w . , wang , h.y . , & ferland , g. j. 2015 , apj , in press ( arxiv:1510.04945 ) weymann , r. j. , morris , s. l. , foltz , c. b. , & hewett , p. c. 1991 , , 373 , 23 zhang , s. , wang , t .- g . , wang , h. , et al . 2010 , , 714 , 367 zhang , k. , dong , x .- b . , wang , t .- , & gaskell , c. m. 2011 , , 737 , 71 zhang , s. , wang , h. , wang , t. , et al . 2014 , , 786 , 42 zhang , s. , zhou , h. , wang , t. , et al . 2015 , , 803 , 58 ccccccccccccccc 100552.6 + 493448 & 1.1217 & 3150 & 173 & 6.4 & 310 & 35 & 103 & + 153542.4 + 090341 & 1.5330 & 1230 & 545 & 8.5 & 782 & 186 & 49 & 330 + 163246.6 + 340526 & 0.8306 & 880 & 280 & 3.2 & 140 & & & + 212546.9 + 004455 & 1.4240 & 8300 & 82 & 2.9 & 103 & 3 & 6 & 2 + ccccccc 100552.6 + 493448&0.048 & 15 & 60 & -1.5 & 30 + 153542.4 + 090341 & 0.003 & 15 & 45 & -0.5 & 80 + 163246.6 + 340526 & 0.086 & 35 & 60 & -1.0 & 80 + 212546.9 + 004455 & 0.005 & 15 & 60 & -2 & 50 + here we examine the properties of a cold cloud where iron is mainly singly ionized . fe ii emission will be produced by continuum fluorescence under these circumstances . our scenario is that an intervening cloud has extinguished the agn continuum and that this attenuated sed strikes an outer dust - free cloud . the intervening cloud is assumed to have a column density of n(h ) = 10@xmath49 @xmath40 and an unattenuated ionization parameter of log @xmath59 ( although this has little affect on our predictions ) . the sed transmitted through this cloud , which then strikes the outer cloud , is shown in fig a1 . we assume that the outer cloud is dust free . this is to allow fe to exist in the gas phase . fe is strongly depleted in dusty gas ( jenkins 2009 ) so little fe ii emission occurs when grains are present . it is possible that the cloud was at one time too close to the agn for dust to exist . the cloud would be similar to a classical pdr ( rollig et al . 2007 ) were grains present . in a pdr grains heat the gas while absorbing the balmer continuum of the incident sed . since grains are , by hypothesis , absent in our cloud , the gas will be heated mainly by photoelectric absorption by the gas , mainly of second and third row elements with ionization potentials smaller than hydrogen . a low gas kinetic temperature results . the radiation striking the gas is bright at wavelengths longward of the lyman limit , so the main species that will be present in the cloud will be h i , he i , c ii , n i , o i , ne i , na ii , mg ii al ii , si ii , s ii , and fe ii . typical gas temperatures in our resulting models are around 300 k ( figure [ figa3 ] ) , too cool to collisionally excite optical or uv lines . uv line emission is produced by continuum fluorescent excitation in this case . note that it is quite likely that the sed striking the cloud is not visible from our line of sight . this would occur if parts of the torus lie between us and the central object . we use version c13.03 of cloudy , a spectral simulation code last described by ferland et al . verner et al . ( 1999 ) describe the fe ii model in this version of cloudy . fe ii emission is complex ( baldwin et al . 2004 ) and we report it using the sums over wavelength bands described in that paper . results are mainly sensitive to the column density of the outer cloud and we will show predictions resulting from varying this parameter below . in this cool environment , fe ii is excited mainly by absorption of the incident sed by uv lines whose lower levels are close to , or within , the ground term . the level populations within the ground term are affected by the density which then affects which uv lines are efficiently pumped . the density does not affect the emission from high levels so long as it is below the critical density of the excited levels , @xmath60 @xmath16 . at lower densities the emission spectrum is mainly determined by the branching ratios as electrons decay to lower levels . we experimented with a range of densities but settled on @xmath61 @xmath16 . lines absorb mainly over their doppler core so the amount of pumping will depend on the line width ( ferland 1992 ) . we assume 1000 km s@xmath0 in the examples we show although this does not greatly affect the details . the fe ii grows stronger as the column density increases and fe ii lines absorb more of the incident continuum . optical depths in the strong uv lines do affect the branching ratios . as they grow more opaque decays from their upper levels will favor subordinate lines . this is quite analogous to the transition from case a to case b in h i emission ( osterbrock & ferland 2006 ) . finally the intensity of the radiation field give in figure [ figa1 ] must be set . we assume an unabsorbed ionization parameter of @xmath62 . results are shown in figure [ figa2 ] . the lower axis gives the total hydrogen column density . an iron abundances of log fe / h = -4.55 was assumed . a good fit is obtained at @xmath63 @xmath40 . the upper axis shows the optical depth in one of the strong uv lines that is responsible for pumping the atom . the net fe ii emission is shown in [ figa4 ] . fe ii uv 1 , 3 , 4 , 63 , 64 are the strongest lines , while those at wavelengths longer than 2800are much weaker . the spectrum looks similar to the observed fe ii . with these parameters , the thickness of the cloud is one parsec , and the distance to the continuum source is estimated about 1 kpc , following the definition of the ionization parameter , and the inferred intrinsic continuum luminosity of j1535 + 0903 . the total volume of fe ii region is about @xmath41 parsec@xmath64 , and a total gas mass of about @xmath65 m@xmath3 , which is comparable to the reported @xmath66 m@xmath3 molecular gas in outflows of infrared luminous quasars , such as mrk 231 ( feruglio et al . 2010 ; cicone et al . however , the gas density is poorly constrained so these numbers are only very rough estimate .

it has become a widespread practice to study the physics of systems in various numbers of spatial dimensions , not necessarily @xmath0 . for instance , graphene with @xmath1 , and string or membrane theory with @xmath2 as high as @xmath3 , are two examples that immediately come to mind for both their experimental and theoretical interest . moreover , geometric ideas provide a common framework used to pursue such studies . a pedagogical goal of this paper is to encourage students to think along these lines in the context of a familiar subject electrostatics . for example , in three spatial dimensions a uniformly charged straight - line segment gives rise to an electric potential @xmath4 whose equipotential surfaces are prolate ellipsoids of revolution about the segment , with the ends of the segment providing the foci of the ellipsoid . as an immediate consequence of the geometry for these prolate ellipsoidal equipotentials , the associated electric field @xmath5 always normal to surfaces of constant @xmath4 has at any observation point a direction that bisects the angle formed by the pair of lines from the observation point to each of the two foci of the ellipsoid . this beautiful electrostatic example was presented by george green in 1828 @xcite , and it has been discussed in many books since then @xcite-@xcite including at least two texts from this century @xcite . while the straight - line segment is an idealization , nevertheless it provides insight into the behavior of real thin - wire conductors , especially upon approximating those real wires as very narrow , needle - like ellipsoids . but as it turns out , the line segment problem in three dimensions is a very special case , in some sense the most ideal of all possible electrostatic worlds . in any _ other _ dimension of space , uniformly charged segments do * not * produce ellipsoidal equipotentials . conversely , in any other dimension , if the equipotentials are ellipsoidal about a linearly distributed straight - line segment of charge , then that charge distribution can * not * be uniform . one need look no farther than two dimensional systems to see clearly that there are differences between uniformly charged segments and those with charge distributed so as to produce ellipsoidal equipotentials . indeed , @xmath1 is the only ellipsoidal equipotential case which is intuitive in the sense that the associated linear charge distribution has maxima at the ends of the segment , as one might naively expect from the repulsive force between like charges placed on a segment of a real , thin conductor at a finite potential . in contrast , the distribution of charge needed to produce ellipsoidal equipotentials is counter - intuitive in higher dimensions . to produce such equipotentials for @xmath6 the charge distribution must have a maximum at the center and vanish at the ends of the segment . we begin our discussion in section 2 in two dimensions , where the two types of charged segments are readily analyzed . then we compare and contrast uniformly charged line segments with those admitting ellipsoidal equipotentials for any number of spatial dimensions . as a preliminary , we first discuss briefly in section 3 the potential of a point charge in @xmath2 dimensions . we then use this information in section 4 to compute the potentials and electric fields for uniformly charged line segments . we continue in section 5 by considering systems with ellipsoidal equipotentials in @xmath2 dimensions . we then determine the linear charge distributions that produce such potential configurations , and we find the remarkable result that the linear charge density giving rise to ellipsoidal equipotentials is peaked at the center of the segment , for @xmath6 . the discussion of general @xmath2 affords the opportunity to illustrate how continuous @xmath2 can be used as a mathematical device to regulate singular behavior . this too is a widespread practice in theoretical physics . we use @xmath2 in this way in section 5 of the paper to interpolate continuously between intuitive and counter - intuitive charge distributions for segments with ellipsoidal equipotentials . finally , in section 6 , we invoke well - known methods to indicate how the various line segment results also give solutions to a class of electrostatic boundary value problems where the charge is moved outward and distributed on one of the equipotential surfaces that surrounds the original segment . the point - particle electric potential in 2d is well - known to be logarithmic . for a point charge @xmath7 located at the origin , up to a constant @xmath8 that sets the distance scale , @xmath9 where @xmath10 is the two dimensional analogue of coulomb s constant . by linear superposition , a finite length , uniformly charged line segment , with constant charge / length @xmath11 distributed on the @xmath12 axis for @xmath13 , produces a potential @xmath14 this result may be established by integrating the contributions of infinitesimal point - like bits of charge that make up the segment , using ( [ 2dpointcharge ] ) and the indefinite integral @xmath15 written out in full,@xmath16 a plot of the potential surface shows the essential features . eurojphyslinesegmentsv2__1.pdf + figure 1 : potential surface for a uniformly charged line segment in 2d . the top of the potential surface is curved and not a straight line , indicating that the charged segment itself is _ not _ an equipotential . this follows analytically from ( [ uniformline2d ] ) . although points on the segment are not at the _ same _ potential , they are all at _ finite _ values of the potential , for @xmath1 . explicitly , for @xmath17 and all @xmath12,@xmath18 we plot @xmath19 versus @xmath12 to show the shape of the potential along the @xmath12-axis , for @xmath20 and @xmath21 . eurojphyslinesegmentsv2__2.pdf + figure 2 : uniformly charged line segment potential for @xmath1 , along the @xmath12-axis . on the other hand , transverse to the @xmath12-axis the potential has a discontinuous slope across the line segment . that is to say , the electric field normal to the line segment , @xmath22 , is discontinuous due to the presence of the charge density on the segment . for example , for @xmath23 this transverse profile is given by@xmath24 \ .\ ] ] we also plot @xmath25 versus @xmath26 to show the shape of the potential along the @xmath26-axis , for @xmath20 and @xmath21 . eurojphyslinesegmentsv2__3.pdf + figure 3 : uniformly charged line segment potential for @xmath1 , along the @xmath26-axis . moreover , the equipotentials are _ not _ ellipses surrounding the segment on the @xmath27-plane , as is especially clear for points close to the segment . the potential contours actually _ intersect _ the segment . a view of the potential contours from _ below _ the potential surface shows these features graphically . ( see the plot to follow . but note the view in that plot is actually an orthogonal projection of the contours onto the @xmath27-plane , and not the true perspective of an observer on the potential axis a finite distance below that plane . ) eurojphyslinesegmentsv2__4.pdf + figure 4 : potential contours for a uniformly charged line segment in 2d . the components of the electric field produced by the segment are given by@xmath28 where we have used @xmath29 . note that @xmath30 is independent of the scale @xmath8 , since changing @xmath8 just amounts to adding a constant to the potential . but of course @xmath30 does depend on @xmath31 as this sets the physical length scale for the system . as a check , in the units we have chosen the charge density along the line follows from @xmath32 , integrated over a horizontal rectangle containing an infinitesimal portion of the @xmath12-axis , as the height of the rectangle is taken to zero . thus , using @xmath33,@xmath34{cl}0 & \text { \ \ if \ \ } x > l/2\\ \pi & \text { \ \ if \ \ } -l/2<x < l/2\\ 0 & \text { \ \ if \ \ } x<-l/2 \end{array } \right . \ . \label{uniformcharge2dderivation}\ ] ] that is to say , the charge / length between @xmath35 and @xmath36 is the constant @xmath11 , as expected . elsewhere , the charge density vanishes , as follows from @xmath37 for all points not coincident with the segment . it is instructive to make a vector field plot of @xmath38 , especially near the charged segment . again it is evident graphically that the segment itself is not an equipotential , since the electric field lines are not perpendicular to the segment as the @xmath12-axis is approached for @xmath39 , except at the single point @xmath23 . here are vector field plots for @xmath20 . eurojphyslinesegmentsv2__5.pdf eurojphyslinesegmentsv2__6.pdf + figures 5 & 6 : vector field plots of @xmath40 for a uniformly charged line segment situated between @xmath41 and @xmath42 . the field is evaluated at the center of each arrow . in contrast to the uniformly charged segment in 2d , consider a distribution of charge along the segment such that equipotentials _ are _ ellipsoidal . the relevant charge distribution turns out to be@xmath43{cl}0 & \text { \ \ if \ \ } x > l/2\\ 1 & \text { \ \ if \ \ } -l/2<x < l/2\\ 0 & \text { \ \ if \ \ } x<-l/2 \end{array } \right . \ , \label{lambda(x)2d}\ ] ] as we shall confirm in the following . since @xmath44 , the total charge on the segment is still @xmath45 , the same as for the uniformly charged case . the corresponding potential is now @xmath46 where @xmath47 is a sum of two distances , from the observation point @xmath48 to each of the two ends of the segment . that is,@xmath49 because the positional dependence of the potential is given entirely by @xmath47 , the equipotentials are ellipses on the @xmath27-plane , with the ends of the segment serving as the foci of each equipotential ellipse . again , we plot the potential surface . eurojphyslinesegmentsv2__7.pdf + figure 7 : potential surface for a non - uniformly charged line segment in 2d , with @xmath50 . the top of the potential surface is now a straight line , indicating that the charged segment itself _ is _ an equipotential , and the points on the segment are at a _ potential for @xmath1 , namely , @xmath51 . this follows analytically from ( [ ellipsoidalline2d ] ) . explicitly , for @xmath17 and all @xmath12,@xmath52 we plot @xmath53 versus @xmath12 to show the shape of the potential along the @xmath12-axis , for @xmath20 and @xmath21 . eurojphyslinesegmentsv2__8.pdf + figure 8 : non - uniformly charged line segment potential for @xmath1 , along the @xmath12-axis . transverse to the @xmath12-axis , the potential again has a discontinuous slope across the line segment due to the presence of the charge density on the segment . for example , for @xmath23 this transverse profile is given by@xmath54 we plot @xmath55 versus @xmath26 to show the shape of the potential along the @xmath26-axis , for @xmath20 and @xmath21 . eurojphyslinesegmentsv2__9.pdf + figure 9 : non - uniformly charged line segment potential for @xmath1 , along the @xmath26-axis . a view of the potential contours from below the potential surface shows the equipotential ellipses . ( but again note the view in that plot is actually an orthogonal projection of the contours onto the @xmath27-plane , and not the true perspective of an observer on the potential axis a finite distance below that plane . ) eurojphyslinesegmentsv2__10.pdf + figure 10 : ellipsoidal potential contours for a non - uniformly charged line segment in 2d . the components of the electric field are now@xmath56 whose final form in rectangular coordinates follows from @xmath57 in terms of @xmath58 and @xmath59,@xmath60 writing everything out in terms of @xmath12 and @xmath26 gives@xmath61 the components of @xmath62 are not elegant in these coordinates @xcite but their properties are fully encoded and amenable to machine computation.@xmath63 although the geometric features of the electric field may not be transparent from these expressions , nevertheless the fact that the equipotentials are ellipsoidal allows one to immediately visualize the direction of @xmath62 as normal to those surfaces of constant @xmath4 . to confirm the charge density along the line , we again use @xmath64 , and integrate over a horizontal rectangle containing an infinitesimal portion of the @xmath12-axis , as the height of the rectangle is taken to zero . the crucial features here are @xmath65 as @xmath66 for @xmath67 ( away from the segment ) , but @xmath68 as @xmath66 for @xmath39 ( on the segment ) . more precisely , as points on the segment are approached transversely , @xmath69 thus for @xmath39 , @xmath70 combining this with the more obvious @xmath71 for @xmath39 gives@xmath72{cl}0 & \text { \ \ if \ \ } x > l/2\\ 1 & \text { \ \ if \ \ } -l/2<x < l/2\\ 0 & \text { \ \ if \ \ } x<-l/2 \end{array } \right . \ . \label{nonuniformcharge2dderivation}\ ] ] that is to say , the charge / length between @xmath35 and @xmath36 is@xmath73 as anticipated above in ( [ lambda(x)2d ] ) . elsewhere , the charge density vanishes , again as follows from @xmath74 for all points not coincident with the segment . once more it is instructive to make a vector field plot of @xmath75 , especially near the charged segment . the fact that the segment itself is an equipotential is evident graphically since the electric field lines are perpendicular to the segment as the @xmath12-axis is approached for @xmath39 . here are vector field plots for @xmath20 . eurojphyslinesegmentsv2__11.pdf eurojphyslinesegmentsv2__12.pdf + figures 11 & 12 : vector field plots of @xmath40 for a non - uniformly charged line segment situated between @xmath41 and @xmath42 . the field is evaluated at the center of each arrow . as previously stressed , the geometry of the ellipsoidal equipotentials ensures that the electric field at any observation point always has a direction that bisects the angle formed by the two lines from the end points of the segment to the observation point @xcite . this result is also manifest in an integral expression for @xmath62 that follows from linear superposition of the field contributions from infinitesimal @xmath76 charges along the segment , upon choosing an appropriate integration variable . that is,@xmath77 the unit vector @xmath78 points from the infinitesimal charge on the segment to the observation point , with angle @xmath79 measured from the @xmath12-axis in the usual counterclockwise sense on the @xmath27-plane . thus @xmath80 . correspondingly , @xmath81 are the angles from the ends of the line segment to the observation point , as given by @xmath82 . since the integration over @xmath83 weights @xmath78 by an _ even _ function of @xmath83 , and the range of integration is _ symmetric _ about @xmath84 , it follows that the resulting direction of @xmath30 will be proportional to @xmath85 which points in direction @xmath86 . that is to say , @xmath87 the change of variables needed to obtain ( [ efieldpunchline ] ) will be discussed more fully below , for non - uniformly charged segments giving rise to ellipsoidal equipotentials in any number of dimensions . a direct graphical comparison of the uniformly charged segment and the non - uniformly charged segment in 2d is obtained by superimposing their equipotentials in a true orthogonal projection of the @xmath4 surface contours onto the @xmath27-plane . eurojphyslinesegmentsv2__13.pdf + figure 13 : 2d equipotential contours for a uniformly charged segment , with constant @xmath11 for @xmath88 , for @xmath89 , @xmath90 , @xmath91 , @xmath92 , @xmath93 , and @xmath94 , as inner to outer black curves , and ellipsoidal contours for a coincident non - uniformly charged segment , for @xmath95 , @xmath92 , and @xmath94 , as inner to outer red curves . for large distances from the line segments , whether uniformly charged or otherwise , the potential approaches that of a point charge as given in ( [ 2dpointcharge ] ) . so asymptotically both sets of contours become circles . regarding this , recall the total charge on either segment under consideration is the same , namely , @xmath96 . thus for large distances from the segment the equipotentials in figure 13 will coalesce , although it is perhaps surprising how rapidly this occurs . in figure 13 the equipotential contours for both the uniform and non - uniform charge distributions , for the same value of @xmath4 , are very nearly coincident if @xmath97 the exact locations of the two sets of contours never differ by more than a few percent if @xmath97 well before the contours reach their asymptotic circular form . for a point charge @xmath7 located at the origin of coordinates , in @xmath98 dimensions , the scalar potential at an observation point @xmath99 is hypothesized to be @xcite @xmath100 where @xmath10 is the @xmath2-dimensional analogue of coulomb s constant . integrating the radial gradient of @xmath101 over the surface of a sphere fixes the normalization , by gauss law.@xmath102 where the total solid angle ( i.e. the area of the _ unit radius _ sphere , @xmath103 , embedded in @xmath2 dimensions ) is given by @xmath104 for example , @xmath105 , @xmath106 , @xmath107 , @xmath108 , etc . to put it differently , in terms of a dirac delta in @xmath2-dimensions,@xmath109 and in fact , this gives the correct result even for @xmath1 , by taking a limit : @xmath110 , hence @xmath111 . so for @xmath1 the point particle potential is logarithmic , as previously noted in section 2 . perhaps the simplest convention would be to set @xmath112 which would require , for @xmath98,@xmath113 this is singular at @xmath1 where the potential is a logarithm , not a power . in that case the corresponding choice would be @xmath114 , so @xmath115 and again @xmath116 . the line of charge and observation point are shown in red in the following figure . eurojphyslinesegmentsv2__14.pdf + figure 14 : segment coordinates . from the figure ( @xmath117 and @xmath118 in the formulas correspond , respectively , to the bottom point b and the top point t in the figure ) @xmath119 the relations in the third line above follow from the law of sines.@xmath120note that the last expression must be used with care since @xmath121 is multi - valued . in particular , if @xmath122 , then as the segment is approached it is always true that @xmath123 , no matter how the segment is approached . more generally , when @xmath124 decreases to cross the surface of the sphere for which the segment is a diameter , the value of @xmath121 increases through @xmath125 . this follows from @xmath126 so that the argument of the @xmath121 in ( [ arcsin ] ) is just @xmath127 for @xmath128 , and @xmath129 at this radius . reducing @xmath124 below @xmath36 increases @xmath130 above @xmath125 , i.e. @xmath121 has moved onto another branch of the function . this change of branch can be explicitly taken into account through the use of heaviside step functions , @xmath131 , to write @xmath132 where the @xmath121 in this expression is the principal branch of the function . assuming linear superposition for the potential , a uniformly charged line segment as shown above in figure 14 , of length @xmath31 , centered on the origin , with top ( t ) and bottom ( b ) ends at @xmath133 ( i.e. @xmath134 ) and @xmath135 ( i.e. @xmath136 ) , will produce in @xmath98 dimensions a potential @xmath137 given by @xcite @xmath138 here @xmath11 is the constant charge / length on the line segment , and @xmath139 are the polar angles for vectors from the top and bottom endpoints of the line segment to the observation point @xmath140 , as in figure 14 . the potential has no dependence on the additional @xmath141 angles needed to specify the location of a point in @xmath2 dimensions using spherical polar coordinates . that is to say , in @xmath2 dimensions the equipotentials are always higher dimensional surfaces of revolution about the line segment . the corresponding electric field is given by@xmath142 and it depends manifestly on @xmath124 and @xmath143 . the only dependence of the electric field on the additional @xmath141 angles in @xmath2 dimensional spherical polar coordinates is carried by the unit vectors @xmath144 and @xmath145 . to see that ( [ phid ] ) is correct , we need only sum the potential contributions for infinitesimal pieces of the line segment with charge @xmath146 located on the vertical axis of figure 14 at position @xmath147 , and at a distance @xmath148 from the observation point @xmath149 as given by @xmath150 but writing @xmath151 and @xmath152 , for fixed @xmath153 and @xmath154 , we have @xmath155 , and @xmath156 on the other hand , @xmath157 . hence the result ( [ phid ] ) . for instance , in the special case @xmath0 the potential of the uniformly charged segment is @xmath158 note that the potential is infinite for all points on the charged segment itself , for which points @xmath159 and @xmath160 . be that as it may , after a bit of algebra @xmath161 can be reduced to @xmath162 where @xmath163 is the sum of the distances from the end points of the segment to the observation point . thus equipotentials in this case are given by constant @xmath47 , and as is common knowledge , this defines an ellipsoid of revolution about the line segment with focal points @xmath164 and @xmath165 . a view from below the potential surface again clearly shows the equipotentials are ellipses . eurojphyslinesegmentsv2__15.pdf + figure 15 : @xmath166 for @xmath20 , where @xmath167 and @xmath168 . eurojphyslinesegmentsv2__16.pdf + figure 16 : contours of constant @xmath169 for @xmath20 , plotted versus @xmath167 ( vertical axis ) and @xmath168 ( horizontal axis ) . for other dimensions , however , the geometrical shapes of the equipotentials are not so easily discerned . for @xmath6 , it remains to evaluate the angular integral in ( [ phid ] ) . define the indefinite integral @xmath170 , and compute@xmath171{ll}i\left ( 4\right ) = \vartheta\ , & i\left ( 5\right ) = -\cos\vartheta\ , \\ & \\ i\left ( 6\right ) = \frac{1}{2}\vartheta-\frac{1}{2}\cos\vartheta\sin \vartheta\ , & i\left ( 7\right ) = \frac{1}{3}\cos^{3}\vartheta-\cos \vartheta\ , \\ & \\ i\left ( 8\right ) = \frac{3}{8}\vartheta+\left ( \frac{1}{4}\cos^{3}\vartheta-\frac{5}{8}\cos\vartheta\right ) \sin\vartheta\ , \ \ \ \ \ & i\left ( 9\right ) = -\frac{1}{5}\cos^{5}\vartheta+\frac{2}{3}\cos^{3}\vartheta-\cos\vartheta\ , \end{array}\ ] ] etc . for odd @xmath172 the integral is always a polynomial of order @xmath173 in @xmath174 , while for even @xmath175 the integral always has a term linear in @xmath176 plus , for @xmath177 , a term with @xmath178 multiplying a polynomial of order @xmath179 in @xmath174 . for any @xmath2 we therefore obtain the angular integral in ( [ phid ] ) in terms of @xmath180 , @xmath181 , @xmath182 , @xmath183 , and @xmath184 . these quantities may then be expressed in terms of @xmath143 and @xmath124 upon using the relations in ( [ trig]-[arcsin ] ) . for example , in @xmath185 the potential of the uniformly charged segment is@xmath186 unlike @xmath187 , the potential @xmath188 is _ not _ a function solely of the variable @xmath47 as defined in ( [ s ] ) . consequently , the equipotentials are _ not _ ellipsoidal for this four dimensional example . indeed , a plot now shows that the equipotentials are not ellipsoidal , especially for points close to the uniform line of charge , although the shape of the potential surface is similar to that for @xmath0 . eurojphyslinesegmentsv2__17.pdf + figure 17 : @xmath189 for @xmath20 , where @xmath167 and @xmath168 . note that @xmath190 is infinite for all points on the segment . also note , to obtain the correct shape of the potential surface it is important to interpret @xmath121 in the formula for @xmath190 as a multi - valued function , as given explicitly by ( [ arcsinmulti ] ) . a view from below the potential surface shows that the potential contours are _ not _ ellipses in this case , although the difference is somewhat subtle . this next graph should be compared closely to the ellipsoidal case for @xmath185 as presented below in section 5 . eurojphyslinesegmentsv2__18.pdf + figure 18 : contours of constant @xmath191 for @xmath20 . the electric field is given by ( [ linee ] ) and in this case has @xmath124 and @xmath143 components@xmath192 all the complication in these expressions lies in the derivatives @xmath193 and @xmath194 whose geometrical significance is not yet easy to visualize . on the other hand , the last terms involving @xmath195 in ( [ uniformer ] ) and ( [ uniformetheta ] ) have simple geometrical interpretations since they define a vector that is perpendicular to the axis of the segment and points away from that axis for positive @xmath11 ( i.e. in the direction of @xmath196 if we were to use cylindrical coordinates ) . in any case , it is worthwhile to compute @xmath197 and @xmath198 since these derivatives appear for all even @xmath175 and always have the same form . after some algebra we find @xmath199 note the difference in sign for @xmath200 . once again @xmath196 , so the bulk of the contributions from these derivatives again gives a vector that is perpendicular to the axis of the segment . but there remains a contribution that is entirely in the @xmath144 direction if @xmath201 is written as @xmath202 , or else entirely in the @xmath145 direction if @xmath203 is written as @xmath204 . choosing the first of these options gives@xmath205 to compare to the three dimensional ellipsoidal case , it suffices to express @xmath144 and @xmath145 in terms of @xmath206 and @xmath207 . although the latter two unit vectors are not orthogonal , they are independent except at @xmath133 and @xmath135 , and at those particular angles either one of @xmath206 and @xmath207 will suffice to give the direction of the electric field , since @xmath208 . we find@xmath209@xmath210 consider the same geometry as in figure 14 , but rather than assuming uniform charge density on the line segment , suppose the potential produced by the line segment depends positionally only on @xmath47 in any number of dimensions . including some convenient numerical and dimensionful factors , suppose@xmath211 in this case , equipotentials are ellipsoidal by assumption , since the positional dependence is only on @xmath47 . the only issue is to determine @xmath212 . if the only charge present is on the line segment , then for other points the potential function must be harmonic , @xmath213 . so , using @xmath214 and @xmath215 ( see eqns(5)-(7 ) in @xcite for more details ) , we compute@xmath216 the potential will then be harmonic for points not lying on the segment if and only if@xmath217 for @xmath218 . for general @xmath2 _ the relevant solution _ of this second order , ordinary differential equation , is given by a gauss hypergeometric function represented by the standard series around @xmath219 , namely,@xmath220 the various factors in the last line are useful to determine the exact expression for the charge distribution along the segment , as presented below ( see ( [ chargedistribution ] ) ) . note the large @xmath47 behavior of @xmath221 implies@xmath222 since @xmath223 . the normalization of @xmath221 and the @xmath224 factors in ( [ ellipsoidalanyd ] ) were chosen so that the total charge on the segment is the same as in the uniformly charged case , namely , @xmath96 . this total charge appears in the limiting form of the potential for @xmath225 , i.e. @xmath226 . alternatively , closed - form expressions for @xmath212 in terms of more elementary functions are sometimes more easily obtained by evaluating the corresponding hypergeometric series around @xmath227 , as given by @xcite@xmath228 and then analytically continuing to @xmath218 . here @xmath229 and @xmath230 are constants chosen so that the result ( [ alternate3pointpotential ] ) is real - valued for @xmath218 , and so that any constant term ( the trivial harmonic ) is eliminated from @xmath221 as @xmath231 . for integer @xmath2 the hypergeometric functions that appear in the solutions ( [ 3pointpotential ] ) and ( [ alternate3pointpotential ] ) always reduce to elementary functions . for example , in various dimensions the relevant harmonic functions are:@xmath232 by construction , the equipotentials are always ellipsoidal for any @xmath2 , except for the trivial one dimensional case where equipotentials consist of just pairs of points on the line . in contrast to the uniformly charged line segment for @xmath233 , if the potential depends only on @xmath47 then the charged line itself where @xmath234 for all points on the segment is _ always _ an equipotential , albeit with _ infinite _ @xmath4 when @xmath235 , while for @xmath236 and @xmath1 the potential along the charged line is _ finite_. for example , consider graphically the case for @xmath185 . qualitatively the potential surface is similar to that for the uniformly charged segment in three dimensions . potential contours are ellipsoids that surround @xmath237 and _ never _ intersect the charge segment . also , the top of the potential surface is just an exact copy of the segment itself , albeit at an infinite value of @xmath4 . that is to say , the charged segment is itself an equipotential , but in fact the potential is infinite for points on the segment . again , a view from below the potential surface shows more clearly the ellipsoidal potential contours . eurojphyslinesegmentsv2__19.pdf + figure 19 : @xmath238 for @xmath20 , versus @xmath167 and @xmath168 . eurojphyslinesegmentsv2__20.pdf + figure 20 : contours of constant @xmath239 for @xmath20 . admittedly , in 4d it takes a discerning eye to see differences in the shape of the ellipsoidal equipotential surface compared to that for the uniformly charged segment , if graphs of the two cases are viewed separately . but if viewed side - by - side the difference _ is _ evident in contour plots for the potentials . perhaps even more clearly , the difference is highlighted by plotting both cases in the same graph . eurojphyslinesegmentsv2__23.pdf + figure 21 : comparison of @xmath4 contours in 4d for a uniformly charged segment ( black ) and a non - uniformly charged segment with ellipsoidal equipotentials ( red ) for @xmath20 . upon doing so , it is also evident that the non - uniform charge distribution must be greater near the center of the segment rather than at the ends of the segment , a counter - intuitive feature . the ellipsoidal equipotentials are shorter in the direction of the line segment , and wider transverse to the segment , than those of the uniformly charged line for the same value of @xmath240 . this is exactly the opposite of what happens in 2d , where the ellipsoidal equipotentials were longer and narrower than those of the uniformly charged segment for the same value of @xmath4 . the non - uniform charge distribution on the segment that produces ellipsoidal equipotentials can now be determined using the integral form of gauss law in @xmath2 dimensions , ( [ integralgausslawind ] ) , in complete parallel to the calculation leading to ( [ nonuniformcharge2d ] ) in 2d . here we only give the results of that calculation . in @xmath2 dimensions the charge density for a line segment that produces ellipsoidal equipotentials is@xmath241 where we have used the ratio of total solid angles in @xmath2 and @xmath242 dimensions,@xmath243 for example , @xmath244 to give @xmath245 in agreement with ( [ lambda(x)2d ] ) , while @xmath246 to give the uniform distribution , @xmath247 . these charge densities are all normalized so that the total charge on the segment is the same for any @xmath2 , namely,@xmath248 this follows from@xmath249 a geometrical construction to obtain the distribution ( [ chargedistribution ] ) is to consider a uniformly charged @xmath103 , i.e. a hypersphere embedded in @xmath2 dimensions , with radius @xmath250 and with _ constant _ hypersurface charge density @xmath251 . an orthogonal projection of the charge , @xmath252 , from a hyper - cylindrical ribbon of revolution about a diameter of the hypersphere onto the underlying surrounded segment @xmath253 of the diameter , gives precisely @xmath254 for @xmath255 . this follows directly from @xmath256 , where the hypersurface area @xmath257 of the hyper - ribbon that surrounds @xmath253 is @xmath258 with @xmath259 , @xmath260 , and @xmath261 . this construction is easily visualized for @xmath1 and @xmath262 . we plot the densities for various integer @xmath2 , from @xmath1 up to @xmath263 , shown respectively as the lower to upper curves ( for @xmath264 ) in the following graph . eurojphyslinesegmentsv2__24.pdf + figure 22 : linear charge density @xmath265 leading to ellipsoidal equipotentials , in various dimensions . in this graph the 2d case is the only one which is intuitive in the sense that the infinitesimal pieces of charge repel each other so we would expect excess charge / length to be pushed towards the ends of a finite length , equipotential segment of real conductor with small but nonzero transverse size , all parts of which are at the same finite @xmath4 . for the idealized line segment in 3d this is not so the charge density is uniform but then the segment itself is at an infinite @xmath4 so physical intuition based on real conductors is perhaps difficult to apply in this situation @xcite . in higher dimensions , the result is even _ more _ counter - intuitive for idealized line segments . the charge distribution is peaked at the center of the segment . of course , in making these statements , we are assuming the physical properties of the idealized segment itself can be completely understood mathematically by taking the limit of the surrounding equipotential ellipsoids of revolution as their girth goes to zero . to gain more insight about the behavior of the charge distribution and the resulting potential , we may think of @xmath2 in ( [ chargedistribution ] ) as a continuous variable not just an integer to be used as a regulating parameter by which the @xmath0 case can be approached as a limit . this point of view shows that all @xmath266 behave intuitively , with @xmath267 peaked at the ends of the segment . the extreme case is @xmath236 where all charge is located _ only _ at the ends of the segment . moreover , for all @xmath266 the potential along the segment , as given by ( [ 3pointpotential ] ) , is a _ finite _ constant , so there is nothing pathological about the potential that might obviate physical intuition based on real conductors . on the other hand , all @xmath6 behave counter - intuitively , with @xmath268 peaked at the center of the segment and vanishing at the segment ends . and for all @xmath6 the potential along the segment , as given by ( [ 3pointpotential ] ) , is _ infinite _ , so real - world physical intuition is not guaranteed to be reliable for these idealized situations . the uniformly charged 3d case , also with infinite @xmath4 along the segment , acts as a separatrix between intuitive and counter - intuitive charge distributions . a plot of @xmath265 as a surface over the @xmath269 plane helps to visualize these features . eurojphyslinesegmentsv2__25.pdf + figure 23 : the @xmath265 surface as a continuous function of both @xmath147 and @xmath2 for a straight - line segment from @xmath270 to @xmath271 . constant density for a uniformly charged @xmath0 segment is represented by a thick red line . we close this section with a discussion of the direct calculation of the potential and the electric field as linear superpositions of the @xmath272 s and @xmath273 s due to infinitesimal bits of charge along the segment , @xmath274 , for a line segment in @xmath2 dimensions . in terms of the obvious rectilinear coordinates , as shown in figure 14 of section 4 , the results are@xmath275@xmath276 where @xmath154 is the transverse distance from the segment and @xmath153 is the @xmath147-coordinate of the observation point , and where @xmath277 is a unit vector pointing from the location @xmath278 of the bit of charge on the segment to the observation point @xmath99 . now , using the charge distribution ( [ chargedistribution ] ) that produces ellipsoidal equipotentials , we eventually obtain@xmath279@xmath280 where we have changed integration variables from @xmath147 to an angle @xmath176 according to@xmath281 note that @xmath176 here is the polar angle _ measured from the location of the bit of charge _ , and in general this is not the @xmath143 shown in figure 14 of section 4 . nonetheless , @xmath282 and @xmath283 _ are _ the angles shown in figure 14 . these last relations , in particular ( [ crucialid ] ) and ( [ electricfieldthetaintegral ] ) , are useful to establish _ the direction _ _ of the electric field _ _ without actually performing the integration_. to verify this statement , let @xmath284 in ( [ electricfieldthetaintegral ] ) to find@xmath285 the integration here weights @xmath78 by an _ even _ function of @xmath83 , with the range of integration symmetric about @xmath84 , so it follows that the resulting direction of @xmath286 will be proportional to @xmath287 . that is to say , at any observation point around the segment the direction of @xmath288 bisects the angle formed by the two lines from the end points of the segment to the observation point , thereby confirming that the equipotentials are ellipsoids of revolution about the segment . note that this direct calculation of @xmath288 using the integral expression ( [ elineind ] ) generalizes prior calculations for @xmath0 where the charge density is uniform , as given in @xcite , to cases where the charge distribution on the segment is _ not _ uniform for @xmath233 . we leave it as an exercise for the reader to evaluate the integral in ( [ elineind ] ) to obtain explicit expressions , and to show that these are in agreement with @xmath289 . even before attempting to evaluate the integral in ( [ elineind ] ) , however , it should be evident that a determination of @xmath288 by taking the gradient of the potential is easier to carry through . having found the equipotentials for various charged line segments , we have in hand a set of solutions for a variety of electrostatic boundary value problems . if all the charge from the line segment is moved outward along electric field lines and placed on a single surface selected from among the equipotentials , in such a way as to preserve @xmath4 on that surface , then the surrounding equipotentials are unchanged as well . this is widely known , and was in fact a feature emphasized by green @xcite . the charge distribution on the selected surface , as required to carry out this feat , is of course given by the electric field normal to that surface as provided by the original charged line segment solution . in particular , ellipsoidal equipotentials in any @xmath2 yield hypersurface charge densities that are expressible as elementary functions . the results are straightforward generalizations of the well - known 3d ellipsoidal case . in fact , it is not even necessary for the hyperellipsoids to be surfaces of revolution . for a general ellipsoid , as defined in @xmath2 dimensions by @xmath290 the charge distribution on that hyperellipsoid such that it is an equipotential is given by@xmath291 this can be established by a straightforward adaptation of the argument that applies to the 3d case , as given for example in smythe @xcite , chapter 5 , 5.00-5.02 . if this charged hyperellipsoidal surface is squashed to obtain an equipotential hyperdisk ( see smythe 5.03 for the 3d case ) , say by letting @xmath292 while maintaining the constraint ( [ ellipsoidd ] ) , then the hypersurface charge density on the resulting hyperdisk ( counting charge on both sides of the disk ) becomes@xmath293 this charge distribution may now be projected along any of the remaining unsquashed principal axes of the original ellipsoid by elementary integrations . the final result is exactly the linear charge distribution given by ( [ chargedistribution ] ) , namely , @xmath294 where the projection has been made onto the @xmath10th axis . alternatively , the surface charge distribution ( [ ellipsoidsurfacecharge ] ) of the unsquashed ellipsoid may be projected directly without being squashed , again by elementary integrations , to obtain the same result for @xmath295 . in the context of a familiar subject electrostatics we have carried out several elementary calculations in various numbers of spatial dimensions , @xmath2 , to encourage students to think more critically about the role played by @xmath2 . although for @xmath0 a uniformly charged straight - line segment gives rise to an electric potential @xmath4 whose equipotential surfaces are prolate ellipsoids of revolution about the segment , with the ends of the segment providing the foci of the ellipsoid , this is a very special case . if @xmath233 , uniformly charged segments do not produce ellipsoidal equipotentials . if @xmath233 , a non - uniform distribution of charge is required to produce equipotentials that are ellipsoidal about a straight - line segment of charge . we have illustrated these @xmath233 features in detail for @xmath1 and @xmath185 , and we have provided a framework as well as explicit formulas to carry through the same level of detail for any @xmath2 . we have shown how @xmath1 is the only ellipsoidal equipotential case which is intuitive in the sense that the associated linear charge distribution has maxima at the ends of the segment , as one might naively expect . in contrast , we have shown that the distribution of charge needed to produce ellipsoidal equipotentials is counter - intuitive for @xmath6 , becoming all the more so as @xmath2 is increased , in the sense that the requisite charge distribution has an absolute maximum at the center of the segment and vanishes at its ends . for @xmath0 and @xmath185 , the potential surface plots were found to have similar features that can be distinguished only by careful inspection . differences in @xmath4 between uniformly charged line segments and those giving rise to ellipsoidal equipotentials were shown to be somewhat subtle . distinctions between the two cases were more easily drawn by examining the corresponding electric fields close to the segment , or , relatedly , by directly comparing the charge densities . we also took the opportunity to illustrate , albeit briefly , how continuous @xmath296 can be viewed as a mathematical tool to regulate singular behavior and to interpolate between intuitive and counter - intuitive charge distributions . in the course of our discussion , we employed some basic geometrical ideas at an elementary level , such as in the determination of the direction of the electric field associated with ellipsoidal equipotentials . finally , we indicated how the line segment results may be generalized to solve a class of electrostatic boundary value problems by distributing the charge on hypersurfaces , rather than straight lines , while maintaining the line segment equipotentials outside the charged hypersurface . for more on the history of the 3d line segment problem , including citations of some recent pedagogical papers on the subject , see t s van kortryk , on the fields due to line segments arxiv:1410.6832 [ physics.hist-ph ] . throughout our discussion we use rectangular cartesian , spherical polar , and cylindrical coordinates , in view of student familiarity with these coordinate systems , even though prolate ellipsoids of revolution are more naturally described in terms of prolate spheroidal coordinates . it is tempting to introduce the variable @xmath297 , the coordinate locally orthogonal to @xmath47 in a prolate spheroidal coordinate system . this would make the result for @xmath62 just as compact as that for @xmath190 when the equipotentials are ellipsoidal . however , the results are sufficiently manageable in terms of rectangular cartesian coordinates , at least for graphical purposes , that there is no real numerical advantage achieved by using prolate spheroidal coordinates . beginning students can appreciate why the normal to a prolate ellipsoidal surface bisects the angle formed by the two vectors from the foci to any point on the ellipsoid by thinking about the law of reflection applied to a light ray that strikes an ellipsoidal mirror while en route from one focus to the other . alternatively , a few banked shots on any good ellipsoidal billiard table should clarify the situation . the form in ( [ alternate3pointpotential ] ) is also more useful to obtain the correct result for @xmath1 since the form in ( [ 3pointpotential ] ) requires taking a limit @xmath298 to obtain a nontrivial result . just how real conductors of finite length and girth reduce to uniformly charged ideal line segments in 3d is a venerable problem with a dignified history , beginning with j c maxwell , on the electrical capacity of a long narrow cylinder , and of a disk of sensible thickness proc . london math . ix ( 1878 ) 94101 , reproduced as paper xcii , pp 672 - 680 , in _ the scientific papers of james clerk maxwell _ , vol . ii , w d niven , editor , cambridge university press ( 1890 ) . for a guide to the literature since then , see j d jackson , charge density on a thin straight wire : the first visit am . 70 ( 2002 ) 409 - 410 .

since strogatz and watts @xcite found the small - world effect and barabsi and albert @xcite found the scale - free property of the world wide web , researches on complex networks have greatly increased . these pioneering efforts sought to find some universal principles and invariants , but when people encounter a specific complex network , some questions often arise : `` can we identify the most important nodes in the network?'';``are these nodes more important than other nodes ? '' etc .. these problems are related to data analysis of complex networks . in this field , the node importance is a foundational problem . only when we know how to measure the importance of nodes , can we answer the questions mentioned above . the problem of the node importance appears in various fields . for example , studies of the most important scientists@xcite , the most dangerous terrorists@xcite , the most critical proteins or genes@xcite , etc .. in some circumstances , we may also need to know the relative important scientists , the relative dangerous terrorists , the relative critical proteins , etc .. besides , even if some real - world complex networks are not very large and with the help of computer visualization technologies , people can not yet understand their holistic features . for example , the internet@xcite . we need to find some subnetworks which are combined with some special nodes to obtain holistic understanding of the whole network@xcite . `` the importance of nodes '' is a vague concept , although many academic papers use it to describe the properties of networks . researchers have reached little consensus on this concept , but fortunately most concede that it can indeed be described by some rules or the intuitive ideas . as a reference , studies of social networks , which are certainly regarded as a kind of complex networks , inspired us . in the field of social networks , the importance of nodes is related to `` centrality''@xcite . moreover , in the field of web search engines , the importance of nodes often relies on the importance of neighborhoods , such as in pagerank@xcite and hits@xcite . some papers also propose new ideas on the definition of importance , such as the failure of system owing to the deletion of one node@xcite . considering that all these definitions are very different , when people deal with some specific applications , they may choose a subset of these definitions . based on the investigations on these definitions , we suggest that a set of four rules can be used to characterize the node importance in common circumstances . of course , the rules can be inserted or removed as needed to accommodate the particular real world situations . however , these rules often conflict to the others . we need a solution to resolve the conflicts . mathematically , each of the four rules defines an order relationship on the importance of nodes . for example , if we define the important nodes as those with more neighbors ( larger degree value ) then the degree value actually is a measure of the node importance . because the node importance needs more than one rule to characterize , when we have characterized the concept and chosen corresponding rules and formulas , we need to focus on how to deal with the conflicts among these order relationships and aggregate the results of the rules . usually , the aggregation should not have bias on any rule . there are two relative works on the aggregation . in the field of web search engines , people use a smart method named `` rank aggregation''@xcite to deal with this problem . in another way , if we treat every rule as an optimization object , this problem can be transferred to a multi - objective optimization problem . in the field of evolutionary algorithms for multi - objective optimization problems , this is commonly tackled with a mathematical tool , called the dominance relationship . the dominance relationship can categorize the nodes into the equivalence classes . every equivalence class will indicate the same ordinal number of the nodes in it . the first equivalence class would only include the most important nodes , which are also called `` the skyline''@xcite in the field of database management system ( dbms ) and `` the pareto front''@xcite in the field of evolutionary computation . here we suggest that the equivalence classes can be used to resolve the conflicts among the indicators based on the following reasons : 1 ) this approach has no bias on any rule ; 2 ) this approach can obtain diverse and representative important nodes ; 3 ) this approach can guarantee that a good node would certainly have a good ordinal , which is very important to the explanation of the ordered results ; 4 ) this approach has a mathematical foundation , i.e. , it is related to the maximum vector problem@xcite ; 5 ) this approach has been used in various fields , such as dbms and evolutionary computation . because pagerank and hits are very successful in web searching , some researchers believe that they will perform well in ranking the nodes of complex networks and then use them as benchmark algorithms to measure the effectiveness of the other algorithms @xcite , but whether they can perform well or not when applying them for very different purposes has not been discussed before . though pagerank and hits are perfect under their assumptions , the impossibility theorems restrict their extensions@xcite , so that they can not be taken for granted as benchmarks . assume that we set the intuitive rules as the benchmark , thus , are the results of pagerank and hits similar to the rules of the node importance here ? considering that both the results of the rules and the results of these algorithms can be represented as sequences , we suggest the kendall s @xmath0@xcite as the measure indicator . based on this idea , we define a measure indicator and three sub - indicators , and introduce an algorithm to calculate them . these sub - indicators can be regarded as the effectiveness of the ranking algorithms relative to the framework . to validate our ideas , we chose three well - known real - world networks , i.e. , the metabolic network@xcite , the dolphins network@xcite and the zachary karate club network@xcite as the examples to carry out the experiments . the experimental results show that the framework can produce representative and diverse nodes . moreover , we also calculate the effectiveness measure sub - indicators . the experimental results show that pagerank and hits perform well even though they were not designed for the undirected networks . however , the results also show that both algorithms have bias . in general , our work tries to provide a framework for issues on ranking / sorting the node importance . this framework addresses three main issues . first , how to reasonably define the node importance ? our solution is to clarify the concept of the node importance from the intuition by analyzing the concept with rules that define every aspect of this concept respectively . second , how to resolve conflicts among rules ? our solution is to aggregate the results of the rules by importing a mathematical tool that categorizes them into the equivalence classes and then assemble them into a partially ordered sequence . third , how to measure the effectiveness of the compared algorithms ? our solution is to compute the similarity between the results of the framework and that of the compared algorithm . we uses three widely used networks to validate the approach . finally , we apply the proposed approach to analyze the structure of the internet . the importance of nodes is related to many fields , and it is currently attracting more and more interests . the evaluation of the node importance is based on the researches in graph theory and graph - based data mining@xcite , and we can say that this field can be regarded as a branch of graph - based data mining . however , research on the node importance origins from the study of social networks . after the emergence of complex networks , other applications in technical networks , biological networks , etc . were also proposed . to clarify the node importance , we need to solve three main problems . the first one is that what are the suitable rules to define the node importance . the second one is how to resolve the conflicts among all the rules . the third one is how to measure the effectiveness . the importance of nodes was discussed in social networks as `` centrality '' . before the work of freeman , there is no unanimity on what is `` centrality '' and what is the conceptual foundation of `` centrality '' ; of course , there is no unanimity on what is the proper measure of `` centrality''@xcite . in 1976 , freeman reviewed the measure of `` centrality''@xcite and suggested three measure indicators for point ( node ) centrality . first , `` degree '' is a proper measure , it can be used to measure the communication activity in a network . second , `` betweenness '' is based upon the frequency with which a node falls between pairs of other nodes on the shortest or geodesic paths connecting them , and is used to exhibit a potential for control of the communication . third , `` closeness '' is based upon the degree to which a node is close ( approximates ) to all other nodes in the network and is used to exhibit the independence or efficiency of communication . these three measure indicators are referenced as three different structural attributes . the work of freeman has achieved a great success . many researchers use the `` centrality '' under freeman s suggestion now . the importance of nodes was defined recursively in pagerank and hits when dealing with the web documents , that is , the importance of a node relies on the importance of neighbors , and the neighbors relies on neighbors neighbors , etc .. essentially , the importance of nodes is related to the degree of nodes . based on this idea , pagerank and hits need to iteratively calculate the node importance . however , both these algorithms were designed to deal with directed graphs . because this paper mainly focuses on undirected graphs , the undirected vertex are treated as a couple of vertexes with opposite directions to make both the algorithms feasible . the importance of nodes also was defined as the failure of system@xcite . if a node is removed from a connected graph , and the new graph is not connected , then this node is important . actually , literature@xcite and the related papers emphasize on the importance of node sets . for a single node , this definition may include less information because the nodes mostly do not break the network only by itself . moreover , this definition can be partially represented by the betweenness . moreover , when considering the spreading process on complex networks , the k - shell@xcite can be used to define the most central , or the most important nodes . notice that this paper focuses on the static structure , the k - shell and the importance definition on failure@xcite are not applicable . moreover , the definitions of pagerank and hits can be represented as the neighborhood importance , therefore , four rules among these definitions are suggested to measure the node importance . if we regard every rule as an optimization object , the problem how to resolve the conflict among the rules could be transferred to a multi - objective optimization problem . in the field of multi - objective optimization , the simplest way is to set weights@xcite , which reflects the users preference for every rule . because the weights in the weighted sum method are uncertain , this method can not be used for a benchmark purpose . maybe we can refer to a smart technology named `` rank aggregation''@xcite in web search engine . the aim of the rank aggregation is to obtain a sequence which has a minimal distance to all ranking results from different measure indicators or ranking algorithms . if we use this method , that means we actually assume that the sequence with minimal distance would be most satisfactory to the definition of the node importance . this assumption is not necessary . moreover , this method would not guarantee that the good node has a good ordinal . for example , if node @xmath1 is more important than node @xmath2 under all the rules , the rank aggregation may set that node @xmath2 is more important than node @xmath1 to minimize the distance . in the field of multi - objective optimization , `` dominance relationship '' is used to deal with the conflicts@xcite . in contrast to the weighted sum method , this method does not need parameters , and moreover , it is not sensitive to the small difference of scores of nodes , because it is based on order relationships . this method can keep the diversity to obtain representative important nodes . for example , the best node under every rule should certainly be regarded as one of the most important nodes . furthermore , when using it as a benchmark , if node @xmath1 is better than node @xmath2 in all aspects , node @xmath1 would certainly be more important than node @xmath2 . this feature is useful for a reasonable explanation . moreover , the concept of `` pareto front '' is used to describe the set of `` the most important nodes '' . cotta and merolo@xcite referred to the work of scientific collaboration network @xcite and computed the degree , the betweenness and the closeness of the scholars majoring in evolutionary computation and obtained interesting results . they also listed the `` pareto front''the most important scholars . in this paper , we extend this method and used it as a good benchmark , and formally describe the detailed mechanism and introduce how to pack the results into a sequence and how to compare to other algorithms . when we have obtained the results from the benchmark and compared algorithm , how to define the effectiveness measure indicator(s ) is an important problem . because both the results can be expressed as the sequences , the effectiveness actually is the `` distance '' between two sequences . two popular measures are related to this issue : the spearman s @xmath3 and the kendall s @xmath0@xcite . kendall s @xmath0 is based on such an idea : how many times to transfer a sequence to the other by neighbor swapping . when applying kendall s @xmath0 to compute the effectiveness , the compared algorithms can `` cheat '' the measure indicator by generating too many equivalent nodes , so we use two sub - indicators to bound it and a sub - indicator to indicate the degree of `` cheating '' . here , we have introduced the problems , then we will introduce more detailed information on the solutions in the next subsections . people reached little consensus on the concept of `` the importance of nodes '' , but most concede that the node importance can be described by intuitive ideas . for example , `` degree '' is used as a measure indicator for the communication activity in human communication . a person with more social relationships is treated as more important ; and in a pure network , from one aspect , the nodes with larger degree are regarded as more important . this description on the node importance in networks is obviously based on intuition . therefore , when we discuss the node importance , we must extract the formal representation from the intuition . we choose four rules to evaluate the node importance according to the existing researches . 1 . 1 : commonly , the nodes with larger degree would have a larger influence on the network , so we think that the nodes with larger degree are more important . that is , if the degree of node @xmath1 is larger than that of node @xmath2 , then node @xmath1 is more important than node @xmath2 . 2 . 2 : if a node connects two or more communities , it is a key node , because this node can control the communication between the communities . we know , the potential of controlling the communication can be measured by betweenness . as such , if the betweenness of node @xmath1 is larger than that of node @xmath2 , then node @xmath1 is more important than node @xmath2 . 3 . 3 : if a node is closer to `` the center '' of the network , it is more important . because closeness is the proper indicator , we have , if the closeness of node @xmath1 is larger than that of node @xmath2 , then node @xmath1 is more important than node @xmath2 . 4 . 4 : if a node has neighbors with larger influence , then this node is more important . that is , if the neighbors of node @xmath1 have larger influence than that of node @xmath2 , then node @xmath1 is more important than node @xmath2 . the intuition of rules 1 , 2 and 3 come from the field of social networks , and the intuition of rule 4 from the web search engine . though we list four rules to describe the concept of the node importance , mathematical formalization still is necessary . we denote `` the importance of node @xmath1 is greater than that of node @xmath2 '' as @xmath4 , and `` the importance of node @xmath1 equals that of node @xmath2 '' as @xmath5 . according to the definitions above , the rules can be formalized as follows , rule 1 involves the degree as the measure indicator . we note the degree of node @xmath1 as @xmath6 . @xmath7 where , @xmath8 rule 2 can be measured by betweenness , @xmath9 would represent the betweenness of node @xmath1 and is defined as follows , assume that @xmath10 is the number of shortest paths ( geodesic paths ) between node @xmath11 and @xmath12 , and have @xmath13 , @xmath14 is the number of shortest paths between node @xmath11 and @xmath12 and through the node @xmath1 . then we have , @xmath15 rule 3 actually is based upon the closeness , we use @xmath16 to represent the closeness measure of node @xmath1 , @xmath17 to represent the length of the shortest paths between node @xmath11 and @xmath12 and @xmath18 to represent the set of all nodes in the network . then we have , @xmath19 as to rule 4 , we use a function of degrees of the neighbors of the node to measure it . we use @xmath20 to represent the set of the immediate neighbors of node @xmath1 , that is , all elements in @xmath20 would have and only have a hop / transimition from node @xmath1 . and we suggest such a function , @xmath21{{\sum\limits_{i \in nb(a ) } { degree(i)^{nk } } } } \ ] ] if @xmath22 then the neighboring nodes with larger degree would contribute more to the value of this indicator , but this may make a dilemma that the nodes with larger degree may be inferior to its neighbors with larger degree under this rule . for example , when the network structure is a star or a wheel , the central nodes would have an inferior indicator value . and v.s . , if @xmath23 , then the neighboring nodes with larger degree would contribute less , which is unfair , unless for some specific purposes . so we suggest @xmath24 . based on the definitions above , the rules would be depicted as follows , 1 . 1:@xmath25 2 . 2:@xmath26 3 . 3:@xmath27 4 . 4:@xmath28 5 . 5:@xmath29 four rules tell us the order relationships under different measure aspects . because they may be conflicting , we need a method to aggregate their results . here we employ a kind of strict partial order relationship called dominance relationship to order the nodes . for an illustration , we assume only two rules 1,2 and three nodes @xmath1 , @xmath2 , @xmath30 . assume that under rule 1 , @xmath31 ; and under rule 2 , @xmath32 ; then , for rule 1 , we set values of @xmath1 , @xmath30 , @xmath2 as 1 , 2 , 3 ; for rule 2 , we set values of @xmath30 , @xmath1 , @xmath2 as 1 , 2 , 3 . so @xmath1 , @xmath2 , @xmath30 would have coordinates @xmath33 . here , smaller number represents more importance . [ fig.ranking.sketch ] . + because node @xmath1 is more important than @xmath2 for rule 1 and rule 2 , so we say node @xmath1 * _ dominates _ * node @xmath2 or , @xmath2 * _ is dominated _ * by node @xmath1 . obviously , @xmath1 and @xmath30 are not dominated by any other nodes , so we say they rank 1 . if we erase all non - dominated nodes , that is , node @xmath1 and node @xmath30 , from fig . [ fig.ranking.sketch ] , then node @xmath2 would not be dominated by any other nodes , so we say node @xmath2 ranks 2 . for each node , it is assigned four values to represent the ordinals under the corresponding rules . that is , the node importance can be represented by a vector with four elements . for node @xmath1 , we denote @xmath34 , here @xmath35 represents the ordinal value under rule 1 , @xmath36 represents the ordinal value under rule 2 and so on . therefore , rule 1 to rule 4 can be rewritten as follows , 1 . 1:@xmath37 2 . 2:@xmath38 3 . 3:@xmath39 4 . 4:@xmath40 according to the rules , the set of the importance vector of all nodes is a strict partial order set based on the dominance relationship and satisfies the irreflexive , transitive , asymmetric relationships . because the dominance relationships are a kind of strict partial order relationship , and a strict partial order relationship should respond to a non - strict partial order relationship '' over a set which is reflexive , antisymmetric , and transitive ; a strict partial order relationship `` @xmath41 '' is a binary relation over a set which is irreflexive and transitive , and therefore asymmetric . ] . under the non - strict partial order relationship , the nodes can be packed into the equivalence classes , that is , all the nodes in the same equivalence class are equivalent in certain relationship . we give formal definition for the dominance relationship below , given node @xmath1 and node @xmath2 , @xmath42 ( m is the number of the selected rules ) is said to dominate @xmath43 ( denoted by @xmath44 ) , if and only if @xmath1 is less than or equivalent to @xmath2 for each element and @xmath45 , i.e. + @xmath46 given a set a , the non - dominated set ( denoted as nds ) is the set that contains all the nodes which are not dominated by any other nodes . mathematically , _ nds _ is defined as : @xmath47 here nds is `` the pareto front '' , or `` the skyline set''@xcite . mathematically , nds is the maxima of a set of vectors@xcite . given a set a , the elements in @xmath48 would rank 1 . @xmath49 is the dominated set , and if it is not empty , it would have its nds , all the elements in @xmath49 s nds would rank 2 . every nds in different hierarchies is an equivalence class . when we set the ordinal to every node , we can categorize all the nodes into the equivalence classes based on iterative non - dominated sets , and every equivalence class means a ranking number . therefore , the procedure to rank the nodes by the importance actually is the procedure to iteratively obtain the non - dominated set from the remaining set which has eliminated the non - dominated nodes . once we determine the measure indicators of the importance , then we have determined the results of the equivalence classes based on the dominance relationship , that is , the iterative non - dominated sets . so we need an algorithm to compute the equivalence classes . in the field of dbms , several algorithms have been proposed to compute the skyline set efficiently . for examples , block - nested - loops algorithm , divide - and - conquer algorithm and b - tree algorithm proposed by brzsnyi et al.@xcite , the progressive skyline algorithms proposed by tan et al.@xcite and the online algorithm proposed by kossmann et al.@xcite . these algorithms can also be used to compute the equivalence classes efficiently by iteratively computing the skyline set . here we just introduce a simple algorithm , regardless of efficiency , to categorize the nodes into the equivalence classes . assume that the input is a set named @xmath50 , we define a set named @xmath48 to store the non - dominated nodes , and a set named @xmath51 to store dominated nodes in current iteration step . when an iteration step is executed , the equivalence classes and the rank numbers of the equivalence classes can be output step by step . the pseudocode of this algorithm is depicted as fig . [ fig.pseudocode.ec ] . + once the rules are determined , the nodes could be categorized into the equivalence classes by their importance . the nodes in the same equivalence class will share the same rank number , so all the nodes could be packed into a partially ordered sequence . but the further problem would be , because several algorithms have been used to sort / rank the nodes by the importance , how to measure the effectiveness of algorithms ? i.e. , how to measure the similarity between the aggregation of the results of the selected rules from intuition and the results of the ranking algorithms ? pagerank@xcite and hits@xcite are the most famous algorithms and they were designed to automatically determine the importance of web pages . here we use them for the analysis of complex networks like the previous works@xcite , and we try to investigate how much pagerank and hits are similar to the equivalence classes approach . pagerank and hits are the `` global ranking / sorting '' , that is , it will generate a totally ordered sequence . a totally ordered sequence means that every element in this sequence can be ordered by a property and has an identical ordinal . because both the results of the rules and the results of the compared algorithm can be expressed as sequences , we can use the distance between the partially ordered sequence generated by the equivalence classes approach and the totally ordered sequence generated by the compared algorithms to measure the effectiveness of the algorithms . if we use the partially ordered sequence as a benchmark for comparison , every opposition to the partially ordered sequence is regarded as a unit of distance . according to this idea , the distance can be defined . actually , the distance is an indicator of the similarity . when we know the maximum of distance , we can define the similarity ratio . we use the term `` coverage '' to represent the similarity ratio . we use a short example to illustrate the concept of coverage . assume that there is a partially ordered sequence , denoted as @xmath52,[3,4]\}$ ] which means @xmath53 is an equivalence class , and @xmath54 is another , and a totally ordered sequence , @xmath55 which means @xmath56 . then how to measure the distance between two sequences ? we use the counting of swapping between the neighboring nodes to define the distance , that is , how many times we swap the neighbors to transfer the totally ordered sequence to meet the partially ordered sequence , the least number of actions can be defined as the distance . as to the example , we need at least 4 times , that is , @xmath57 ; therefore , the distance is 4 . actually , the distance means the violation of the order relationships to the partially ordered sequence . notice that the position relationship between @xmath58 and @xmath59 ( similar for @xmath60 and @xmath61 ) is not considered . once the distance is defined , the coverage can be defined as the opposition of the ratio of violation . as to the example , because 4 is the possible largest value for all possible totally ordered sequence , so the coverage is zero . in many circumstances , the algorithms would estimate some inequivalent nodes as equivalent , so we define two sub - indicators the best and worst coverage to evaluate the coverage in the best and worst circumstances , that is , if we want to get the worst coverage and the compared algorithms obtain a sequence @xmath62 , 1\}$ ] , that means no . 3 node equals to no . 2 node , because for the partially ordered sequence , no . 2 node is prior to no . 3 node , so we rewrite it as @xmath63 , 1\}$ ] , such that the relative position of no . 2 node and no . 3 node is different from that of the partially ordered sequence , here , the coverage is the worst coverage . here we formally depicted the concept of coverage as follows , given two sequences @xmath64 and @xmath65 , if @xmath66 $ ] such that @xmath67 and @xmath68 and for arbitrary @xmath69 , then we say , there exists one time of neighbor swapping to transfer the sequence @xmath1 to @xmath2 , denoted @xmath70 . given two sequences @xmath1 and @xmath2 , if there exist the least k sequences @xmath71 such that @xmath72 , ns(s_j , s_{j+1})=1 \wedge ns(a , s_1)=1 \wedge ns(s_k , b)=1 $ ] , then we say @xmath73 and the distance from @xmath1 to @xmath2 is @xmath74 . a partially ordered sequence can actually be regarded as a set of sequences , so the distance between a sequence and a partially ordered sequence is the minimum distance from a sequence to a set of sequences . given a sequence @xmath1 and a sequence set @xmath50 , the distance from @xmath1 to @xmath50 is @xmath75 . for a given network , assume that the partially ordered sequence of the equivalence classes approach is @xmath76 , and the compared sequence is @xmath77 , and all the possible sequences of @xmath76 would construct a set denoted @xmath78 . then the coverage can be depicted as follows , @xmath79 when the sequences of the compared algorithms are totally ordered sequences , this indicator is enough to characterize the similarity between these sequences to the sequence of the equivalence classes approach . but some algorithms would generate some equivalent nodes , then , how to define the similarity ratio ? as to the sequences with some equivalent nodes , we can regard it as a sequence set and choose the best sequence and the worst sequence in the sequence set to represent it , therefore , we define the best coverage and the worst coverage . if we regard the compared sequence with the equivalent nodes as a set of sequences ( denoted @xmath50 ) , we have the definition of two sub - indicators , best coverage and worst coverage , based on coverage as follows , @xmath80 notice that the best coverage does not measure the best effectiveness , because when all the nodes are in an equivalence class , the best coverage is 1 , here it is meaningless , because the worst coverage is 0 . the worst coverage is a more proper measure indicator to measure the worst effectiveness . and the gap between the best coverage and the worst coverage would mean the number of nodes which are incorrectly regarded as equivalent nodes . here we define other sub - indicator `` certainty ratio '' that are denoted as `` certratio '' to indicate the incorrect degree . if the compared algorithm can identify the equal nodes and does not regard the inequivalent nodes as equivalent , its certratio is 0 . actually , if the algorithm treats every nodes uniformly , and does not import extra information , the symmetric nodes will not be assigned different ranking values . under such an assumption , the certratio is only related to the gap between the best coverage and the worst coverage . here we define the certratio as equation [ eqn.certratio ] , which is not suitable for some special algorithms which include extra information . @xmath81 the calculation of coverage includes two steps . the first step is the calculation of the distance between @xmath77 and @xmath76 . the minimum neighbor swapping seemly is difficult to be counted , but there is an easy algorithm to solve this problem . every node would have two property values . one is the rank number , the other is the importance value assigned by the compared algorithms . if we sort the nodes by the importance value , then the distance is the counting of neighbor swapping to sort the nodes by the rank number . for example , if there are four nodes @xmath82 , the rank numbers are @xmath83 . assume that the compared algorithm has generated a sequence , that is , @xmath84 , so the corresponding rank number sequence is @xmath85 , the minimum counting of neighbor swapping is the counting of neighbor swapping to ascendingly sort the rank number sequence . to obtain the minimum of neighbor swapping , no neighbor swapping could be duplicated . we use the greedy strategy to deal with the problem . that is , we define an array which records the difference between every node and its successor , then the maximum difference is preferred to be eliminated , that is , this corresponding node and its successor is firstly swapped . so we use the array @xmath86 $ ] ( here n is the number of the nodes ) to record the difference of the array of the nodes @xmath87 $ ] , and iteratively swap the nodes with maximum difference to its successor . notice that when the rank value sequence are sorted , the elements in array @xmath86 $ ] are -1 or 0 , so the stopping criteria of this algorithm is that the maximum of elements in @xmath86 $ ] is smaller than zero . the algorithm can be depicted as fig . [ fig.pseudocode.distance ] . + the second step is the calculation of maximum possible distance to the partially ordered sequence . the intuitive idea is to construct the furthest sequence and count the neighbor swapping . but there is a simper formula . assume that the nodes are categorized into @xmath88 equivalence classes denoted as @xmath89 , and the sizes are @xmath90 . notice that the maximum distance between arbitrary nodes is @xmath91 , each equivalence class @xmath92 would result in a subtract of @xmath93 , so we have , @xmath94 based on the methods above , the best coverage and the worst coverage is easy to calculate , because we only need to deal with the nodes with equivalent importance . before the calculation of the minimum of the distance , if we sort the nodes with the same importance according to the same orders in the sequences of the equivalence classes approach , then we obtain the best sequence . if according to reverse orders , the we obtain the worst sequence . we choose three real - world networks to validate the proposed approach . the selected networks are the metabolic network , the dolphins network and the zachary karate club network . these three real - world networks are widely used as examples . protein interaction networks are deeply investigated in biology . the metabolic network is a kind of protein interaction network . here we chose a single module in the metabolic network of _ thaliana _ , because the module is simple , but it has many typical features . there are 30 nodes and 34 edges in this network and the edge represents that two nodes participate in the same metabolic activities . see in fig . [ fig.netdraw.metabolic ] . + the dolphins network owes to the work of d. lusseau . dr . d. lusseau observed a community of 62 bottlenose dolphins ( tursiops spp . ) over a period of 7 years from 1994 to 2001 . according to the observation they constructed the dolphins network . the nodes represent the dolphins , and the ties between nodes represent associations between dolphin pairs occurring more often than expected by chance . node 37 is a very important node , when node 37 disappeared , the network broke into two communities , one big and one small , but when it returned , two communities got united . the whole network has 62 nodes and 159 ties . see in fig . [ fig.netdraw.dolphins ] . + the zachary karate club network is a famous social network . wayne zachary observed a karate club in a university for a period of three years , from 1970 to 1972 . in addition to direct observation , the history of the club prior to the period of the study was reconstructed through informants and club records in the university archives . the nodes represent the members of the club , and the edges represent the friendship between the nodes . the 34th node , john a. is the chief administrator and the 1st node , mr . hi is the karate instructor . because of the conflicts between mr . hi and mr . john , the network broke into two clubs . this network includes 34 nodes and 78 edges . see in fig . [ fig.netdraw.zachary ] . + based on the proposed approach , we calculated the equivalence classes of the tested networks . here we set @xmath95 . besides , we also computed the results of the measure indicators of degree , betweenness , closeness with the tool ucinet@xcite . the equivalence classes of the metabolic network are listed in table [ tab.po.metabolic ] . @xmath96 [ tab.po.metabolic ] from table [ tab.po.metabolic ] , we can see that no . 9 node belongs to the first class of nodes , because no . 9 node has better values of degree , betweenness , closeness and neighbors indicators . 9 node undoubtedly is the most important protein . according to the same approach , we list the equivalence classes of the dolphins network and the zachary karate club network in table [ tab.po.dolphins ] and table [ tab.po.zachary ] . @xmath97 [ tab.po.dolphins ] from table [ tab.po.dolphins ] , we can see that the nodes no . 2 , 15 , 37 , 38 and 41 are the most important . here node 37 is the most important node between two communities , node 2 belongs to the small community and the nodes no . 15 , 38 and 41 belong to the big community . here this result shows that the equivalence classes approach can obtain representative nodes . @xmath98 [ tab.po.zachary ] from table [ tab.po.zachary ] , we can see that nodes no . 1(mr . hi ) and 34 ( the chief administrator ) are the most important , because they are the kernel nodes in the separated clubs . all the results show that the equivalence classes approach can identify the most important nodes . pagerank and hits are two famous algorithms to sort / rank web documents by importance . web documents are linked by the hyperlinks and construct a directed network . here , they are used to deal with the undirected network , the same as that the other researchers did . we calculate their effectiveness with the proposed framework to check this kind of applications . moreover , to compare their results to four chosen basic measure indicators , i.e. , degree , betweenness , closeness , neighbors , we compute the coverage values of these four indicators . when computing the importance score by pagerank and hits , we set the transition probability as 0.15 and the iteration times as 200 for pagerank , the maximum of iteration times as 500 for hits . the results here is obtained by the tool jung@xcite . @xmath99[0cm]{pagerank } & best & 0.814721 & 0.868226 & 0.885081 \\ \cline{2 - 5 } & worst & 0.814721 & 0.868226 & 0.885081 \\ \cline{2 - 5 } & certratio & 0.000000 & 0.000000 & 0.000000 \\ \hline \raisebox{-1.50ex}[0cm][0cm]{hits } & best & 0.809645 & 0.759840 & 0.895161 \\ \cline{2 - 5 } & worst & 0.809645 & 0.759840 & 0.895161 \\ \cline{2 - 5 } & certratio & 0.000000 & 0.000000 & 0.000000 \\ \hline \raisebox{-1.50ex}[0cm][0cm]{degree } & best & 0.984772 & 0.926982 & 0.991935 \\ \cline{2 - 5 } & worst & 0.723350 & 0.848260 & 0.866935 \\ \cline{2 - 5 } & certratio & 0.261422 & 0.078722 & 0.125000 \\ \hline \raisebox{-1.50ex}[0cm][0cm]{betweenness } & best & 0.977157 & 0.937250 & 0.977823 \\ \cline{2 - 5 } & worst & 0.878173 & 0.918426 & 0.868952 \\ \cline{2 - 5 } & certratio & 0.098984 & 0.018824 & 0.108871 \\ \hline \raisebox{-1.50ex}[0cm][0cm]{closeness } & best & 0.829949 & 0.881346 & 0.943548 \\ \cline{2 - 5 } & worst & 0.819797 & 0.869937 & 0.913306 \\ \cline{2 - 5 } & certratio & 0.010152 & 0.011409 & 0.030242 \\ \hline \raisebox{-1.50ex}[0cm][0cm]{neighbors } & best & 0.890863 & 0.879064 & 0.907258 \\ \cline{2 - 5 } & worst & 0.832487 & 0.858528 & 0.899194 \\ \cline{2 - 5 } & certratio & 0.058376 & 0.020536 & 0.008064 \\ \hline \end{array}$ ] [ tab.effectiveness ] from table [ tab.effectiveness ] , we can see that all the certainty ratios of pagerank and hits are equal to zero , meaning that both algorithms can successfully determine the nodes with equivalent importance and do not disregard the inequivalent nodes as equivalent . moreover , hits outperforms pagerank in the zachary network , but fails in the metabolic network and the dolphins network . from table [ tab.effectiveness ] , we can see that both algorithms outperform the degree indicator . moreover , the closeness indicator outperforms both the algorithms in all the best coverage values and the worst coverage values . hits gets a worst score in the dolphins network and pagerank gets a worst score in the metabolic network . in general , as to pagerank and hits , they get scores of about 76% - 90% for similarity ratio , and 100% for certainty ratio . this proves their good effectiveness , although they were not designed for the undirected networks . the statistic results need an explanation in details . here we choose the metabolic network as the anatomic object . the experimental results can illustrate the preferences of both the algorithms . here we list the results of both two algorithm in table [ tab.pagerankmetab ] and [ tab.hitsmetab ] . @xmath100 [ tab.pagerankmetab ] from table [ tab.pagerankmetab ] we can see that pagerank prefers to the degree indicator and marginal nodes . for examples , though node 7 is more important than node 5 in all aspects , pagerank sets a worst ranking value ; the relationships among the nodes no . 1 , 4 , 6 , 8 , 17 and 20 clearly demonstrate the marginal preference . this is the reason that pagerank does not obtain higher scores . @xmath101 [ tab.hitsmetab ] from table [ tab.hitsmetab ] , we can see that hits prefers to the neighbors and closeness . for examples , in the results of hits , node 8 and node 17 are more important than node 5 , node 25 is more important than node 26 . these preferences would lead to a result that important bridge and small community may be somehow ignored . the bad score of hits on the dolphins network also can be explained on these preferences . the bias of hits can be explained by itself . hits determines two values for a node : the authority and the hub value , which are mutually recursively defined . the authority value of a node is the sum of the scaled hub values that point to this node . the hub value of a node is the sum of the scaled authority values of the nodes that this node points to . when all the links are two - way , because of the iterations , the central nodes are emphasized and certainly , the nodes connecting to the central nodes are also emphasized . generally speaking , pagerank and hits are both good to identify the node importance , but we must be aware of their bias . based on bgp tables posted at archive.routeviews.org , mark newman reconstructed a network representing the structure of the internet . in this network , each node represents a domain and each edge represents an inter - domain interconnection . this network has 22963 nodes and 48436 edges . this paper uses the equivalence classes approach to analyze the internet , and lists the top 10 equivalence classes as table [ tab.po.internet ] . @xmath102 [ tab.po.internet ] for table [ tab.po.internet ] we can see that the most important nodes are no . 4 , 15 , 23 , 27 . as we have known , the internet is a scale - free network@xcite . a few hub nodes have a great amount of links , and most nodes have a few links . the internet also have the rich - club phenomenon@xcite , that is , the hub nodes tend to link to the others . this paper retracts a sub - network which only contain the nodes in table [ tab.po.internet ] with the proposed approach , and the result demonstrates that the internet has the rich - club phenomenon with the visualization technology as fig . [ fig.internet ] . from fig . [ fig.internet ] , the average degree of this subnetwork is 31.0423 , contrarily , the whole network is 2.1093 , meaning that the hub nodes forms a dense kernel . from fig . [ fig : subfig : a ] , the most important 4 nodes are red , and they form a complete graph . moreover , the proposed approach can demonstrate the hierarchy of the internet . as shown in fig . [ fig : subfig : b ] , the red nodes form a very dense graph , however , the graph lacks some links to become a complete graph , that is , if the node is more important , it is more likely linked by the kernel . this paper proposes a framework to investigate the node importance . the proposed framework suggests four rules to characterize the node importance and suggests the use of the equivalence classes to describe the relationship among nodes because of the desired features of the equivalence classes approach . this paper also suggests that the equivalence classes approach can be used as a benchmark to measure the effectiveness of the other ranking / sorting algorithms , and proposes three sub - indicators based on the coverage indicator . this paper demonstrates how to apply this framework . three real - world networks are used to calculate the effectiveness of pagerank and hits , and the experimental results show that both algorithms perform well . moreover , the analysis on the metabolic network showed that : pagerank would wrongly order some nodes because of the bias on degree and marginal nodes ; hits has the biases on neighbors and closeness . these results imply that we must be aware of the bias of pagerank and hits when using them as a benchmark in such a field . on the other hand , considering the scores of both the algorithms , they may be capable of being used as rapid algorithms on this issue . from the experimental results we can see that this framework is feasible . moreover , this paper applies the proposed approach to analyze the internet . the experimental results show that the internet has a kernel with dense links and have the rich - club phenomenon with the computer visualization technologies . in the future , we will extend it to more complicated cases , for example , the directed and weighted networks . the proposed method is potential to be applied into the analysis of critical proteins and find the drug targets and other fields . the authors thank for ucinet and jung for the convenience . the authors gratefully thank the editors and the anonymous reviewers for the improvement of the quality of this paper . c. winter , j. roy , g. kristiansen , s. kersting , d. aust , t. knosel , p. rummele , b. jahnke , v. hentrich , f. ruckert , m. niedergethmann , w. weichert , m. bahra , h. j. schlitt , u. settmacher , h. friess , m. buchler , h .- saeger , m. schroeder , r. grutzmann , c. pilarsky . , google goes cancer : improving outcome prediction for cancer patients by network - based ranking of marker genes , plos computational biology 8 ( 5 ) ( 2012 ) e1002511 . j. horn , n. nafpliotis , d. e. goldberg , a niched pareto genetic algorithm for multiobjective optimization , in : proceedings of the first ieee conference on evolutionary computation , ieee world congress on computational intelligence , vol . 1 , piscataway , new jersey , 1994 , pp . 8287 . j. han , x. yan , p. s. yu , mining , indexing , and similarity search in graphs and complex structures , in : data engineering , 2006 . proceedings of the 22nd international conference on , 2006 , pp . 106106 .

in recent years the inner structure of dark matter haloes has been the topic of some discussion . observations suggest that the dark matter distribution in disk galaxies has a roughly constant density core , with a typical size of order a few kpc . a recent analysis suggests a shallow power - law mass density distribution @xmath1 with @xmath2 @xcite . on the other hand , numerical simulations based on the ( @xmath0)cold dark matter ( cdm ) paradigm suggest a very steep inner mass density distribution , a so - called `` cusp '' @xcite [ nfw ] . the most recent simulations suggest slopes with @xmath3 @xcite . there is strong consensus that this slope can not be as shallow as the observations suggest . the rotation curves of low surface brightness ( lsb ) galaxies are considered particularly clean tests of the cdm paradigm at galaxy scales . lsb galaxies are dominated by dark matter @xcite , and their dynamics should thus give a more or less direct map of the dark matter distribution . lsb galaxy rotation curves , together with the missing dwarf or substructure problem ( e.g. @xcite ) , currently form the most serious challenges for the cdm model . there has been much debate on this topic in recent years . this is however not the place to review this debate , and the reader is referred to the summaries by e.g. @xcite and @xcite and references therein . on the theoretical side the `` universal mass density profile '' ( with its characteristic @xmath4 cusp ) as proposed by @xcite , has long been a corner - stone of cdm ( but see @xcite ) . recent simulations presented in @xcite and @xcite [ hereafter h04 ] show that simulated cdm haloes may actually contain a range of slopes in their inner regions ( though the slopes remain steep ) . the shallowest inner slope measured is @xmath5 , with an average over the simulated haloes of @xmath6 ( at the innermost reliably resolved radius for galaxy mass haloes ) . with this scatter in mind , h04 argue that simple fitting functions ( such as `` the '' nfw halo ) do not capture the full variety and diversity in shapes and slopes , and suggest that this may partly explain the discrepancy between observations and simulations . they claim that the inner slope is difficult to constrain , both observationally as well as theoretically and point out that rotation curve constraints are strongest where numerical simulations are least reliable . this , as well as the potential for systematic effects in the observations , has , at least according to h04 , led to an `` unwarranted emphasis '' on the value of the inner slope , at the cost of evaluation of the data and simulations over their full radial extent . h04 present a novel method to fit rotation curves derived from their simulations as well as lsb galaxy rotation curves from the literature . their method is able to describe the variety present in data and simulations and from the distributions of respective fitting parameters h04 conclude that the observed curves _ are _ consistent with the cdm paradigm ; a conclusion that contradicts a large body of observational work . for their analysis , h04 make use of the rotation curves presented in @xcite , @xcite and @xcite , and derive circular velocity curves for haloes from their simulations . to quantify both sets of observed and simulated rotation curves they use a three - parameter fitting formula ( see e.g. @xcite ) : @xmath7 here @xmath8 is the asymptotic velocity of the flat part of the rotation curve , @xmath9 is a scale radius ( the transition radius between the rising and flat part of the rotation curve ) and the parameter @xmath10 describes the abruptness of the turn - over between the rising and flat parts of the rotation curve . a higher value of @xmath10 results in an abrupt transition , a low value in a more gradual turn - over . note that this function was designed to fit observed rotation curves : it specifically assumes a solid - body rise in the inner parts and a flat rotation curve in the outer parts . h04 fit the two sets of rotation curves with this function , constraining the fit parameters slightly by insisting that 1 . @xmath11 , 2 . @xmath12 , 3 . @xmath13 , with @xmath14 the maximum observed velocity in the rotation curve . fits with @xmath15 [ rejected by condition ( 1 ) ] correspond with very abrupt transitions between the rising and flat parts of the curve . condition ( 3 ) prevents some of the solid - body rotation curves where the asymptotic velocity is not well - constrained , from producing fits with extremely large @xmath8 . conditions ( 1)(3 ) thus help removing unrealistic values for the fit parameters . figure [ fig : hayashidata ] ( after h04 , their fig . 9 ) shows the results presented in h04 . the lsb rotation curves show a broad distribution in @xmath10 with two pronounced peaks : one around @xmath16 and a narrower one at @xmath17 . some 70 per cent of the fits have @xmath18 . the @xmath10 distribution of the simulated cdm haloes is markedly different . it is narrow and centered around @xmath19 with a dispersion of @xmath20 . at first sight the two distributions seem quite incompatible , and one might conclude that the lsb rotation curves are not consistent with the cdm simulations . h04 argue however that for most of the lsb galaxy rotation curves the reduced @xmath21 distribution is quite broad and shallow , so that by constraining the fit parameters slightly more , a fit can be found that is consistent with the simulations , at the cost of only a small increase in @xmath22 . they define these cdm - compatible constraints as : 1 . @xmath23 2 . @xmath12 3 . @xmath13 4 . @xmath24 . here @xmath25 is the average density within the radius where the velocity reaches half its maximum value ( alam et al 2002 ) . see h04 for a justification of these additional constraints , here it is sufficient to note that for the lsb galaxies @xmath25 is derived from the fit of eq . ( 1 ) to the rotation curves , whereas @xmath26 is the predicted value for a galaxy with @xmath27 in the @xmath0cdm model presented in @xcite and @xcite ( see also fig . 2 in @xcite and fig . 11 in h04 ) . the constraint on @xmath25 ensures that only fits are considered with average densities close to those predicted in a standard @xmath0cdm universe . h04 find that for the majority ( @xmath28 70 per cent ) of the galaxies fits can be found for which both the best - fit @xmath29 [ using constraints ( 1)-(3 ) ] and the constrained @xmath30 [ using constraints ( i)-(iii ) ] . this group of galaxies ( group a using the terminology in h04 ) is thus considered to be consistent with cdm . for the remaining galaxies no cdm - compatible fit could be found ( group b in h04 ) , or the curves were too irregular to be well fitted by eq . ( 1 ) ( group c in h04 ) . the main claim in h04 is thus that most of the lsb galaxies can still be fitted reasonably well with these extra constraints and without a large increase in @xmath22 . they conclude , based on the agreements of the constrained , cdm - compatible @xmath10 and @xmath25 distributions with those of the simulations , that `` this sample of lsb rotation curves is not manifestly inconsistent with the predictions of @xmath0cdm cosmological models . '' this conclusion thus contradicts most observational work on lsb rotation curves which argues the opposite . in this paper i discuss these conclusions and investigate the method with which they are derived . i find that the method presented in h04 can not distinguish between shallow and steep mass - density slopes , and illustrate this by showing that with this method even core - dominated , pseudo - isothermal halos would be inferred to be consistent with cdm . i introduce an additional criterion that does take the inner mass - density slope into account , and find that only a quarter of the lsb galaxies investigated are possibly consistent with cdm . however , most of these galaxies fit parameters are so weakly constrained that this is not a very strong conclusion . only 3 out of the 51 galaxies investigated have tightly constrained solutions that are not significantly inconsistent with cdm . however , two of them are high surface brightness galaxies that are likely dominated by their stellar population . the net result is that in the total sample there is only one galaxy that is likely to be dark matter dominated and not significantly inconsistent with cdm . lsb galaxies that are consistent with cdm seem to be rare indeed . in sect . 2 i derive an analytical expression for the slope of a rotation curve , based on eq . 3 compares the slopes in both data and simulations at identical physical radii . in sect . 4 i show that the constraints imposed by h04 are not sufficient to prove or disprove agreement with cdm . in sect . 5 a new constraint is introduced that does take the slope into account . the results are summarised in sect . equation ( 1 ) is a flexible , analytical function that can describe a wide range of rotation curves ( see @xcite for more examples ) . one can , analogous to h04 , use the fitting function to investigate the properties of the underlying data . assuming a spherical halo dominated by dark matter we can use the inversion used in e.g. @xcite to derive the corresponding mass - density distribution : @xmath31 in combination with eq . ( 1 ) yields @xmath32.\ ] ] the logarithmic slope of the mass density distribution can now be derived from eq . ( 3 ) : @xmath33 the logarithmic , or power - law slope @xmath34 thus _ depends on both @xmath10 and @xmath9_. fig . [ fig : courslope ] shows @xmath34 for a number of values of @xmath10 and @xmath9 , where the values have been chosen to cover the range shown by the lsb rotation curve fits from h04 . changing @xmath10 has the effect of changing the shape of the slope - radius relations ; changing @xmath9 shifts the relations horizontally without affecting the shape for any particular @xmath10 . most of the plotted slopes are remarkably _ shallow _ at small radii . an inner asymptotic value of @xmath35 is predominant . the cdm slope @xmath5 can only be reproduced with eqs . ( 1 ) and ( 4 ) for small values of @xmath9 and/or extremely small values of @xmath10 . similarly , at large radii the slope converges to @xmath36 . these are not naturally occurring values for @xmath0cdm mass profiles , with asymptotic slopes @xmath37 in the inner parts and @xmath38 in the outer parts . equation ( 1 ) is really built to mimic the same striking features of rotation curves that led to the introduction of the pseudo - isothermal ( iso ) halo , namely a solid - body rise in the inner parts , and a constant ( `` flat '' ) rotation velocity in the outer parts . one needs to make very specific choices of fitting parameters to make this function resemble an nfw profile . the h04 simulations predict steep slopes , even at the smallest reliably resolved radii , defined by the so - called convergence radius @xmath39 @xcite . for the h04 dwarf galaxy models @xmath40 kpc or @xmath20 kpc for @xmath41 . the slope at @xmath39 in the simulations varies between @xmath5 and @xmath42 . equation ( 4 ) can be used to derive the slopes implied by the fits to the observed lsb rotation curves . for this one needs to choose a radius at which the slopes will be evaluated . in the rest of this paper two choices will be used . firstly , the radius of the innermost measured point of the rotation curve as given in @xcite and @xcite will be considered . here a value of @xmath43 is used ( converted using the appropriate distance ) , which is the typical spacing between data points for the curves presented there . ] . one might argue that at @xmath44 a different radius is probed for each rotation curve , so , secondly , a constant radius @xmath45 kpc @xmath46 kpc for all curves will be used as well . one should not expect dramatically different results from both choices though : the average value of @xmath47 is @xmath48 kpc , very much comparable with @xmath39 . it should be kept in mind that even though we evaluate the slope at these particular choices for the radius , the @xmath10 and @xmath9 parameters ( and therefore the slope ) are constrained by the _ entirety _ of the curve . figure [ fig : hayslope3 ] shows the distribution of inferred mass density slopes derived using the lsb rotation curves and the best - fit @xmath10 and @xmath9 parameters listed in table 2 of h04 . the slopes for both @xmath49 and @xmath50 kpc are shown . it is clear that the distributions are heavily biased towards shallow slopes and very different from the values found in the simulations at these radii @xmath51 . the small values of @xmath52 derived in h04 indicate that eq . ( 1 ) fits the data well , and the problem is thus not with the quality of the fits . in fig . [ fig : hayslope3 ] is also indicated the distribution of slopes for the galaxies that h04 claim are consistent with cdm ( their group a with @xmath53 and @xmath54 ) . this distribution is not markedly different from that of the entire sample . from fig . [ fig : hayslope3 ] one could thus conclude that the majority of slopes are inconsistent with the cdm prediction . how can this be reconciled with the h04 conclusions ? as discussed above , h04 add an extra constraint to the @xmath10-parameter , and argue that the resulting agreement with the range of @xmath10-values shown by the simulations implies consistency with cdm . this point will be explored in more detail below , but one can already see here that this does not solve the problem of the discrepancy between observed and simulated slopes . consider the distribution of slopes of the subset of group a for which @xmath55 ( and both @xmath56 ) , i.e. galaxies for which the best fit is ( apparently ) already consistent with cdm . the distribution of slopes of this sub - group is also plotted in fig . [ fig : hayslope3 ] . the selection criteria for this sub - group favour galaxies with steeper slopes , and it should thus come as no surprise that the dominant peak at @xmath57 has disappeared . nevertheless it is remarkable that the distributions are still dominated by slopes @xmath58 . the distribution using @xmath59 peaks at @xmath60 . the distribution using @xmath61 kpc peaks at @xmath62 . as we are probing the slope at @xmath63 kpc and not in the centre , we should not expect all iso haloes to have a flat slope at this radius . the distribution of slopes found is consistent with that expected at @xmath45 kpc for an ensemble of pseudo - isothermal ( iso ) halos with core - radii between @xmath64 and @xmath65 kpc . the decrease in mass - density at this radius is thus much less steep than expected for a cdm halo . the steepest slope expected for a realistic iso halo ( i.e. consistent with the smallest value @xmath66 kpc measured in real lsb galaxies ; see results in @xcite ) is @xmath67 , with the large majority of iso haloes having slopes less steep than that . the slopes measured in the h04 simulations at @xmath45 kpc vary between @xmath68 and @xmath42 , firmly inconsistent with the observed distributions . the reason for the discrepancy can best be appreciated by realizing that according to eq . ( 4 ) the slope is a function of both @xmath10 and @xmath9 . this is shown graphically in fig . [ fig : gammartfits ] which shows iso - slope contours in the @xmath69 plane , as evaluated at @xmath63 kpc . it is immediately obvious that @xmath70 is not a unique function of @xmath10 , but depends equally strongly on @xmath9 . most combinations of @xmath10 and @xmath9 with @xmath71 and @xmath72 yield slopes shallower than @xmath73 . only a very small part of parameter space results in steep , cdm - compatible slopes ( essentially only the area with @xmath74 and/or @xmath75 ) . constraining @xmath76 does not actually constrain the slope to steep , cdm - like values , but still allows values up to @xmath77 . over - plotted in fig . [ fig : gammartfits ] are the best - fit @xmath10 and @xmath9 values of the lsb rotation curves as determined by h04 . a distinction is made between the entire sample ( groups a , b and c ) , the `` cdm - compatible '' group a , and the sub - group discussed above where the best fit is apparently already cdm - compatible . the distribution of the points is not very different from group to group , though the @xmath78 subgroup tends to have slightly steeper slopes , as already shown in fig . [ fig : hayslope3 ] . regardless of how the total sample is divided up , the very large majority of the galaxies have slopes inconsistent with the values @xmath37 derived in the h04 simulations . note again that the slopes in simulations and observations are compared at _ identical radii _ , and that the @xmath10 and @xmath9 values were derived using the _ entirety of the rotation curves_. the extra constraints introduced by h04 are thus insufficient conditions for agreement with cdm , as constraining to @xmath76 does not automatically imply a mass - density slope consistent with cdm . unless one is prepared to allow for shallow slopes in cdm at these radii ( making it inconsistent with its own simulations ) an extra constraint on the slope is needed to ensure agreement with cdm adding this extra constraint to the four introduced by h04 is only necessary if these prior constraints are insufficient to make a unique distinction between cdm and non - cdm models . if , for example , rotation curves known to be inconsistent with cdm are not rejected as such , the prior constraints are clearly insufficient to address cdm - related questions . one can test the strength of these constraints by `` observing '' a sample of simulated iso haloes . this will show that the h04 constraints are not sufficient as they imply that the majority of these iso haloes , all with easily detectable shallow slopes , are consistent with cdm . iso haloes form of course the one model known to be certainly inconsistent with cdm ( if they were not , there would be no cusp / core controversy ) . the modeling procedure as described in @xcite [ hereafter db03 ] is used . the reader is referred to that paper for an extensive description of these models . the aim there was to estimate the impact of systematic effects on the observations . for that purpose many different rotation curves were simulated and `` observed '' after the addition of observational uncertainties due to e.g. resolution effects and random motions . db03 used the same datasets as the basis of their models as h04 use for their analysis , and the db03 procedure can be directly applied here to test the fitting method and make comparisons with the h04 results . the iso halo parameters are chosen in such a way that the resulting curves resemble minimum disk iso fits to the observed rotation curves . fig . 10 from @xcite shows that mass models of lsb galaxies have asymptotic velocities @xmath79 ) < 2.2 $ ] ( between @xmath80 and @xmath81 km s@xmath82 ) . central dark matter core densities vary between @xmath83 ) < 0 $ ] ( between @xmath84 and @xmath85 @xmath86 pc@xmath87 $ ] ) . the core radius @xmath88 follows from these two parameters , and has a range @xmath89 ) < 0.83 $ ] ( between @xmath90 and @xmath91 kpc ) . for the models a random @xmath92 and @xmath93 are chosen from the respective logarithmic intervals . note that these intervals encompass the entire observed parameter range . no account is taken of any correlations that might exist between these parameters , nor of any other uncertainties that might be present in the determination of halo parameters . this is reflected in the range of core - radii , which is somewhat larger than that found from observations ( typically @xmath94 kpc ) . the choice of parameter ranges is therefore liberal , but sufficient for our purposes : this is not an attempt to model individual galaxies but a _ reductio ad absurdum _ proof of principle . for each model halo an inclination , resolution and sampling interval are chosen randomly from the observed distributions described in db03 . error - bars are assigned to each sampled point , again modeled on the observed distribution of uncertainties , with a minimum value of 4 km s@xmath82 . the curve is then corrected for inclination with a corresponding increase in the size of the error - bars . to each data point a random velocity component between @xmath95 and @xmath96 km s@xmath82 is added . this is slightly simpler than the method used in db03 , but results in similarly sized random motions . lastly , as real rotation curves do not all extend out to the same radius , for each galaxy a random outer radius between 3 and 20 kpc is defined , out to which the galaxy is `` observed '' . figure [ fig : panel ] shows a small selection of the artificial iso curves . over - plotted is the eq . ( 1 ) fit using the best - fit constraints ( 1)-(3 ) , as given in sec . 1 . the reduced @xmath21 and the fit parameters are also given in the figure . figure [ fig : modelgamma ] shows the distribution of the best - fit @xmath10 parameters for 100 different realisations of the model . some @xmath97 per cent of the iso galaxies have @xmath18 . this number should be compared with h04 where it was found that @xmath98 per cent of the lsb rotation curves had @xmath18 . the @xmath10-distribution of lsb galaxies is at least in this respect consistent with that of iso haloes . [ fig : modelslopes ] shows a histogram of the slopes at @xmath63 kpc , both for the input iso models and the `` observed '' output models , the latter derived using the best - fit @xmath10 and @xmath9 parameters . both the input and output slopes show a clear peak at @xmath35 with a tail towards steeper slopes where the data are probing the edge of the dark matter core kpc and not in the center , we do not expect to find flat slopes for all iso haloes . the steepest iso slopes expected are nevertheless significantly less steep than the predicted cdm values . cf . the remarks in sect . 3 the output distribution has a less pronounced peak at @xmath35 , and a longer tail towards steeper slopes , but as the discussion in @xcite shows , this is fully consistent with the effects of the limited resolution used to `` observe '' the input distribution . the range in slopes found is slightly larger than that exhibited by real lsb galaxies shown in fig . [ fig : hayslope3 ] . this is entirely due to the larger range in core - radii used . fits of eq . ( 1 ) to a set of model iso rotation curves `` observed '' under the same conditions as the samples of lsb galaxies under investigation here are thus able to retain and retrieve the presence of shallow inner slopes . in order for the h04 criteria to make a valid discrimination between cdm and non - cdm models they thus need to be able to reject the majority of the iso models presented here . applying these criteria to the iso models and again fitting using eq . ( 1 ) , one finds that reducing the upper limit of @xmath99 to @xmath100 has only a modest effect on @xmath22 and almost the entire iso sample can still be fitted reasonably well with eq . ( 1 ) and constraints ( i)-(iii ) , as can be seen from the representative examples in fig . [ fig : panel ] . cdm constraint ( iv ) turns out to be the most restrictive one . it rejects about @xmath101 per cent of the simulated iso haloes with @xmath102 ( note that h04 show that a similar fraction of the lsb galaxies have incompatible densities as well ) . the bottom row of fig . [ fig : panel ] shows a few of the rejected models . these do not look any different from the cdm - compatible iso curves : condition ( iv ) merely removes some fraction of the iso haloes that do not have a cdm - compatible density , without any particular selection against specific values of @xmath10 , @xmath9 , @xmath8 or other relevant parameters . the most surprising outcome though is that of this sample of random iso haloes almost 60 per cent match the h04 cdm criteria . a comparison of the distribution of slopes in this cdm - compatible sample of haloes with the total distribution shows that there has been no selection against shallow slopes ( fig . [ fig : slopescomp3 ] ) . of the total sample of 100 observed iso models , @xmath103 per cent has an observed slope @xmath104 . once condition ( iv ) is applied , 57 haloes remain , of which 82 per cent have best - fit slopes @xmath105 . even when the slopes are derived using the constrained `` cdm compatible '' fit parameters , one still finds that 70 per cent has a shallow slope @xmath104 . as the range in core - radii used for the models is slightly larger than found observationally , the fractions mentioned here should be considered lower limits . regardless of how the slope is defined , shallow slopes are still profusely present ( and detectable ! ) in the models . a comparison with the input iso halo parameters shows that the rejected haloes are not abnormal in any way . in fig . [ fig : rho0 ] @xmath25 is plotted against the input iso halo central density @xmath93 . there is a clear relation , as might be expected : the average density of a halo is related to its central density . galaxies are plotted with different symbols depending on whether they meet the @xmath25 constraint or not . it is clear that the halo models that do not meet the cdm criteria are only guilty of having average densities that are too high or too low . they are not degenerate , nor do they suffer from e.g. larger error bars . a more finely - tuned choice of halo parameters could bring the rejection rate down significantly , but as the input ranges of @xmath93 and @xmath106 are based on observed distributions , it is more likely that real galaxies have a wider range of densities than predicted by cdm , and condition ( iv ) merely selects from a much broader spectrum of galaxy properties those galaxies that happen to lie in the cdm - compatible range . figure [ fig : chi2 ] compares the values of @xmath22 for the best - fit parameters and the cdm - compatible parameters . haloes that were rejected by the @xmath25 criterion are indicated in the right - hand side of the plot . following h04 one can divide the haloes in three categories : _ ( group a ) : _ rotation curves that are well fit by eq . ( 1 ) and for which a good fit with apparently cdm - compatible parameters can also be found , i.e. @xmath107 and @xmath108 . _ ( group b ) : _ rotation curves that are well fit by eq . ( 1 ) but for which no good fit with cdm - compatible parameters can be found , usually because of a discrepant value of @xmath25 ; these curves have @xmath109 and @xmath110 or undefined ( i.e. , with @xmath25 out of range ) . _ ( group c ) : _ rotation curves that are poorly fit by eq . ( 1 ) , i.e. @xmath111 and @xmath110 . the latter category only contains a few galaxies , as there are no large - scale asymmetries in the model rotation curves , and most rise smoothly and monotonically with radius . note the numerical agreement between the observational @xmath112 values presented in h04 and those of the iso models , showing that the additional uncertainties introduced in those models are a good representation of the true observational uncertainties . models in group a are , according to h04 , `` consistent with @xmath0cdm in terms of their inferred central densities and the shape of their rotation curves . '' one now has to add several caveats to this statement . firstly , even with @xmath113 and @xmath114 a steep slope is not guaranteed . the constraints used to define group a are not sufficient to filter out the shallow slopes that are still present and detectable in the iso models , even after the addition of observational uncertainties . secondly , as the inner slopes are well - defined and not necessarily steep , the shapes of the rotation curves are also not necessarily consistent with @xmath0cdm . constraints ( i)-(iii ) thus seem to have had hardly any effect in improving the agreement with cdm . the only restrictive constraint is condition ( iv ) , the @xmath25 criterion , but even this seems to have only selected models with a certain range in average density from amongst a much larger range of models . this is illustrated in fig . [ fig : bullock ] where a comparison is made between average @xmath25 densities of the iso models from group a and b with those expected in a @xmath0cdm universe . galaxies in groups a and b are indicated separately . in the h04 interpretation the iso halos from group a are consistent with cdm in terms of shape and average density , while group b is consistent in terms of shape but not density , even though groups a and b both consist of pure iso haloes , with easily detectable shallow slopes . the cdm constraints ( i)-(iv ) fail to take this into account and are thus insufficient to distinguish between cdm and non - cdm models . they can not exclude the possibility that lsb galaxies have shallow inner mass - density slopes , even at @xmath115 . additional constraints are needed to make a unique identification of galaxies compatible with cdm . now that it is established that constraints ( i)-(iv ) are not sufficient to uniquely identify cdm - compatible galaxies , the next step is to identify additional conditions that can . figures [ fig : courslope ] and [ fig : gammartfits ] show that the inner slope and therefore the shape of the rotation curve do not depend on @xmath10 alone , but that there is an equally strong dependence on @xmath9 . these figures also explain why h04 came to the conclusion that their four constraints were sufficient : implicit in their analysis is the assumption that the shape of the rotation curve depends on @xmath10 alone . while it is true that for @xmath116 kpc a smaller @xmath10 implies a steeper slope , this does not ensure a steep slope in an absolute sense . for @xmath76 , depending on the value of @xmath9 , slopes as shallow as @xmath117 are still possible . furthermore , it is shown above that the h04 method does not take these shallow slopes into account , and thus allows models with easily detectable and prominent shallow slopes purpose - built in to pass the cdm test with flying colours . in order to identify galaxies that are really compatible with cdm , an additional constraint is needed , in such a way that only @xmath118 combinations are allowed that yield a steep slope [ as well as pass conditions ( i)-(iv ) ] . in practice this means that @xmath69 parameter space needs to be searched for the minimum @xmath21 value that still results in a steep slope , one needs to take that parameter into account as well . this is done by stepping through values of @xmath8 at fixed @xmath10 and @xmath9 , and choosing the value that yields the lowest @xmath21 . therefore @xmath119 $ ] . ] . in order to compare directly with the simulations the slopes are again evaluated at @xmath63 kpc . at this radius the h04 simulations show slopes @xmath120 , and strictly speaking one ought to restrict the search of @xmath121 space to slopes @xmath122 . however , to take into account uncertainties in data and simulations , and give the cdm models as much leeway as possible , a more liberal range of @xmath123 will be used . [ fig : slopeonly ] shows that this is the steepest slope we can expect at this radius from an ultra - compact iso halo with @xmath124 kpc , and this value thus cleanly separates the iso and cdm domains . at this point a short digression on the significance of @xmath112 is useful . in testing the goodness - of - fit of a model , the reduced @xmath21 , that is , the total @xmath21 divided by the degrees of freedom of the data , is often used . a value of @xmath125 is then taken as an indication that the model is a good description of the data . the reduced @xmath21 is an appropriate statistic when comparing different models to a single set of data , or data with a similar number of degrees of freedom . an example would be the testing of an nfw model versus an iso model for a single rotation curve , or a set of curves with a similar number of data points . h04 determine for each lsb rotation curve the reduced @xmath21 for the best - fitting @xmath118 model , as well as the minimum @xmath112 value using their cdm constraints . they argue that for most of the galaxies the cdm constraints only cause a small increase in the _ reduced _ @xmath21 , and that therefore the cdm - constrained models are not much worse than the best - fit models . this is not necessarily so . remember that the reduced @xmath21 is the @xmath21 divided by the degrees of freedom of the data ( the number of data points minus the number of parameters in the model ) . it is easy to see that in order to cause the same change in _ reduced _ @xmath21 the change in _ absolute _ @xmath21 must be much larger for a rotation curve with a large number of data points than for a rotation curve with a small number of data points . in other words , a small change in @xmath22 can be very significant ( that is , unlikely to be consistent with the uncertainties in the data points ) for a curve with a large number of data points , while it can be insignificant ( entirely consistent with the uncertainties ) for a curve with only a small number of data points . the observed sample we are dealing with here contains a large number of rotation curves , with a wide range in the number of data points , varying from @xmath126 to over 350 . the relevant parameter when judging whether a change in fit parameters yields a model that is still consistent with the data is thus _ not _ the reduced @xmath21 , but the _ change in absolute @xmath21_. the confidence level of any parameter combination is given by the corresponding change in @xmath21 with respect to the minimum value . for a model with three free parameters @xmath127 as used here , the @xmath128 confidence intervals for any fit are given by @xmath129 , respectively . using these criteria , a rotation curve can be defined to be consistent with cdm if the best - fit values of @xmath10 and @xmath9 lie within @xmath130 of the closest @xmath118 combination with a @xmath131 slope . in order for the fit to give any meaningful constraints , one should furthermore require that the area enclosed by the @xmath130 confidence level is a small fraction of the total area of parameter space investigated is not crucial . we have tested the analysis with a @xmath132 level as well . though the number of badly constrained fits increases , the relative proportions of cdm consistent _ vs _ inconsistent fits does not change appreciably . ] . if the latter is not the case , then any conclusions regarding agreement or disagreement with cdm are only very weak at best . here the group a rotation curves , defined in h04 to be consistent with their definition of cdm , are investigated again using the new constraints . groups b and c are already inconsistent with cdm using the more relaxed h04 constraints , and will therefore not be investigated further . for each rotation curve from group a the minimum @xmath21 was found using eq . ( 1 ) , and the change in @xmath21 necessary to be consistent with a slope @xmath123 was determined . the @xmath133-confidence interval corresponding to this @xmath134 was also calculated , as well as the ratio @xmath135 of the area enclosed by the @xmath130 contour and the total area of @xmath121 parameter space searched @xmath136 kpc ) . as described above , the ratio @xmath135 is used to determine whether a solution is well - constrained or not . small values of @xmath135 indicate that the @xmath130 area is only a small fraction of total parameter space , and therefore yield significant solutions . large values of @xmath135 indicate that the fit does not constrain models in a significant way . the results are given in table [ table : fit ] , where the galaxies are sorted in order of the number of data - points in each individual rotation curve . [ fig : chi2surface ] shows the @xmath128 confidence contours for a number of representative galaxies as described below . from studying the plots of the @xmath21 contours for all galaxies , it was found that a value @xmath137 is a reasonable discriminator between well - constrained and badly constrained fits , but the precise value of @xmath135 is not critical ( see fig . [ fig : chi2lim ] for some examples of @xmath138 fits ) . the galaxies in table [ table : fit ] fall in four distinct categories : 1 . _ well - constrained solutions , consistent with cdm : _ objects with best fits within @xmath130 of a steep slope @xmath123 parameter combination , and with a well - constrained @xmath139 solution . an example is ngc3274 , as shown in fig . [ fig : chi2surface ] . this category consists of 3 out of the 36 galaxies in group a. 2 . _ well - constrained solutions , inconsistent with cdm : _ objects with best fits more than @xmath130 away from a steep slope @xmath123 parameter combination , and with a well - constrained @xmath139 solution . an example is ugc 11819 , as shown in fig . [ fig : chi2surface ] . this category contains 17 out of 36 galaxies . badly constrained solutions , consistent with cdm : _ objects with best fits within @xmath130 of a steep slope @xmath123 parameter combination , but with a badly constrained @xmath140 solution . an example is f563 - 1 , as shown in fig . [ fig : chi2surface ] . this category contains 12 out of 36 galaxies . badly constrained solutions , inconsistent with cdm : _ objects with best fits more than @xmath130 away from a steep slope @xmath123 parameter combination , and with a badly constrained @xmath140 solution . an example is ddo 185 , as shown in fig . [ fig : chi2surface ] . this category contains 4 out of 36 galaxies . it is clear that only galaxies from the first two categories , i.e. with @xmath139 , can help constrain halo models . a glance at table [ table : fit ] shows that most of the solutions that are consistent with cdm are actually badly constrained . this is investigated further in fig . [ fig : resolve ] . here @xmath135 is plotted against the number of data points in each rotation curve . a distinction is made between rotation curves that are consistent with cdm and those that are inconsistent with cdm using the constraints introduced here . table [ table : fit ] and fig . [ fig : resolve ] lead to several conclusions . firstly , of the 15 galaxies that are consistent with cdm , 13 have less than 30 data points . furthermore , of these 13 galaxies , only 1 has @xmath139 ( i.e. is well - constrained ) . looking at the 21 galaxies that are inconsistent with cdm , only 10 have less than 30 data points . of these 10 galaxies , all except one have @xmath139 . rotation curves with more than 30 data points tend to be better constrained on average , as well as have a tendency to be inconsistent with cdm . it is also interesting to note that even amongst galaxies with only a limited number of data points , the most constrained solutions ( small @xmath135 ) tend to be those inconsistent with cdm . the majority of solutions @xmath141 per cent ) that are consistent with cdm have only a small number of data points _ and _ are ill - constrained . these ill - constrained solutions therefore do not decide between any model one way or the other , and statements that these galaxies are consistent with cdm are thus not very strong . of the well - constrained solutions the large majority is inconsistent with cdm . furthermore , of the 13 galaxies with more than 30 data points only 2 are consistent with cdm . these two galaxies ( ngc 3274 and ngc 4455 ) also happen to have the smallest optical scale - lengths and highest surface brightnesses from all galaxies with available photometry in the @xcite , @xcite and @xcite samples : both galaxies have @xmath142 mag arcsec@xmath143 . these galaxies are therefore unlikely to be dominated by dark matter in their inner parts and it is very likely that the steep slopes in these two galaxies simply reflect the stellar mass distribution ( see also the discussion on ngc 3274 in sec . 9.2.2 of @xcite ) . amongst the rotation curves with less than 30 points there is only one galaxy with @xmath144 that seems consistent with cdm . this is ugc 731 , a bona - fide lsb dwarf , and is the best and only candidate in the entire observed sample for a cdm - consistent , dark - matter - dominated galaxy with a well - constrained solution . the level of agreement is not perfect though , ugc 731 is consistent with a @xmath145 slope at the @xmath146 confidence level , with the confidence level decreasing to @xmath147 for an @xmath4 slope , which is still only the upper limit of the range of slopes expected from the simulations . once the steep slope constraint is added to the h04 constraints there are thus only 3 galaxies left that have a well - constrained fit consistent with a steep slope and @xmath0cdm . of these 3 galaxies , 2 are of very high surface brightness and are likely to be dominated by stars in the inner parts , leaving only one possible candidate with a well - constrained cdm - compatible fit . the large majority of the well - constrained fits strongly prefer shallow slopes . the 2@xmath133 limit used here already indicates that for many galaxies it is difficult to adjust the fit parameters to get a cdm - compatible fit . restrictive as this criterion is , it really only gives an upper limit to the true probability . while one has the freedom to adjust the parameters of an individual fit to test the likelihood of an hypothesis for one particular rotation curve , this becomes problematic when dealing with many fits . if every fit is individually adjusted to fit a particular model , does that still constitute a proper test ? an analogue can perhaps be found in some of the first supernova results that indicated the existence of a cosmological constant ( e.g.@xcite ) . the majority of the supernovae in that early work were each individually consistent within @xmath130 with the _ _ non-__existence of @xmath0 . even though the data showed a systematic trend , each measurement could have been corrected individually to fit the no-@xmath0 model . however , the joint result of these small inconsistencies has formed the basis for the discovery of the effects of @xmath0 on the expansion of the universe . while it would be presumptuous to compare those results with the current analysis , it does show that it is important to pay attention to the global conclusion that is forced upon us by the data . one thing that is clear is that galaxies that are significantly consistent with @xmath0cdm in terms of their inferred central densities and the shapes of their rotation curves seem to be rare indeed . i have investigated the claim by h04 that the majority of rotation curves of lsb galaxies are not inconsistent with cdm . i have shown that the method used is unable to distinguish between shallow and steep slopes , even when these are obviously present and easily detectable . i illustrate this by showing that with the h04 method one would infer even pseudo - isothermal , core - dominated haloes to be consistent with cdm . an extra constraint based on the mass - density slope is thus needed to make more definitive statements on any ( dis)agreement with cdm . using this extra criterion i show that only a quarter of the lsb galaxies can be said to be consistent with cdm , as opposed to the three quarters found by h04 . the majority of these cdm - consistent lsb galaxies do however have fit parameters that are so ill - constrained , that they are consistent with virtually anything . restricting the analysis to galaxies with well - constrained solutions , i find that only 3 out of the 20 well - constrained galaxies are possibly consistent with cdm ; two of these are high surface brightness dwarf galaxies that are likely to be dominated by stars . in summary , a comparison of the circular velocity profiles of cdm haloes with rotation curves of lsb galaxies indicates that the shapes and inferred central densities of most lsb galaxies are mostly _ inconsistent _ with those of simulated haloes within the limitations imposed by observational error . alam , s. m. k. , bullock , j. s. , weinberg , d. h. 2002 , , 572 , 34 binney , j. 2004 , iau symposium , 220 , 3 bullock , j. s. , kolatt , t. s. , sigad , y. , somerville , r. s. , kravtsov , a. v. , klypin , a. a. , primack , j. r. , & dekel , a. 2001 , , 321 , 559 courteau , s. 1997 , , 114 , 2402 de blok , w. j. g. 2004 , iau symposium , 220 , 69 de blok , w. j. g. , bosma , a. 2002 , , 385 , 816 de blok , w. j. g. , mcgaugh , s. s. 1997 , , 290 , 533 de blok , w.j.g . , mcgaugh , s.s . , rubin , v.c . 2001 , , 122 , 2396 de blok , w. j. g. , mcgaugh , s. s. , bosma , a. , & rubin , v. c. 2001 , , 552 , l23 de blok , w. j. g. , bosma , a. , & mcgaugh , s. 2003 , , 340 , 657 fukushige , t. , kawai , a. , & makino , j. 2004 , , 606 , 625 ghigna , s. , moore , b. , governato , f. , lake , g. , quinn , t. , & stadel , j. 2000 , , 544 , 616 hayashi , e. , navarro , j.f . , power , c. , jenkins , a. , frenk , c.s . , white , s.d.m . , springel , v. , stadel , j. , quinn , t.r . 2004 , , 355 , 794 jing , y. p. , mo , h. j. , borner , g. , & fang , l. z. 1995 , , 276 , 417 klypin , a. , kravtsov , a. v. , bullock , j. s. , & primack , j. r. 2001 , , 554 , 903 mcgaugh , s. s. , rubin , v. c. , & de blok , w. j. g. 2001 , , 122 , 2381 moore , b. , governato , f. , quinn , t. , stadel , j. , & lake , g. 1998 , , 499 , l5 moore , b. , ghigna , s. , governato , f. , lake , g. , quinn , t. , stadel , j. , & tozzi , p. 1999 , , 524 , l19 navarro , j.f . , frenk , c.s . , white , s.d.m . 1996 , , 462 , 563 navarro , j.f . , frenk , c.s . , white , s.d.m . 1997 , , 490 , 493 navarro , j. f. , et al . 2004 , , 349 , 1039 power , c. , navarro , j. f. , jenkins , a. , frenk , c. s. , white , s. d. m. , springel , v. , stadel , j. , & quinn , t. 2003 , , 338 , 14 riess , a. g. , et al . 1998 , , 116 , 1009 swaters , r. a. , madore , b. f. , van den bosch , f. c. , & balcells , m. 2003 , , 583 , 732 wechsler , r. h. , bullock , j. s. , primack , j. r. , kravtsov , a. v. , & dekel , a. 2002 , , 568 , 52 f563 - 1 & 10 & 1.92 & 0.67 & 0.54 & m01 + eso 4250180 & 11 & 24.94 & 0.25 & 4.31 & m01 + eso 4880490 & 12 & 1.07 & 0.27 & 0.28 & m01 + eso 3020120 & 12 & 1.72 & 0.62 & 0.48 & m01 + ddo 189 & 14 & 0.75 & 0.79 & 0.18 & db02 + eso 1870510 & 15 & 0.02 & 0.45 & 0.0011 & m01 + ugc 4115 & 15 & 0.36 & 0.87 & 0.064 & m01 + f568 - 3 & 16 & 43.27 & 0.13 & 5.99 & m01 + f583 - 1 & 17 & 15.12 & 0.43 & 3.14 & m01 + f571 - 8 & 19 & 93.60 & 0.10 & @xmath148 & m01 + f563 - 1 & 19 & 0.02 & 0.56 & 0.0011 & db02 + ugc 11583 & 19 & 4.31 & 0.57 & 1.20 & m01 + eso 1200211 & 19 & 2.09 & 0.45 & 0.59 & m01 + f583 - 4 & 19 & 19.67 & 0.13 & 3.72 & m01 + eso 2060140 & 20 & 9.57 & 0.06 & 2.28 & m01 + ugc 11819 & 23 & 181.61 & 0.07 & @xmath148 & m01 + f730-v1 & 24 & 18.66 & 0.04 & 3.60 & m01 + ugc 4173 & 24 & 0.00 & 0.62 & 0.00 & db02 + ugc 11616 & 26 & 105.32 & 0.05 & @xmath148 & m01 + ddo 47 & 26 & 0.05 & 0.91 & 0.0040 & db02 + ugc 731 & 28 & 4.95 & 0.11 & 1.35 & db02 + ugc 11454 & 28 & 218.96 & 0.05 & @xmath148 & m01 + ugc 5005 & 28 & 1.47 & 0.87 & 0.40 & db02 + ugc 1230 & 34 & 20.27 & 0.31 & 3.79 & db02 + ugc 628 & 37 & 8.92 & 0.05 & 2.17 & db02 + ddo 185 & 42 & 19.23 & 0.44 & 3.67 & db02 + ugc 10310 & 49 & 32.78 & 0.20 & 5.09 & db02 + ugc 3371 & 57 & 11.68 & 0.57 & 2.63 & db02 + ngc 3274 & 66 & 0.00 & 0.01 & 0.00 & db02 + ugc 3137 & 72 & 24.17 & 0.15 & 4.23 & db02 + ngc 4455 & 80 & 1.63 & 0.10 & 0.45 & db02 + ugc 4325 & 84 & 76.41 & 0.24 & @xmath148 & db02 + ddo 52 & 86 & 41.8 & 0.02 & 5.87 & db02 + ngc 100 & 92 & 21.16 & 0.03 & 3.90 & db02 + ngc 5023 & 149 & 95.96 & 0.01 & @xmath148 & db02 + ngc 4395 & 355 & 23.51 & 0.01 & 4.16 &

one of the major long - term goals of astronomy is to place the solar system within a greater context , finding for example whether habitable planets like the earth are typical , and whether these earth analogues have conditions suitable for alien life . such goals are ambitious and many obstacles must be overcome for them to come to fruition . for the particular case of directly imaging earth - like planets around other stars , a major unknown is the level of photon noise from the dust that resides in the target system itself , specifically dust located in the `` habitable zone '' ( hz , e.g. * ? ? ? * ) . such dust populations are generically referred to as `` exo - zodi '' by analogy with the solar system s zodiacal cloud , though they may have different origins . to be clear , we consider the dust populations seen around 10 - 25% of nearby stars with near - infrared ( ir ) interferometry to be a largely unrelated phenomenon , based on their small grain size and hot temperatures , and lack of correlation with both mid - ir detections of hz dust ( mennesson et al . , apj , in press ) and far - ir detections of cool dust . if hz dust levels an order of magnitude or so greater than the solar system level are typical , this noise source could seriously hinder an earth analogue - imaging and characterisation mission ( which we will refer to as earth - imaging for brevity , e.g. * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) . currently , the brightness distribution ( which we also call the luminosity function ) of exo - zodi is largely unknown . limits have been set by the keck interferometer nuller ( kin , * ? ? ? * ; * ? ? ? * mennesson et al . , apj , in press ) , but the distribution has only been characterised at levels 3 - 4 orders of magnitude above the solar system by photometric methods @xcite . enter the large binocular telescope interferometer ( lbti ) , a mid - ir instrument specifically designed to characterise exo - zodi at dust levels just a few times greater than the solar system s zodiacal cloud . for a brief description of the instrument see defrre et al . ( submitted ) . by observing of order 50 nearby stars at this level of sensitivity ( weinberger et al . , submitted ) , the lbti will identify specific targets with low dust levels that are suitable for future earth - imaging , and moreover will characterise the exo - zodi luminosity function with sufficient detail to provide new information on the dust origin and evolution . this latter point is important because the lbti will observe a limited number of stars and can not access the entire sky , and population - level information such as correlations with other system properties will be needed to assess the suitability of targets that are not observed with the lbti . of course the results will also be scientifically valuable and interesting , particularly when combined with existing and future observations ( e.g. defrre et al . submitted ) . mid and far - ir imaging using telescopes such as _ spitzer , herschel _ , and jwst will provide important information on the existence , location , and structure of cooler dust belts that lie outside the hz . interferometric observations with longer baselines ( e.g. vlti / matisse ) and/or at different wavelengths ( e.g. chara / fluor , vlti / pionier ) will provide information that complements the lbti to help build a complete picture of the inner regions of individual systems . the goal of this paper is to outline a modelling framework for interpreting lbti observations . such models are needed because lbti observations yield limited information on the spatial structure of any disk that is detected . it is therefore important that the model used to interpret the observations is thoroughly characterised so that the constraints placed on its various parameters can be understood in terms of their implications for the disk structure and surface brightness . the primary goal is to use the model to make a useful statement about the level of dust in a system , or the limits on undetected dust , in the habitable zone where earth - imaging will be attempted in the future . in what follows , we outline a parameterised dust model that can be used to approximate the solar system s zodiacal dust cloud . we show how this model can be used to derive the distribution of dust levels given an lbti observation of thermal dust emission . we then determine the corresponding implications for scattered light levels at visible wavelengths . we finally show the expected levels of dust that could be detected by the hunt for observable signatures of terrestrial planetary systems ( hosts ) survey and the corresponding scattered light levels . for a detailed description of the lbti sample we refer the reader to weinberger et al . ( submitted ) , and for discussion of the instrument and the first scientific results to defrre et al . ( submitted ) . the primary goal of our exo - zodiacal cloud model is to allow easy comparison of the results among survey stars , which span a range of distances and luminosities . in general we expect that detections will be near the sensitivity limits , and that there will be little information about any warm dust detected ( or not detected ) other than from lbti . therefore , while warm dust may arise from various different processes , for example _ in situ _ asteroid - belt like evolution , stochastic collisions , delivery of comets from elsewhere , or some combination of all three ( e.g. * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) , we will have little power to distinguish among them . some scenarios could result in clumpy non - axisymmetric structures , but there again is little hope of distinguishing these from axisymmetric structures in all but the dustiest systems where lbti detections will be relatively easy and at high signal to noise ratios . therefore , the models considered here are axisymmetric . for specific stars , the stellar luminosity @xmath1 ( in solar units ) is easily derived by fitting stellar atmosphere models to photometry , and the flux density @xmath2 at any wavelength can be inferred from these models . given a distance @xmath3 , the stellar radius can be estimated from the same model , though in some cases it may have been directly observed . in any case , the contribution of the resolved stellar disk to an lbti observation will generally be negligible . here we use the stellar fluxes for hosts survey stars derived by weinberger et al . ( submitted ) . lbti observations are sensitive to the surface brightness distribution @xmath4 in the n band at 11@xmath5 m ( the bandpass spans 9.81 - 12.41 @xmath5 m ) . our model can be considered a means of parameterising this surface brightness distribution using parameters that have some physical relevance . unless stated otherwise , in what follows our calculations are made at 11 @xmath5 m . the surface brightness profile of the disk , if viewed face - on , is modelled as @xmath6 here @xmath7 is the temperature of disk material that behaves like a black body @xmath8 where the distance to the star , @xmath9 , is in au . the numerical factor in eq . [ eq : sb ] is a conversion ( 1au@xmath10/1pc@xmath10 ) so that the surface brightness is in units of jy arcsec@xmath11 . the parameter @xmath12 is representative of the disk s face - on surface density of cross - sectional area ( in au@xmath10/au@xmath10 ) , so is analogous to optical depth and is assumed to have a power law distribution @xmath13 between radii @xmath14 and @xmath15 ( both in au ) . the normalisation @xmath16 is to be set at some @xmath17 ( in au ) such that the surface density is in units of zodis @xmath18 ( see section [ ss : zodi ] ) . for generality we include an inclination @xmath19 , which is half the total opening angle of the disk , and represents the maximum inclination of the parent bodies ( whose nodes would be assumed to be randomised ) . here we assume that our disks have negligible opening angles so @xmath19 is not used , but that @xmath19 is always sufficiently large that the disk is radially optically thin . if all of the disk particles behave like black bodies , the parameter @xmath12 can be interpreted as the disk s surface density of cross - sectional area . its interpretation becomes more complicated for disks with more realistic particle optical properties because the observed dust can be several times hotter than the black body temperature , and not all radiation incident on a particle is absorbed and thermally re - emitted . however , as the surface brightness is typically assumed to exhibit a power law dependence across the relevant radius range , we have not lost any generality by expressing the disk surface brightness in this way , even if the disk s temperature profile differs from that of a black body . these issues will be discussed in more detail in sections [ ss : other ] and [ ss : scat ] . the final parameters , which specify the direction to the observer relative to the disk , are the disk inclination @xmath20 and the position angle @xmath21 . the inclination is defined such that @xmath22 is face - on and the position angle is measured anticlockwise ( i.e. east ) from north . the range of inclinations is 0 to @xmath23 and the range of position angles 0 to @xmath24 , ) means that if the disks are not spatially resolved we in fact only need to consider angles between 0 and 90@xmath25 . we retain the larger range because we may wish to set @xmath21 based on the position angle of the star or an outer disk component . ] with these ranges set because we can not distinguish between the near and far side of a disk . that is , the faint disks considered here are optically thin viewed from any direction , and in thermal emission look the same mirrored in the sky plane or for 180@xmath25 rotations . the model parameters and their meanings are summarised in table [ tab : params ] . .model parameters [ cols="<,<,<,<",options="header " , ] ` zodipic ` is an implementation of the @xcite zodiacal cloud model . this parametric model has many parameters , whose values were derived via fitting to cobe / dirbe observations . ` zodipic ` produces two - dimensional images of the solar system s zodiacal dust cloud as might be seen by an external observer , and has been used as a reference model for calculating the limits set by kin observations @xcite . though it is a complex model and can include dust components such as the earth s resonant ring and trailing blob , only the main smooth axisymmetric component is normally used . because this main component is simply a radial power law , with an approximately gaussian vertical density distribution , it is possible to reproduce the ` zodipic ` surface brightness using our parameterised model . the vertical dust distribution is the main difference between the @xcite model and ours . the @xcite model assumes that the radial and vertical components of the cloud are separable @xmath26 where @xmath27 and @xmath28 and @xmath29 is the height above the disk midplane in au . the best fit values for the cobe data were @xmath30 , @xmath31 , @xmath32 , @xmath33 , and @xmath34 . the units of the normalisation are au@xmath10/au@xmath35 ( i.e. a volume density of surface area ) , and the integrated length of the vertical term is not constant with radius , so their exponent @xmath36 is not the same as our exponent @xmath37 . numerically integrating equation ( [ eq : zodipic2 ] ) over @xmath38 yields 0.63 , so the surface density at 1 au is @xmath39 ( au@xmath10/au@xmath10 ) . assuming for mathematical simplicity that @xmath40 , the integral of the exponential term in equation ( [ eq : zodipic ] ) is @xmath41 for both cases of @xmath42 , and hence @xmath43 and @xmath44 . therefore , the ` zodipic ` model can be reasonably expressed in terms of our model with @xmath45 au @xmath46 , and @xmath47 . there remain a few small differences in the surface brightness of our model relative to ` zodipic ` , as shown in the right panel of fig . [ fig : mod ] . the main difference is that the ` zodipic ` model is slightly brighter at larger radii because the radial temperature profile is flatter . setting the ` zodipic ` temperature profile to equal our black body prescription leads to nearly indistinguishable face - on surface brightness profiles . some minor differences arise because the @xcite model defines radius as @xmath48 , whereas our narrow opening angle means that our model is effectively cylindrical . here we introduce a reference disk model , the parameters of which are summarised in table [ tab : params ] and based on the @xcite solar system model described above . the model extends from 0.034 - 10 au with a power - law index @xmath49 , and is scaled to @xmath50 at @xmath45 au ( the location where @xmath51 k ) . the inner edge is set by a black body temperature of 1500k , approximately the sublimation temperature of silicates , though we show in section [ ss : other ] that this choice is irrelevant for our model as long as the inner disk edge is inside the iwa of 40 mas . the outer edge of 10 au is chosen to be sufficiently large that it does not affect the model at the lbti wavelength of 11 @xmath5 m . as noted above we treat the disk as two dimensional so the opening angle is small . our reference model is based on the solar system so the star is sun - like , but as described in section [ ss : zodi ] the reference parameters are later varied to model disks around different stars . figure [ fig : mod ] shows the distribution of temperature , surface density and surface brightness at 11@xmath5 m in this reference model , where we have assumed a distance of 10 pc . survey targets are chosen so that the lbti is sensitive to dust in their habitable zone ( weinberger et al . , submitted ) , so the dotted lines show the range of radii implied by a temperature range of 210 - 320k , a rough indication of the habitable zone location for the sun ( e.g. * ? ? ? the right panel shows that the surface brightness of our model closely matches ` zodipic ` , and as noted above this difference only arises due to different assumptions about the radial temperature profile . therefore , for the same assumptions and aside from the differences in scale height our model yields the same results as ` zodipic ` for a sun - like star . the disk normalisation ( i.e. brightness ) is set by the parameter @xmath16 , which is the disk surface density at @xmath17 . the goal here is to set what @xmath52 and @xmath17 are so that they scale sensibly with stellar spectral type , and so that our disk model is somehow sensibly related to zodi units , and therefore the solar system dust level . these parameters , along with @xmath14 , @xmath15 , @xmath37 , and @xmath19 define our reference disk model . in making this definition , it is important to remember the ultimate goal of the lbti , which is to constrain the level of dust in the habitable zone around nearby stars . that is , disk surface brightness is the most important measure of the impact of exo - zodi on future earth - imaging missions . a desirable definition might be one of `` constant hindrance '' for an earth - imaging mission , where @xmath16 is set such that a @xmath53 disk impacts all observations in search of exo - earths at the same level . however , for this imaging any exo - earths will be unresolved , whereas the exo - zodi that accompany them will be resolved , so the same physical disk has the same surface brightness regardless of distance , but the brightness of the unresolved planet decreases with distance . therefore , any zodi definition that attempts to provide constant hindrance will be distance dependent , which is an undesirable property since it is unrelated to the physical disk structure . to set @xmath16 therefore requires a choice between using either the total disk brightness , or the surface brightness at @xmath17 . both approaches have their merits . for example , using the total disk brightness allows comparison with photometric observations and therefore easy construction of the warm dust luminosity function ( e.g. * ? ? ? * ) . however , as we show below in fig . [ fig : fnutvsr ] , lbti observations are not necessarily sensitive to total disk brightness ( e.g. when a significant fraction of the disk emission lies inside the first transmission peak ) . we therefore prefer a zodi definition linked to the disk surface density in the habitable zone . for the spectral type scaling , it then makes sense for @xmath17 to scale with @xmath54 , so that the surface brightness expressed by @xmath16 corresponds to the radial distance where the equilibrium temperature is the same as at earth ( i.e. an `` earth - equivalent '' distance ) , therefore @xmath55 au . scaling @xmath14 and @xmath15 by @xmath56 is also appropriate , since this results in a common temperature at the inner and outer edges of the disk , and maintains the disk inner edge at the sublimation radius . the disk edges are of relatively little importance here because as we show below the lbti is relatively insensitive to their location . with this definition , @xmath16 is fixed at the surface density of the solar system s zodiacal cloud at 1 au , but the location where this surface density applies depends on the stellar luminosity . this definition is different to ` zodipic ` , which fixes @xmath17 at 1 au for all stars and fixes @xmath14 at the sublimation radius . because the dust temperature at @xmath17 is the same for all spectral types , the thermal surface brightness at @xmath17 ( i.e. in the habitable zone ) for a 1 zodi disk is also the same . because it is based on surface brightness , this zodi definition is well suited for use as an important metric for future earth - imaging missions . an lbti observation is not necessarily sensitive to the radial extent of a disk , so the derived zodi level depends on what is assumed for @xmath14 and @xmath15 , particularly if the disk is relatively narrow . for example , below we consider a `` worst case '' scenario where the dust emission only originates in the habitable zone , in which case the derived dust surface density ( and hence @xmath18 ) is higher for a given null depth . to avoid confusion we recommend that zodi levels primarily use our reference model to allow consistent comparisons . the original concept for nulling interferometry was put forward by @xcite , and for a reasonably non - technical description of the instrument we refer the reader to defrre et al . ( submitted ) . basically speaking , the lbti combines the beams from the two lbt mirrors , with one beam having a half - wavelength phase shift . light in the viewing direction , and along lines perpendicular to the mirror baseline vector , is therefore suppressed . light from off axis sources at odd multiples of the angular distance @xmath57 is transmitted ( @xmath58 is the baseline length ) . similarly , light from sources at even multiples of this angular distance is suppressed . the transmission ( or `` fringe '' ) pattern is therefore a @xmath59 function parallel to the baseline vector , and is constant perpendicular to this vector . a model disk , the transmission pattern , and the transmitted disk image are illustrated in fig . [ fig : trx ] . this illustration is for an object at transit ; at other times the fringe pattern is not aligned with north . the key measurement from an lbti observation is the ratio of the transmitted flux to the total photometric flux . because the transmission pattern is designed to suppress light from the star , leaving light from a much fainter disk , this ratio is approximately the ratio of the transmitted disk flux to the stellar flux . deriving the transmitted disk flux is not straightforward however , as some on - axis ( i.e. stellar ) flux is always transmitted because the instrument is not perfect and the star has a finite angular size . removal of these effects is an integral part of the data analysis ( see defrre et al , submitted ) , so the quantity derived from a given observation is the `` calibrated null depth '' or `` source null depth '' . in what follows we generally refer to this measurement and the same quantity derived from our models as simply the `` null depth '' . because the total flux is almost exactly the stellar flux and dominates over any disk emission , we therefore compute the null depth for our models as the transmission for a starless model divided by the stellar flux density . our disk model is radially extended , but to illustrate how surface brightness is transmitted through the lbti transmission pattern as a function of angular scale we first consider a series of discrete annuli of angular radii @xmath60 . we later combine these annuli to model the transmission for our disk models . the angular distance to the first transmission peak is @xmath61 ; for the lbti @xmath62 @xmath5 m and @xmath63 m , so the distance is 79 mas . the transmission function perpendicular to the fringes @xmath64 is @xmath65)$ ] , and is shown as the grey line in figure [ fig : trvsr ] , where the x axis is the distance from the annulus center in units of the distance to the first transmission peak ( i.e. @xmath66 ) . a common definition for the inner working angle of an instrument is where the sensitivity first reaches half of the peak value , which for lbti is therefore at @xmath67 mas . ) . the grey line is the projected transmission function parallel to the lbti baseline , the blue dashed line is the fraction of emission for face - on ( @xmath22 ) annuli after summing around the azimuthal angle . dots show a population of annuli with random inclinations and position angles , and the red line shows the average transmission of these dots.,scaledwidth=50.0% ] to calculate the transmission for a point at some azimuth around the annulus requires finding the sky - plane component of this vector point ( relative to the star ) that is perpendicular to the fringe pattern . this component can be calculated using three rotations , and is @xmath68 where @xmath69 is the angle from the sky plane around the annulus to the point of interest , and @xmath70 is the position angle of the annulus relative to the lbti fringe pattern . the angle @xmath70 varies with hour angle , and because the lbti baseline is always perpendicular to the local vertical ( i.e. a great circle through the target and zenith ) , is equal to @xmath21 for an object at transit . the transmission from a point in the annulus is then \right)/2\ ] ] where @xmath71 are the sky offsets from the null center and @xmath72 the corresponding spatial frequencies ( i.e. @xmath73 ) . lbti transmission is simpler to compute because the length of the sky - projected baseline is always the same due to the common mount for the mirrors . ] @xmath74 ) \ , .\ ] ] for a face - on ( @xmath22 ) annulus at radius @xmath75 , the transmission is clearly a function of azimuth around the annulus . averaging around an annulus yields the total transmission for that annulus , and repeating this calculation for annuli of different angular sizes gives the blue dashed line in figure [ fig : trvsr ] . this exercise is finally repeated for a large number of annuli with random orientations , so that @xmath76 is distributed evenly from 0 to 1 , and @xmath70 is evenly distributed from 0 to 180@xmath25 . again , the transmission around each annulus is azimuthally averaged , which results in the dots shown in figure [ fig : trvsr ] . averaging these points yields the average transmission as a function of annulus radius for a population of disks with random orientations , shown as the red line . as can be surmised from the decreasing amplitude , the average transmission tends to 0.5 at large separations ( i.e. the average of @xmath59 is 0.5 ) . the minimum separation at which this transmission is achieved is @xmath77 , twice the inner working angle . the origin of the dot distribution can be understood by considering how annuli of different orientations are transmitted through a given transmission peak . for example , the upper envelope of dots at about 70% transmission are all transmitted through the first transmission peak , and those at higher @xmath66 are nearer to edge - on with position angles closer to @xmath78 ( i.e. perpendicular to the baseline and parallel to the fringes ) . this effect is relatively common because the average inclination is about 60@xmath25 ( i.e. biased towards edge - on ) . by comparing the phase of the grey line with the red line , it is clear that the peak average transmission is actually about a quarter of the way beyond a transmission peak , and that the peak transmission for face - on annuli lies somewhere in between . the phase shift of the face - on transmission can be understood by realising that an annulus with a radius @xmath75 that is slightly greater than @xmath77 has more emission in the peak transmission region than an annulus with @xmath79 . the average transmission for random orientations is phase shifted slightly further because most disks are inclined , and therefore on average appear somewhat smaller on the sky than they actually are . . flux is lost at all radii due to the transmission pattern , but most is lost from inside the habitable zone . the total and transmitted disk fluxes are 0.11 and 0.031 mjy.,scaledwidth=50.0% ] using the reference model of section [ sss : ref ] , fig . [ fig : fnutvsr ] shows the total and transmitted disk flux as a function of radius for a face - on geometry and a distance of 10 pc . that is , the figure is a histogram showing where the total and transmitted flux originates , so the solid line is an azimuthally summed radial profile created from the right panel of fig . [ fig : trx ] . much of the total disk emission originates from the inner regions , and since the disk inner edge is well inside the first transmission peak significant flux from these inner regions is not transmitted . some flux is also lost at larger radii , and in total only about 30% of the disk flux is transmitted . overall therefore , the radial distance over which the disk emits strongly in the mid - ir and can be detected is not particularly large ; emission is strongly reduced inside the inner working angle , and the faint wien side of cooler emission means that the surface brightness drops steeply at larger radii . given the spacing of the transmission peaks and the distance to nearby stars , the transmitted flux detected by the lbti is constrained to come from near the hz , even if it can not be certain that it originates within it . ) and randomly distributed orientations . the dotted line shows the average transmitted flux derived in figure [ fig : fnutvsr ] . the disk will sometimes appear fainter than average , but is often a little bit brighter . the mean level is 0.026 mjy , and the medial level 0.029 mjy.,scaledwidth=50.0% ] figure [ fig : ftrhist ] shows how the total transmitted flux is distributed for a random distribution of disk orientations . the concentration of transmitted fluxes near the maximum arises because any disk with a position angle perpendicular to the transmission pattern has the same transmitted flux , regardless of inclination ( and any disk near to face - on also has the same transmitted flux ) . the low transmitted flux tail arises from disks that have a position angle parallel to the transmission pattern and are sufficiently near to edge on such that almost the entire disk lies in the central null transmission region . such a geometry is relatively unlikely , as can be seen by the lack of dots with low transmission in fig . [ fig : trvsr ] . these relatively rare low transmitted fluxes are allowed because we assume disks with negligible scale height . this distribution therefore represents a pessimistic ( but possible ) case , and becomes tighter as the disk scale height increases . the width of the distribution also depends on the range of hour angles over which a target is observed , as discussed in section [ ss : ha ] . the point here is that the transmitted flux depends on the orientation , and since this is expected to be unconstrained for lbti targets ( unless we use stellar inclination or the orientation of a resolved outer disk as an estimate , which will be possible for some of the hosts sample ) , there will be a corresponding uncertainty in any disk parameters that are derived from the observations . since the transmitted flux scales linearly with @xmath16 , the distribution of @xmath16 , and hence @xmath18 , required to reproduce a given transmitted flux is directly related to that in figure [ fig : ftrhist ] . the distribution of possible null depths for our reference model at 10 pc is therefore calculated by dividing the distribution of transmitted fluxes in fig . [ fig : ftrhist ] by the stellar flux . the distribution of zodi levels @xmath18 implied for a given observed null depth ( or upper limit ) is found by dividing the observed null depth by the distribution of model null depths for @xmath53 . this was the method employed by @xcite in modelling kin observations with ` zodipic ` . any real lbti observation takes a finite amount of time , during which any disk will rotate relative to the transmission pattern on the sky . thus , even a vertically thin edge - on disk , which can be instantaneously invisible to the lbti , will be visible when the length of the observation is taken into account . to compute how the disk transmission changes for an observation of a given object requires computing the position angle of the lbti fringe pattern on the sky at a given hour angle . the common mirror mount for the lbt means that this angle is the difference between the vector in the direction of the local vertical and a vector pointing towards equatorial north along a line of constant right ascension through the object in question ( i.e. the hour circle ) . this angle is commonly called the parallactic angle . is not instantaneously detected , but is only about 50% fainter than a face - on disk for a 4 hour observation.,scaledwidth=50.0% ] the null depth is then computed as the average null depth over the range of observed hour angles . for modelling purposes here , these are spaced evenly in hour angle , but future models will use the true distribution for each specific observation . [ fig : orient ] shows the difference between the instantaneous sensitivity for a transiting target ( left panel ) and a 4 hour observation ( right panel ) , for a target at @xmath80 declination over the parameter space of different disk orientations . this calculation is purely related to the changing position angle of the disk , and does not represent the time needed to reach a given s / n . here , as before , we have assumed a vertically thin disk . the dark region in the upper middle of the left panel shows that an edge on disk is not instantaneously detectable when the disk is aligned with the lbti fringe pattern . in the right panel , the rotation of the fringe pattern with respect to north over a four hour observation means that the sensitivity relative to a face - on disk is reduced by about 50% for the worst case disk orientation . , relative to a face - on disk . each line shows the transmission for a different disk inclination . the top axis shows the disk position angle relative to the fringe pattern , which is 180@xmath25 at transit because the lbt is facing north at that time . the curves at transit ( ha=0 ) yield the transmission - inclination relation shown in the left panel of fig . [ fig : orient ] at @xmath81 , while the average value of each curve yields the transmission - inclination relation shown in the right panel . the curve average is always greater than the transmission at transit , so observations over a wider range of hour angles can detect disks with less favourable orientations.,scaledwidth=50.0% ] to understand the origin of this increased sensitivity , fig . [ fig : orient_ha ] again shows the transmission relative to a face - on disk , but now for @xmath81 at each point in the range of hour angles from -2 to 2h . each curve shows how the transmission for a different disk inclination varies with hour angle ( or @xmath70 ) . at ha=0 , the curves correspond to a cut at @xmath81 in the left panel of fig . [ fig : orient ] . the average of each line over the hour angle range corresponds to the same cut , but in the right panel . because the average of each curve is greater than the minimum value at ha=0 , the disk sensitivity is greater . of course , the inclination and position angle dependence in figs . [ fig : orient ] and [ fig : orient_ha ] shows that it would be highly desirable to discern how the transmitted disk flux changes as a function of hour angle in order to learn about the disk geometry and the possibility of non - axisymmetric structure . for axisymmetric disks there is also a potential degeneracy between the disk geometry and the vertical scale height ; an edge - on disk with sufficient vertical extent will vary much less than the curves in fig . [ fig : orient_ha ] , and may be indistinguishable from a less inclined disk that is oriented such that the transmission variation is small . if a large variation in transmission is seen , the disk may be vertically thin , but could alternatively be non - axisymmetric . the detection of such effects will certainly be sought , but given that we expect most lbti detections to be near the sensitivity limits and therefore at low s / n we assume for now that lbti observations can be modelled by simply averaging over the hour angle . the lbti is well suited to observing dust levels in the habitable regions around nearby stars , both in terms of baseline length and observing wavelength . lbti results will therefore broadly address the goal of characterising warm dust levels without requiring additional information from other observations . however , the specific location and radial extent of the dust will be poorly constrained , and as illustrated by fig . [ fig : fnutvsr ] , significant disk emission can arise from regions both inside and outside the habitable zone and still be detected with the lbti . in general , such degeneracy in disk models will be hard to break because the goal of the lbti is to detect disks that are fainter than current detection limits ( hence the need for the modelling framework outlined in this paper ) . in some cases however , lbti results will be compared to other observations , most likely ( spectro)photometric detections or limits for the total disk flux density at the same wavelength , and in some cases limits on disk size from high - resolution imaging or interferometry . detections and/or upper limits from near - ir interferometry will provide constraints on dust levels inside the lbti iwa . lbti will similarly provide very strong constraints on the emission spectrum and location of hot dust detections .0.9 ) , non - detection with high quality mid - ir photometry is already sufficient to restrict this dust to lie well inside the habitable zone ( kennedy et al . , in prep ) . ] mid - ir photometric data will come from _ _ infrared spectrograph ( irs , * ? ? ? * ; * ? ? ? * ) , wise observations @xcite , or in the future the spectrometers on the mid - infrared instrument ( miri ) on the jwst , which cover the same wavelength range as the lbti . such photometry will provide the most useful constraints on the disk location , so we now briefly outline how the total and lbti - transmitted disk fluxes vary with model parameters , in particular the disk inner and outer radii . and @xmath15 , or for two different values of @xmath37 ( the defaults are @xmath82 au , @xmath83 au , @xmath49 ) . the lbti n band is shown as a grey stripe.,scaledwidth=50.0% ] to illustrate the flux density distribution , fig . [ fig : spec ] shows disk spectra for our reference model with @xmath53 , with examples for both sun - like and a - type host stars at 10 pc . because our standard model has the same temperature range all disk spectra have the same shape , but are brighter or fainter depending on the zodi level , the distance to the host star , and the host star spectral type ( which changes the area of the disk and hence total brightness ) . the a5v star has @xmath84 , so the disk around this star is 14 times brighter than for the sun - like star . however , the a - star is only about 5 times brighter at 11 @xmath5 m ( 10 vs. 2 jy ) , so the disk / star flux ratio has increased by a factor of about 3 . [ fig : spec ] also shows how the spectrum varies when @xmath14 , @xmath15 , and @xmath37 are changed from their standard values , and what the effect is at the lbti wavelength of 11 @xmath5 m . larger values of @xmath37 concentrate the disk emission closer to the star and hence shift the spectrum towards warmer temperatures . changing @xmath85 and @xmath15 remove emission from the inner and outer disk regions , leading to a spectrum that is closer to a single - temperature black body spectrum . as noted above and shown by the curve for @xmath86 au , the outer disk radius is large enough that the exact value matters little for the level of emission at 11 @xmath5 m . while our model gives a reasonable idea of the true disk spectrum , and could in principle be used to constrain disk structure based on observations at other wavelengths , it is inaccurate on a detailed level because grains do not emit like black bodies . three important differences are that i ) the disk will be fainter than our model at long wavelengths because grains emit inefficiently at wavelengths longer than their size , ii ) the spectrum may also be shifted to shorter wavelengths due to the presence of small grains with temperatures greater than that of a black body , and iii ) the spectrum may show non - continuum spectral features . these differences pose problems for extrapolating our model to other wavelengths , or at least mean that additional parameters such as grain properties would need to be considered . + such differences are of minor importance here however , as important comparisons with other observations can be made at the lbti wavelength , so are largely independent of the disk spectrum . the most important comparison is between the total and transmitted disk fluxes , or equally the total disk to star flux ratio and the null depth . to illustrate why combining such measurements is important , fig . [ fig : rin ] shows how the total and transmitted model disk flux changes at 11 @xmath5 m as the disk inner and outer radii change . as @xmath14 is increased the total disk flux decreases because hot emission is being removed . however , the lbti - transmitted flux changes little while @xmath14 remains small , because these changes occur behind the central transmission minimum and are invisible . when @xmath14 is outside the first transmission peak ( at about 0.7 au here ) both the total and transmitted fluxes decrease in the same way . as @xmath15 is decreased there is initially little difference in the total and transmitted fluxes because the outer disk is faint at 11 @xmath5 m , but when @xmath15 moves behind the central transmission minimum the flux drops much more steeply than the total flux . as noted earlier , fig . [ fig : rin ] shows that the lbti transmission is insensitive to our choice of disk inner and outer radii , as long as the disk is much wider than the habitable zone . given such relations between the total and transmitted disk fluxes , it is clear that observations of both may constrain the disk location . however , the transmitted disk flux is a function of both the disk size and the orientation , so the best constraints on the disk size require the orientation to be known ( or take some assumed value ) . it is unlikely that the position angle will be inferred from lbti measurements in many cases , so the disk orientation would probably be assumed based on other system information , such as coplanarity with a resolved outer cool disk component , the known inclination and/or position angle of the host star s rotation axis , or of planet orbits , or a combination of all three @xcite . see defrre et al . ( submitted ) for an application of this assumption to @xmath87 crv . currently nasa is focussed on developing concepts for three optical ( 0.4 - 1 @xmath5 m ) exoplanet - imaging missions , exo - c ( a coronagraph ) , exo - s ( a starshade ) , as well as an ambitious precursor , the coronagraph on wfirst - afta . therefore , the thermal dust emission information from the lbti must be converted to predict the impact on scattered light imaging . we now derive a simple prescription for converting the dust levels in the above model into scattered light surface brightness estimates . scattered light predictions are in general difficult , and have largely proven unsuccessful to date ( e.g. * ? ? ? * ) , with debris disks imaged in scattered light generally seen to be much fainter than predicted based on theoretical grain models that match observed thermal emission . the typical minimum grain size in debris disks is thought to 1 - 10 @xmath5 m for sun - like stars ( e.g. * * ; * ? ? ? * ; * ? ? ? * ) , and the steepness of the size distribution means that these grains dominate the surface area . such grains are expected to scatter optical light fairly isotropically , and have fairly large albedoes of @xmath880.5 . the scattered light faintness of debris disks , where albedoes of 0.05 - 0.1 are seen ( e.g. * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) , likely arise from incorrect assumptions about grain properties and sizes , and how these grains scatter starlight . the properties of exo - zodi are not sufficiently well known that considering physically motivated possibilities for different grain sizes and properties would make predictions for their scattered light brightness more certain . therefore , regardless of the physical reason , we will assume that the @xmath890.1 effective albedoes seen for scattered light disks around nearby stars are representative , and assume isotropic scattering . it may be that the properties of warm dust are different to those inferred from scattered light detections , since for example hot dust detected with near - ir interferometry is generally inferred to originate in much smaller grains . for higher / lower albedoes , our scattered light predictions would be higher / lower by the same factor . while the model surface density @xmath12 is connected to the true optical depth and the surface area of grains from which the emission arises , the model surface density is the true optical depth only if the grains behave like black bodies . that is , real grains reflect starlight , so the true total surface area in the disk is always higher than @xmath12 . to be more realistic , and to allow for the possibility of scattered light , the grain absorption and scattering properties need to be considered . consider a disk with particles with a range of sizes @xmath90 , and @xmath91 the true cross - sectional area per unit area of particles in the size range @xmath90 to @xmath92 . using the absorption efficiency as a function of size and wavelength @xmath93 , the thermal emission is @xmath94 ) dd.\ ] ] though the dependence is not included explicitly here for simplicity , all quantities in this equation also vary with location in the disk due for example to the disk structure and changing composition . using the scattering efficiency @xmath95 the scattered light emission can be written in various ways , but a convenient form is @xmath96 which we could also express in terms of albedo @xmath97 by substituting @xmath98 . if we assume the albedo @xmath99 is the empirical value of 0.1 , independent of grain size and wavelength , then because @xmath100 the surface density @xmath101 is approximately the true optical depth , but is underestimated by a factor of @xmath102 ( i.e. @xmath103 $ ] ) . that is , with these assumptions the thermal surface brightness could be written @xmath104 the scattered light brightness is therefore calculated using the dust surface density and the empirical effective albedo . with these assumptions , the predicted scattered light emission from the model of section [ ss : model ] would be @xmath105 for simplicity we use the value of @xmath12 derived from our modelling , and thus include an extra factor of @xmath106 . therefore , with our model the scattered light surface brightness relative to the stellar flux in the habitable zone ( i.e. @xmath107 ) of a 1 zodi disk decreases as @xmath108 , as the habitable zone is pushed farther from the star by the increased luminosity . this is the approach we use below to estimate the limiting scattered light surface brightnesses for hosts , and for the lbti @xmath87 crv detection . an extension to this approach would be to add a phase function that accounts for forward scattering properly . the albedo could then be derived theoretically and calibrated with observations . while in the above case the scattered light can be calculated simply along the radial direction , and for a given orientation turned into an image , use of a phase function @xmath109 requires a three - dimensional calculation that includes the star - particle - observer scattering angle @xmath69 at each location in the disk . given that we expect relatively large uncertainties in disk parameters , even in the case of a detection , use of a detailed three - dimensional calculation including ( assumed ) grain properties and a phase function will in general be unwarranted . to convert the simple scattered light prediction of equation ( [ eq : ssca ] ) to an observable for an arbitrary disk inclination a three dimensional calculation must be made to account for brighter disk ansae . however , the disk inclination will in general be unknown , so a simple approximation for deriving a representative scattered light surface brightness would be to assume an average inclination of @xmath110 , and for a vertically thin disk an increase of @xmath111 . we now show two examples using the model described above . we first use the lbti commissioning measurement for @xmath87 crv to illustrate how zodi levels are derived , and then derive zodi limits for the hosts survey . these levels assume that the disk orientation is not known , but in some cases coplanarity with a resolved outer disk could be assumed to further constrain the disk brightness . this assumption can be made for @xmath87 crv , and we refer the reader to defrre et al . ( submitted ) for an in - depth discussion of these results for this system . using @xmath87 crv with a null depth of @xmath112% as an example , we first show how the distribution of zodi levels is derived assuming an unknown disk orientation . this procedure is generic in that it can be applied to any lbti observation , whether a significant detection or an upper limit was found . the lbti is still in commissioning so the zodi sensitivity derived here is not illustrative of the expected performance . [ fig : etacrvlim ] shows the @xmath113-plane coverage of the lbti observation relative to the disk position angle , and the distribution of zodi levels calculated using our reference model . we include the rotation of the disk relative to the fringe pattern during the observation to compute the average null depth for each disk orientation . as described at the end of section [ ss : trx ] the distribution is essentially the inverse of that shown in fig . [ fig : ftrhist ] . the distribution is strongly peaked at about @xmath114 , with a tail of larger values due to unfavourable disk orientations . the 1@xmath115 zodi uncertainty due purely to the calibrated null depth measurement is given by the width of the `` median @xmath116 null uncertainty '' error bar in the right panel of fig . [ fig : etacrvlim ] . similarly , the 16 and 84% levels from the cumulative distribution give a representative 1@xmath115 range due to the orientation distribution . the upper uncertainty on the zodi level is therefore set by the orientation distribution , while the lower uncertainty is set by the null depth measurement . the median zodi level for @xmath87 crv is @xmath117 if the uncertainty is set by the lbti null depth measurement . including the 1@xmath115 range from the orientation distribution in quadrature , the range covered on either side of the median value is @xmath118 . in general , we expect that the null depth uncertainty will dominate the lower bound on the zodi level , and the orientation distribution will dominate the upper bound . for comparison , based on a kin detection and using the ` zodipic ` model mennesson et al . ( in press ) found @xmath119 for @xmath87 crv at 8.5 @xmath5 m with no assumptions about the disk orientation . our derived zodi level is different for two reasons : i ) @xmath87 crv is hotter than the sun ( 6900k ) , so our luminosity - dependent zodi definition will lead to a zodi level about 1.7 times smaller than ` zodipic ` , and ii ) the mid - ir spectrum of the @xmath87 crv disk increases more steeply with wavelength than a black body ( i.e. has a silicate spectral feature ) , so although our model has nearly the same temperature profile as ` zodipic ` , our derived zodi level will be slightly larger because lbti observes at 11 @xmath5 m . therefore , direct comparisons between our zodi levels and those using ` zodipic ` should not be made . using the simple prescription for the scattered light surface brightness described in section [ ss : scat ] , we can convert the zodi levels shown in fig . [ fig : etacrvlim ] into a prediction for the face - on scattered light surface brightness . here , the value of interest is the surface brightness in the habitable zone , as this is where planets would be sought . the level is then calculated using equation ( [ eq : ssca ] ) , with @xmath120 because @xmath16 is the value at the radial distance where the equilibrium temperature is the same as earth s . adopting @xmath121 , and using @xmath122 jy in v band , @xmath123 pc , @xmath124 au , and @xmath125 yields a scattered light surface brightness of @xmath126 mjy / arcsec@xmath10 , or about 15 mag / arcsec@xmath10 . as noted above , this habitable zone estimate may be increased by a factor of a few to account for the inclination of the disk . given a prediction for the sensitivity of the lbti and a sample of stars that will be observed , we can make predictions for the sensitivity of the hosts survey . in what follows we assume a 1@xmath115 uncertainty on the lbti calibrated null depth of @xmath127 , and hence the limits presented are also at 1@xmath115 . limits are calculated as above , using the median of the distribution of zodi levels over the random distribution of orientations . though we do not account for it here , there is some uncertainty in the stellar flux densities and luminosities used , which we estimate to contribute at about the 5% level . here we calculate the sensitivity assuming an observation at a single hour angle , but when the model is applied to real observations in the future the calculations will account for sky rotation as described in section [ ss : ha ] . these calculations are carried out for both our reference disk model , and a `` worst - case '' scenario where the dust is restricted to lie only within the habitable zone , which we assume to lie between temperatures of 320 and 210k . as shown in figs . [ fig : fnutvsr ] and [ fig : rin ] , the lbti is sensitive to emission that lies interior and exterior to the habitable zone , so the dust surface density must be higher in the latter case for an lbti detection at the same sensitivity . as noted above in section [ ss : zodi ] , the `` zodi levels '' derived with this different radial structure are also different and should really be considered as enhancements over the surface density derived for the solar zodiacal cloud , rather than zodi levels to be compared with other values . the hosts survey sample is described by weinberger et al . ( submitted ) , the key aspect being that targets are chosen such that their habitable zones have larger angular sizes than the first lbti transmission peak , so observations directly probe the levels of habitable zone dust . the sample is split by @xmath128 colour at 0.42 into `` sun - like '' and `` sensitivity '' sub - samples , which simply reflects the levels of dust that can be detected . ) .,title="fig:",scaledwidth=50.0% ] + ) .,title="fig:",scaledwidth=50.0% ] + ) .,title="fig:",scaledwidth=50.0% ] figure [ fig : leak ] shows the predicted disk to star flux ratios , null depths , and sensitivity in zodi units for the hosts sample using our reference model ( red and blue symbols ) and our worst case scenario ( grey symbols , which we discuss below ) . the top and middle panels show the flux ratios and null depths expected for a 1 zodi model around these stars , and the bottom panel shows the sensitivity in zodis for the predicted lbti sensitivity . there are clear trends with stellar luminosity , which can be understood as follows . with our model the disk flux density at fixed wavelength scales with the angular area , i.e. , @xmath129 ( with our zodi definition @xmath16 is constant , but we include it to consider other zodi definitions below ) . the stellar flux , on the other hand , scales @xmath130 in the rayleigh - jeans regime so the total disk to star flux ratio only depends on the star , and is higher for earlier spectral types @xmath131 this dependence in the top panel of fig . [ fig : leak ] arises due to the stronger scaling of stellar luminosity with temperature than flux density at 11 @xmath5 m , with the variations from a perfect correlation arising due to variation in @xmath1 at fixed @xmath132 ( i.e. different stellar radii ) . at fixed distance , as the stellar temperature and luminosity increase the habitable zone is pushed outwards ( @xmath133 ) and its area and brightness increase more rapidly than the stellar flux . alternatively , for a fixed disk angular size ( and hence fixed disk brightness ) , increasing the stellar temperature and luminosity pushes the system to greater distances and the star becomes fainter due to increasing distance faster than it becomes brighter due to an increased temperature . the null depths for a 1 zodi disk are shown in the middle panel , and the trend has a similar origin as the disk to star flux ratio . at fixed distance , increasing stellar luminosity increases the disk surface brightness at fixed angular radius because @xmath17 increases ( i.e. increases if @xmath37 is positive ) . assuming that most of the transmitted disk flux originates from a constant angular scale ( i.e. near the first transmission peak ) , the transmitted disk flux therefore increases as @xmath134 ( assuming that the effect of the changing disk temperature in the first transmission peak is small ) . combining this expression with the stellar flux yields a null depth @xmath135 \ , .\ ] ] hosts stars are chosen to have habitable zones with similar angular sizes and for our reference model @xmath49 , so the term in square parentheses varies relatively little and the null depth almost entirely depends on the stellar temperature . the absolute null depth level of course also varies linearly with the zodi level , so the zodi limits can be derived by dividing the expected sensitivity of @xmath127 by the null depth values in the middle panel ( i.e. by solving equation ( [ eq : leakmod ] ) for @xmath18 ) @xmath136 \ , .\ ] ] the resulting sensitivities show that what we define as solar system levels of zodiacal dust are at the predicted 1@xmath115 noise level for early - type stars , as are 3 - 10 zodi disks around sun - like stars . this discussion of scalings applies equally to the narrow worst case scenario , shown as grey symbols in fig . [ fig : leak ] . as expected for a narrower disk that emits over a smaller total physical area , the disk to star flux ratios are lower than for our reference model , as are the null depths . while the difference between the two models is about a factor of five in disk / star flux ratio , the difference in null depths is only a factor of two to three because most of the flux removed from the disk in the narrower model is hidden behind the central transmission minimum ( e.g. fig . [ fig : rin ] ) . similarly , the difference in zodi levels is a factor two to three higher in this case compared to our reference model . therefore , the effect of this pessimistic scenario in terms of habitable zone dust levels is relatively minor . because our zodi definition is based on constant surface density in the habitable zone , the @xmath18 dependence on luminosity directly shows that the lbti can truly detect lower surface densities of dust in the habitable zones of earlier type stars . this conclusion does not depend on our zodi definition because the lbti is sensitive to @xmath137 . a different definition , for example @xmath138 , makes the zodi limit approximately constant ( @xmath139^\alpha$ ] ) , but also implies @xmath140 ( and the sensitivity to habitable zone surface density is the same as with our definition ) . ) for the reference disk model for hosts survey stars , split into `` sun - like '' and `` sensitivity '' sub - samples ( red circles and blue squares ) . these limits use equation ( [ eq : ssca ] ) and the zodi limits from fig . [ fig : leak ] . grey symbols show a pessimistic narrow - disk scenario where disks only cover the habitable zone ( from 320 to 210k).,scaledwidth=50.0% ] given the detection limits in the bottom panel of fig . [ fig : leak ] , we can also derive the face - on habitable zone ( @xmath141 au ) scattered light surface brightness of disks at the hosts survey detection limits . these limits are shown in fig . [ fig : ssca ] for @xmath121 as defined in section [ ss : scat ] . red and blue symbols show limits for our standard disk model , which again depend on the stellar luminosity . the origin of the dependence on stellar luminosity can be understood by rewriting equation ( [ eq : ssca ] ) using the zodi limit scaling above @xmath142 like the zodi limits , the scattered light levels in the habitable zone at these limits decrease with stellar luminosity . the squared distance and luminosity dependence is simply the geometric effect that accounts for @xmath143 dilution of light and that surface brightness is an angular measure . the extra @xmath37 dependence is because the habitable zone is not at the first transmission peak for all targets . if the lbti observation is dominated by emission near the first transmission peak , but the habitable zone is slightly farther out ( i.e. @xmath144 ) , the @xmath37 dependence represents a model - dependent extrapolation that is relatively unimportant while @xmath37 is small . as with the lbti sensitivity to habitable zone surface density , our zodi definition has no effect on the scattered light predictions and is merely an intermediate step for deriving the true quantity of interest . that is , as described above the lbti sensitivity to @xmath12 in the habitable zone ( i.e. @xmath145 ) , which sets the limits on @xmath146 , will always be the same . for the example of a constant @xmath147 zodi definition therefore , @xmath18 is constant but @xmath148 and equation ( [ eq : scamod ] ) is the same . because the hosts sample is explicitly chosen such that the lbti is sensitive to thermal emission from dust in the region of interest for earth - imaging , the scattered light predictions do not strongly depend on the choice of our disk model parameters ( but of course depends on albedo ) . the predictions are not totally independent of our disk model however , since for example changing the steepness of the radial profile makes the dust in the model more or less concentrated relative to the lbti transmission pattern , thus changing the derived zodi level . varying @xmath37 between -1 and 1 yields changes of @xmath1491 mag in the scattered light predictions , meaning that the model dependent uncertainty is similar to the variation expected from the unknown disk inclination . for solar - type stars , the scattered light levels in fig . [ fig : ssca ] for our face - on reference model at the @xmath894 zodi detection limit is 20 - 21 mag arcsec@xmath11 . for comparison , using a 4 zodi disk the ` zodipic ` model predicts 21.2 mag arcsec@xmath11 at 1 au with default parameters ( a solar analogue with @xmath150 and use of a phase function ) , and 21.6 mag arcsec@xmath11 for @xmath121 and isotropic scattering . our simple model therefore compares well with the more complex calculation made by ` zodipic ` . for a 60@xmath25 inclined disk , ` zodipic ` gives 21.1 mag arcsec@xmath11 , and for 90@xmath25 gives 20.6 mag arcsec@xmath11 , and therefore the increase from face - on to edge - on is about 1 magnitude . our model has smaller vertical extent than ` zodipic ` , so the increase in surface brightness for inclined disks will be larger , roughly a factor of three for a 60@xmath25 inclined disk . the grey symbols in fig . [ fig : ssca ] show the scattered light surface brightness for narrower disks that only lie in the habitable zone . these limits are about 1 magnitude brighter than our reference model , so similar to the variation in brightness with the ( unknown ) disk orientation . therefore , the scattered light limits from lbti observations are fairly robust to disk width . we have outlined a parameterised disk model to be used for modelling and interpreting mid - ir exo - zodi observations with the lbti . using this model , we have illustrated how to derive dust limits and levels for exo - zodiacal clouds , and how these can be converted to scattered light surface brightnesses needed for planning future missions that will image extrasolar earth - analogues . using the hosts sample that the lbti will observe , we illustrate the survey detection limits both in terms of zodi units and the expected scattered light levels at these limits . these limits are around ten times the solar system level for solar - type stars , and thus the lbti is expected to provide stringent limits with key information that will help plan future earth - imaging efforts . the large binocular telescope interferometer is funded by the national aeronautics and space administration as part of its exoplanet exploration program . this work was supported by the european union through erc grant number 279973 ( gmk , op , abs & mcw ) . we thank the reviewer for valuable comments . , o. , defrre , d. , coud du foresto , v. , di folco , e. , mrand , a. , augereau , j .- c . , ertel , s. , hanot , c. , kervella , p. , mollier , b. , scott , n. , che , x. , monnier , j. d. , thureau , n. , tuthill , p. g. , ten brummelaar , t. a. , mcalister , h. a. , sturmann , j. , sturmann , l. , & turner , n. 2013 , , 555 , a104 , d. , absil , o. , augereau , j .- c . , di folco , e. , berger , j .- p . , coud du foresto , v. , kervella , p. , le bouquin , j .- b . , lebreton , j. , millan - gabet , r. , monnier , j. d. , olofsson , j. , & traub , w. 2011 , , 534 , a5 , j. , van lieshout , r. , augereau , j .- c . , absil , o. , mennesson , b. , kama , m. , dominik , c. , bonsor , a. , vandeportal , j. , beust , h. , defrre , d. , ertel , s. , faramaz , v. , hinz , p. , kral , q. , lagrange , a .- m . , liu , w. , & thbault , p. 2013 , , 555 , a146 , r. , serabyn , e. , mennesson , b. , traub , w. a. , barry , r. k. , danchi , w. c. , kuchner , m. , stark , c. c. , ragland , s. , hrynevych , m. , woillez , j. , stapelfeldt , k. , bryden , g. , colavita , m. m. , & booth , a. j. 2011 , , 734 , 67

nano electromechanical systems ( nems ) are test beds for the observation of fundamental quantum behaviour of objects which are huge on the scale of individual atoms . for example , recent experiments @xcite have allowed a detailed study and control of single phonons by cooling a macroscopic resonator mode close to its ground state and coupling it to single electronic degrees of freedom . one fascinating aspect of nems is their conceptual simplicity that nevertheless can give rise to highly complex physics , and the links that can be established to other fields such as molecular electronics or optomechanics @xcite . one of the challenges are the details of the oscillator electron coupling , i.e. in the language of measurement theory to understand , utilize @xcite and control @xcite the ` back action ' effects of the detector ( e.g. , superconducting single electron transistors @xcite ) onto the oscillator . in many cases , even if no further approximations in simple theoretical models are made , the coupling to external electronic reservoirs that link the detector to the outer world is treated perturbatively , i.e. , in the framework of ( quantum ) master equations . this has turned out to be a highly successful approach , in particular to describe such various systems as nems coupled to single electron transistors ( sets ) @xcite , franck condon blockades @xcite in transport through molecules with strong electron phonon coupling , or quantum shuttles@xcite . from the theory of electronic transport through nanostructures @xcite , however , it is known that such approximations usually are reliable only in the limit of high external voltage bias , where non markovian effects @xcite due to quantum coherences between the external reservoirs and the electronic system ( set , quantum dot etc . ) can be neglected . it is therefore desirable to develop tools that allow a description of nems beyond the master equation regime ( weak electron leads coupling ) and at the same time are not merely perturbative in the coupling of the oscillator to the electronic environment @xcite . in the past , the coupling of electrons to a single bosonic mode has been solved exactly for the case where only one single electron is present @xcite , i.e. in an empty band approximation . the inclusion of fermi sea reservoirs at different chemical potential transforms this into a difficult many body problem out of equilibrium , and approximations are necessary @xcite . our approach in this paper is to combine exact solutions of the electronic system with a semi classical expansion , together with an adiabatic approximation for the oscillator dynamics within the feynman vernon influence functional ( double path integral ) theory @xcite . we revise this method , which has first been used for simple nems models by mozyrsky and co workers @xcite , and extend it to allow for the description of a relatively large class of non equilibrium electronic environments . the key idea is a systematic expansion around the classical path in order to obtain a langevin equation for the oscillator . already at the simplest level of this approximation ( neglecting quadratic fluctuations around the diagonal path in the reduced density matrix of the oscillator ) , the coupling to the electronic non equilibrium environment gives rise to non trivial effects such as effective oscillator potentials and non linear friction coefficients leading to both positive and negative damping @xcite . gaussian fluctuations around the classical path are built into the theoretical description , but they have to be evaluated by numerical solutions of the underlying langevin equations which is not done in this paper . we compare two non interacting electronic ` quantum dot ' models with one and two levels between source and drain reservoirs : a single dot , and two dots in series . the oscillator couples linearly to the dot occupation ( single dot ) or to the occupation difference ( double dot ) . one particular feature of the double dot case ( where quantum superpositions of the electrons become important ) is the occurence of limit cycles in phase space caused by a negative damping . the paper is organized as follows : after introducing the path integral formalism with a generic model in sec . ii , we present the single dot case in sec . iii and the double dot case in sec . detailed derivations of the important formulae can be found in the appendices . a large class of nems can be described as a composition of an electronic system @xmath0 , a mechanical system @xmath1 and a linear coupling between the two . thus we set up a generic hamiltonian @xmath2 by @xmath3 here , @xmath1 describes a single oscillator with @xmath4 ( momentum ) and @xmath5 ( position ) operators . the oscillator mode is confined in a potential @xmath6 , @xmath7 denotes an electronic force operator , and @xmath8 labels the oscillator mass . whithin this paper the reduced planck constant is set to one ( @xmath9 ) . our generic model does not include an additional oscillator damping mechanism . in the usual master equation treatment of nems , lindblad - form damping due to external degrees of freedoms is included phenomenologically . in the path integral formalism used here , such degrees of freedom can be easily included at least for linear or weak coupling to the oscillator . in order to elucidate the effect of the electronic environment that we treat in all orders in the coupling to external electronic reservoirs ( contained in @xmath0 ) , we choose not to include additional damping terms in our model here . we describe the oscillator dynamics by the reduced density matrix of the oscillator in position representation @xmath10 , for which we derive a semiclassical equation of motion for the oscillator position by using feynman vernon influence functional theory similar to mozyrsky and co - workers@xcite . assuming that the total density matrix @xmath11 factorises at the initial time @xmath12 into a system and a bath part @xmath13 , the propagation of the reduced oscillator density @xmath14 at time @xmath15 is given by a double path integral@xcite , cf . appendix [ sec.:gen_mod ] . a transformation to center of mass and relative coordinates @xmath16 has the notion to detach the classical trajectory @xmath17 from the quantum mechanical deviations . within a born oppenheimer approximation , change of variables allows us to study a slow oscillator by an adiabatic approximation of the classical trajectory @xmath18 cf . appendix [ sec.:gen_mod ] . in this approach , the typical timescale of the oscillator movement is slow compared to the electronic transition rates . in the subsequent , when having introduced the angular oscillator frequency by @xmath19 and electron transition rates by @xmath20 , @xmath21 we have to satisfy the condition @xmath22 , @xmath21 . in the next step , we derive a stochastic equation of motion for the classical trajectory , taking into account the propagation of the initial reduced density matrix , cf . appendix [ sec.:gen_mod ] . the key step here is a cluster expansion to quadratic order in the off - diagonal path @xmath23 that describes the gaussian fluctuations around the classical oscillator trajectory , where the fluctuations are determined by the properties of the non - equilibrum environment . to achieve a self consistent equation of motion , we then re - insert the full time dependence of the fixed classical trajectory in accordance with the adiabatic approximation and end up with the langevin equation @xmath24(t ) \rangle } } + \dot x_t a[x](t ) = \xi_t\label{eq.:gen_model_langevin}. \end{aligned}\ ] ] here the interaction picture of the electronic operator is given by @xmath25(t ) = \exp[i(\mathcal h_{\rm e}-\hat f x)t]\hat f\exp[-i(\mathcal h_{\rm e}-\hat f x)t]$ ] , and @xmath26 is a stochastic force with zero mean and the correlation function @xmath27(t)\delta\tilde f[x](t ' ) \rangle}}. \end{aligned}\ ] ] the fluctuation of the electronic operator is defined by @xmath28(t)=\tilde f[x](t)-{\ensuremath{\langle \tilde f[x](t ) \rangle}}$ ] . the friction @xmath29 $ ] is given by @xmath30(t ) & = 2\int_{t_0}^{t}dt'\;t'\;{\ensuremath{\operatorname{im}}}{\ensuremath{\langle \delta\tilde f[x](t)\delta\tilde f[x](t ' ) \rangle}}. \end{aligned}\ ] ] in the following we choose @xmath31 as initial time . for the specific cases of single and double dots , we checked that the upper integration boundary can be extended to infinity . the ahm combines a single bosonic mode with a simple electronic transport system . we describe the bosonic part in first quantisation as @xmath32 with the bosonic position and momentum operators @xmath33 and @xmath4 , the phonon frequency @xmath19 and the phonon mass @xmath8 . the oscillator length is defined by @xmath34^{1/2}$ ] . the electronic part is a single dot level confined between two leads : @xmath35 the dot level has energy @xmath36 and creation / annihilation operators @xmath37/@xmath38 . the operators @xmath39/@xmath40 create / anihilate electrons with momentum @xmath41 and energy @xmath42 in a free electron gas in the left ( @xmath43 ) or right ( @xmath44 ) lead . electronic transitions are possible with amplitude @xmath45 between the dot and a state in the lead . between the two subsystems there is a simple linear coupling with coupling constant @xmath46 , such that the total hamiltonian reads @xmath47 for convenience we introduce the dimensionless coupling constant : @xmath48 + here , we regard spinless electrons . note that the generalization to a model including both spin directions requires an onsite interaction term like @xmath49 . in the following , we only consider non interacting electrons . a physical realization of the model discussed here would correspond to either spin polarized electrons or a coupling to the oscillator that effects only electrons of one certain spin direction . sketch of the ahm model . two leads ( grey bars ) with chemical potentials @xmath50 , @xmath51 embed the dot level which couples to an oscillator . here @xmath52 is the bias voltage and @xmath53 the electron charge . the level energy @xmath36 is shifted by the oscillator position @xmath54 , the shifted level energy @xmath55 is stationary , if the oscillator force @xmath56 and the electron force @xmath57 \rangle}}$ ] are in balance . here @xmath8 and @xmath19 describe the oscillator mass and angular frequency , and @xmath46 ( g ) denotes the coupling constant , cf . eq .. for sufficiently small tunneling rates @xmath20 , @xmath21 the stable stationary level energies ( dashed lines ) are located at @xmath58 . the instable points of @xmath59 ( dotted lines ) are located around the chemical potentials . ] when applying our generic form to the ahm , we obtain the langevin equation @xmath60(t ) & = \xi_t \label{eq.:set_langevin_eq } \end{aligned}\ ] ] with the effective force @xmath61 and the friction @xmath29 $ ] : @xmath62,\nonumber\\ a[x](t ) & = 2 \lambda^2 \int_{t_0}^{t}dt'\;t ' { \ensuremath{\operatorname{im}}}{\ensuremath{\langle \delta\tilde n(t)\delta\tilde n(t ' ) \rangle}}. \end{aligned}\ ] ] the effective force has two contributions ; a term proportional to the elongation ( hooke s law ) and the electron force which is proportional to the dot occupation ; which only contributes if the dot is occupied . the friction term @xmath63 $ ] results from stochastic electron jumps between the leads and the dot . left ) density plot of the effective oscillator potential @xmath64 as a function of oscillator position @xmath65 and @xmath52 in units of @xmath19 and @xmath66 with the parameters @xmath67 , @xmath68 , @xmath69 and @xmath70 . for increasing bias voltage , the effective potential shows two , three , two and one minima . this different regions are separated by white lines at @xmath71 . right ) @xmath64 and the two contributions @xmath72 , @xmath73 to the force @xmath61 at bias voltages ( symmetric choice ) @xmath74 . @xmath64 exhibits extrema where the two force contributions ( grey and dashed line ) are in balance . the width of the center plateau in the electronic force contribution @xmath75 grows with increasing bias voltage @xmath52 , cf . fig . for the intermediate occupation @xmath76 . for sufficiently high bias voltage only one minimum remains . ] for finite bias voltage , figure [ fig.:set_force_balance ] shows the positions of the dot energy level and the points of instable balance , which result from the balance of both forces . the effective force and therewith the oscillator potential are determined by the dot occupation @xmath77 , whereas the friction and the stochastic force correlation @xmath78 depend on the imaginary / real part of the dot correlation function . the dot correlation function in terms of the lesser and greater green s function ( which are derived in appendix [ sec.:ahm_corr ] ) reads @xmath79 the occupation of the dot is calculated in an adiabatic approach with the help of the lesser green s function @xmath80 , cf . appendix [ sec.:ahm_corr ] , where we assume constant tunneling rates @xmath81 . for finite temperatures , the dot occupation reads @xmath82 whereby @xmath83 stands for the inverse temperature , @xmath84 is a short notation for @xmath85 , @xmath86 denotes the chemical potetials and @xmath87 . @xmath88 designates the digamma function . by integration we obtain the effective potential @xmath89\bigg|_{\xi=\frac12 + \frac{\beta\gamma}{4\pi}}\ ; + \frac12 \frac{x}{l_0}\bigg[\frac{x}{l_0}-g\bigg]\omega_0 , \end{aligned}\ ] ] with @xmath90 denoting the gamma function . in the absence of friction the dynamics of the oscillator is determined by the effective potential @xmath91 . the high temperature case is of minor interest , because the temperature washes out the structures of @xmath91 . in fig . [ fig.:set_density_composed_sliced ] we present features of the effective potential @xmath91 at zero temperature . @xmath91 has minima when the effective force @xmath92 is zero . this is the case when the oscillator force and the electron force are in balance . we plot the two contributions to the effective force : the force @xmath72 is proportional to the displacement @xmath65 , the contribution @xmath73 results from the dot occupation ( scaled with @xmath93 ) and has two steps at @xmath94,\qquad\alpha\in\{\rm l,\ ; r\}. \end{aligned}\ ] ] in the upper part of fig . [ fig.:set_density_composed_sliced ] we easily recognize that for large bias voltage there will be only one minimum at @xmath95 . by changing the bias voltage or the coupling strength , we can reach situations where two minima at @xmath96 and at @xmath97 are added , like in the middle part of fig . [ fig.:set_density_composed_sliced ] and also situations , where only the two minima at the side remain and the one in the middle vanishes , like in the bottom part of the [ fig.:set_density_composed_sliced ] ( this agrees with mozyrsky et al.@xcite ) . increasing bias voltage shifts the steps in @xmath73 apart , whereas increasing the coupling constant minimises the distance . the positions of the minima / steps are exact in the zero rate limit ( @xmath98 , @xmath99 ) where the averaged occupation is step like . for finite rates , the steps smoothes out . the form of the oscillator potential already hints towards the phase space spanned by position @xmath65 and momentum @xmath100 . in the absence of friction , the minima of the potential correspond to the fixed points of the oscillator motion . the classical trajectory @xmath101 is obtained by neglecting fluctuations due to the stochastic force @xmath26 in ; therefore the equation of motion reads @xmath102(t)-f_{\rm eff}({\ensuremath{\langle x_t \rangle } } ) \label{eq.:stochastical_eq_of_motion_cl } = 0 . \end{aligned}\ ] ] in fig . [ fig.:set_phase_space_comparison ] we show the effective potential and the friction , as well as the solutions of the classical equation of motion in phase space without and with friction . the rows correspond to three different voltages leading to the three typical cases with one , three and two minima of the potential ( the latter case is investigated in @xcite ) . the initial conditions are chosen for each phase diagram separately in order to make the characteristic shapes visible . the phase diagrams without friction follow directly from the shape of the effective potential . the trajectories including the friction follow the ones without friction for a while until a position with a peak in the friction is reached that produces a kink - like damping feature . the friction causing the kinks is maximal , when the shifted energy level @xmath84 is in resonance with the chemical potentials @xmath103 ( fig . [ fig.:set_force_balance ] ) , because at these points the average occupation switches and the electronic fluctuations and therewith the friction itself is very large . between its peaks the damping does not completely vanish at finite bias , so all phase space trajectories end up in spirals and reach stable fixed points after infinite time ( we have terminated all phase trajectories after the same time ) . the position of the peaks ( instable fixed points ) separates the stable fixed points . in the two minima case ( @xmath104 in fig . [ fig.:set_phase_space_comparison ] ) the left fixed point corresponds to the zero occupied electronic level in fig . [ fig.:set_force_balance ] and the right fixed point to the fully occupied level . for increasing bias voltage , the friction splits and we obtain a third fixed point in between , corresponding to the average occupied state . for sufficiently high bias voltage ( large transport window ) only the average occupied state survives . apart from the stable fixed points we also observe saddle points that repell the trajectories near the kinks where the damping is large . they result from the kinks in the dot occupation ( fig . [ fig.:set_density_composed_sliced ] , right ) and correspond to the dotted lines in fig . [ fig.:set_force_balance ] and the bright v - structure in fig . [ fig.:set_density_composed_sliced ] , left . at large bias the saddle points move out of the range of allowed positions , the same happens to the peaks in the friction . at infinite bias the effective potential becomes a simple parabola and the friction vanishes completely ( this can be checked analytically ) . the dqd that we treat in this section consits of two single dot levels coupled by a tunnel barrier . again we assume a coupling to a single bosonic mode . the total hamiltonian is composed of the oscillator part @xmath105 , cf . eq . , the electronic part @xmath106 and an interaction part which describes the coupling between the oscillator and the two dots . in contrast to the ahm , here the oscillator couples to the difference of the occupation numbers with the coupling strength @xmath46 . the total hamiltonian therefore reads @xmath107 containing the electronic part @xmath108 . \end{aligned}\ ] ] in this model , @xmath109/@xmath110 denotes the creation / annihilation operator of the @xmath111th ( @xmath112 ) dot . @xmath113 designates the corresponding level energy and @xmath114 describes the tunnel coupling matrix element between the two dots . note that we include no coulomb interaction terms here . the stationary average of the population difference @xmath115 corresponds to the electronic force operator @xmath7 of the generic model , eq . . the occupation of the @xmath111th dot @xmath116 can be calculated by using keldysh s equation , see section [ sec.:dqd_green ] . therewith the occupation difference of the left / right dot follows from @xmath117 , \end{aligned}\ ] ] with the abbreviations @xmath118 and @xmath119 , \hspace{0.2 cm } b= \big[{\ensuremath{|t_c|}}^2+\tilde\nu_{\rm l}^2+(\gamma/4)^2\big ] , \end{aligned}\ ] ] whereas we assumed @xmath120 . calculating the integrals leads to @xmath121 . \nonumber\\ \end{aligned}\ ] ] they are expressed in terms of the auxiliary functions @xmath122 defined in appendix [ sec.:dqd_aux_int ] . when applying the generic model , cf . , to the dqd hamiltonian @xmath123 , we obtain the langevin equation @xmath60(t ) & = \xi_t \label{eq.:dqd_stoch_motion_eq } \end{aligned}\ ] ] with the effective force @xmath61 and the friction @xmath29(t)$ ] . in contrast to the ahm , the effective force is affected by the population difference and not only by the occupation number . explicitly , @xmath124(t ) & = 2\lambda^2 \int_{t_0}^{t}dt'\;t ' { \ensuremath{\operatorname{im}}}{\ensuremath{\langle \delta\tilde\sigma_z(t)\delta\tilde\sigma_z(t ' ) \rangle}}\label{eq.:def_dqd_friction}. \end{aligned}\ ] ] in appendix [ sec.:dqd_friction ] we derive the explicit expression for the friction . in the case of infinite bias ( ib ) and @xmath120 we obtain @xmath125(t ) = - 8 \ \left|t_{c}\right|^{2 } \frac{\lambda^{2}}{\gamma } \ \widetilde{\nu}_{\rm l } \ \frac{\widetilde{\nu}_{\rm l}^{2 } + \left|t_{c}\right|^{2 } + 5 \frac{\gamma^{2}}{16}}{\left(\widetilde{\nu}_{\rm l}^{2 } + \left|t_{c}\right|^{2 } + \frac{\gamma^{2}}{16}\right)^{3}}.\ ] ] in contrast to the ahm the friction does not disappear for infinite bias . the second difference that we can recognise by regarding the prefactor @xmath126 is that we obtain regions where the friction is _ negative_. the latter also holds for the finite bias case . the real part of the correlation function according to the population difference @xmath127 determines the correlation function of the stochastic force @xmath128(t)\delta\tilde \sigma_z[x](t ' ) \rangle}}\label{eq.:dqd_stoch_forces_average } , \end{aligned}\ ] ] with fluctuation @xmath129(t)=\tilde\sigma_{z}[x](t)-{\ensuremath{\langle \tilde\sigma_{z}[x](t ) \rangle}}$ ] . the effective potential determines the behavior of the oscillator trajectories in the phase space ( @xmath130@xmath65plane ) , as seen in section [ subsec.:ahm_eff_pot ] for the anderson holstein model . in the following we examine the differential equation of the system by studying its fixed points @xcite . some further theoretical details of this kind of investigation are explained in appendix [ sec.:dqd_fix_point ] . for the double dot we obtain the dynamical system @xmath131(t ) \right].\quad \end{aligned}\ ] ] fixed points occur under the condition @xmath132 , i.e. @xmath133 and following from that , the fixed points position coordinates are equal to the roots of the effective force @xmath134 the jacobian matrix is obtained from @xmath135 & - a[{\ensuremath{\langle x^{\ast } \rangle } } ] \\ \end{array } \right)$ ] evaluated at the fixed point @xmath136 , whereby @xmath137 . + determinant @xmath138 and trace @xmath139 become @xmath140 , \hspace{0.3 cm } \tau = - a[{\ensuremath{\langle x^{\ast } \rangle}}].\ ] ] the trace decides about the stability of a fixed point and is equal to the negative friction here . for the case without friction the trace is equal to zero , therefore only centers occur in the phase plane . in the case with friction the trace can be either positive or negative leading to both , stable and unstable fixed points . for comparison see figure [ fig.:dqd_phase_space_comparison ] , where row a depicts the results of a fixed point analysis . the effective force @xmath61 is plotted in the upper part of each plot in row a together with trace and determinant . therewith the characteristics of the fixed points are determined and it is possible to predict the shape of the phase space portrait . in the lower parts of the plots in row a these predictions are illustrated , fixed points are marked by black lines . for small tunnel coupling @xmath141 ( diagram a1 ) we obtain seven fixed points . these can be characterised as three stable and one unstable spirals , each seperated by one of the three saddle points . stable spirals correspond to the different rest positions for the oscillator and the saddle points to the points of instable balance . increasing the tunnel coupling leads to five fixed points . in graph a2 we obtain a stable spiral in the middle enclosed by two saddle points and followed by a stable spiral on each side . in the diagrams a3 and a4 the mean point changes to an unstable spiral . with further increased tunnel coupling ( a3 ) there remain only three fixed points and for even higher values of @xmath141 ( a4 ) the number of zeros in the effective force reduces to one . for the oscillator the latter means that it is shifted to a new rest position independent from its inital position . in figure [ fig.:dqd_phase_space_comparison ] the rows b and c depict phase space portraits for the double dot system without and with friction . the four columns correspond to four different values of the tunnel coupling @xmath141 . as initial condition , the momentum was set to zero and the positions were chosen in order to show the various shapes of the trajectories . in the case without friction we recognize periodic cycles which are stable and run around one or more fixed points . these centers were also expected from the analysis in section [ subsec.:dd_fix ] . the fixed points correspond to certain states of the double dot system . if the left dot is occupied the rest position of the oscillator is shifted to the right and correspondingly to the left for an occupied right dot . these points turn to stable spirals when the friction is turned on . the states when both dots are occupied or empty correspond to the fixed point in the middle . the lowermost row shows what happens when we include the friction in our calculations . in contrast to the single dot system , here the friction has positive as well as negative values depending on the position of the oscillator . this means that the oscillator is either decelerated or accelerated . both can be interpreted as inelastic jumps of the electrons , where energy is transfered between the electrons and the oscillator in both directions like it has been observed in @xcite . there , the authors consider a resonator coupled to a superconducting single electron transistor ( sset ) . as a result of the interplay of positive and negative damping in a certain parameter range they observed limit cycles and bistability in the phase plane , in our work we obtain a likewise behaviour for the oscillator . in contrast to our semiclassical approximation they investigate the wigner function of the system with numerical master equations . they compare these results to a mean field evaluation of the expectation value of the oscillator position @xcite . for weak coupling the mean field approach gives quantitatively correct results , and for higher coupling it still describes the dynamics qualitatively correct . these results suggest that our use of average oscillator positions and momenta is qualitatively correct for a description of the oscillator instabilities . consider again row c in figure [ fig.:dqd_phase_space_comparison ] , where the results for the dynamics of the oscillator with friction are plotted . the outer left and right stable spiral do not change by increasing the tunnel coupling @xmath141 . by contrast the oscillator s behaviour between these stable rest positions changes a lot . in graph c1 we observe a stable and an unstable spiral , like we expect from the fixed point analysis . in the neighbourhood of the unstable fixed point the friction is negative , so the oscillator trajectory is repelled and ends up in the right stable spiral . in diagram c2 we recognize that the latter path becomes stable . the limit cycle appears when the unstable spiral has disappeared and exists as long as the tunneling coupling is in the range of @xmath142 . there we observe a bistability : as the initial position gets closer to the fixed point the limit cycle turns into a stable spiral . by further increasing @xmath141 , the middle spiral becomes unstable and a second limit cycle appears ( c3 ) . this limit cycle with a smaller radius exists in the range of @xmath143 . for @xmath144 the system undergoes a hopf bifurcation @xcite , which happens when a pair of complex eigenvalues from the dynamical system , which determine the evolution in the phase plane , see section [ sec.:dqd_fix_point ] , cross the imaginary axis from the left to the right half - plane . in other words , the trace @xmath139 changes its sign and at the bifurcation point the eigenvalues are purely imaginary @xmath145 , see equation . in the last graph of row c both limit cycles have disappeared and the oscillators path ends up in the left or right stable point . in our calculations we choose @xmath146 , standing in some contrast to our adiabatic approach , which implies a slow oscillator ( @xmath147 ) . if @xmath148 the interaction between the current and the oscillator is strongest @xcite , because both act on the same timescale . interesting effects still appear with a slightly enlarged @xmath149 like in our plots , but for @xmath147 there remains only one stable spiral . this means , that the oscillator rest position is shifted from its ground position caused by the stochastic processes initiated by the current . we presume that our approach is useful also for a comparative fast oscillator and we will accomplish further investigations with a non - adiabatic approximation to reconsider our results . we have derived a stochastic equation of motion that describes the dynamics of a single oscillator coupled to an electronic environment out of equilibrium . we studied two cases , namely the single dot level and double dot ( two level ) electronic system . for both cases we have explained the features of effective potential and friction for the ensemble averaged oscillator motion . the effects we recognize fit well together with former works . in the dqd model limit cycles and bistabilities appear . until now the master equation has been used for most investigations of the oscillator behaviour in nems . we have used a method that gives us access to regions where the master equation has problems : we naturally include finite bias , and arbitrary electron coupling to external reservoirs . we had to stay in a regime with a relatively fast oscillator in order not to miss the interesting physical effects . the validity of our method in this regime could still be improved with a non - adiabatic calculation . this work was supported by project dfg br 1528/5 - 2 , the we heraeus foundation and the rosa luxemburg foundation . we acknowledge discussions with c. emary . the reduced density matrix @xmath150 , obtained from the total density matrix @xmath11 by tracing out the bath degrees of freedom , describes the mechanic subsystem @xmath151 , cf . . by using a factorising initial condition @xmath152 the elements of the reduced density matrix can be expressed in a path integral representation@xcite @xmath153(\tau)\label{eq.:red_dense_matrix_path_rep}. \end{aligned}\ ] ] hereby @xmath154 $ ] denotes the classical action in the path integral for eq . and the feynman vernon influence functional is defined by @xmath155(t_0,t)={\ensuremath{\operatorname{tr}}}_{\rm b}\big\{\hat u^\dag[q'](t , t_0)\hat u[q](t , t_0)\rho_{\rm b}\big\}. \end{aligned}\ ] ] the time evolution operators are defined by @xmath156(t , t_0)={{\ensuremath{\operatorname{t}_{\leftarrow}}}}\exp\big[-i\int_{t_0}^{t}dt'\ ; \mathcal{h}_{\rm res}[q](t')\big ] \end{aligned}\ ] ] corresponding to the reservoirs @xmath157(t)=\mathcal{h}_{\rm e}-\hat f q_t$ ] . in the next step we establish an effective interaction picture by @xmath158(t , t_0 ) & = e^{i\mathcal{h}_0(t - t_0 ) } \ ; \hat u[q](t , t_0)\nonumber\\ & = { { \ensuremath{\operatorname{t}_{\leftarrow}}}}\exp\big[-i\int_{t_0}^{t}dt'\;\tilde v[q](t')\big],\nonumber\\ \tilde v[q](t)&= e^{i\mathcal{h}_0(t - t_0 ) } v[q](t ) e^{-i\mathcal{h}_0(t - t_0 ) } \label{eq.:gen_time_evol}. \end{aligned}\ ] ] hereby the reservoir is decomposed in an unperturbed part @xmath159 and the perturbation @xmath160(t)= -\hat f(t{\dot x}_0+\frac12 y_t)$ ] and @xmath161(t)= -\hat f(t{\dot x}_0-\frac12 y_t)$ ] , where we wrote @xmath162 using the adiabatic approximation . by using eq . , the influence functional can be expanded to second order in perturbation theory . denoting @xmath163 , @xmath164(t_1)$ ] and @xmath165(t_1)$ ] , the feynman vernon influence functional reads @xmath166(t_0,t ) = { \ensuremath{\langle \tilde u[q'](t , t_0)\tilde u[q](t , t_0 ) \rangle}}\nonumber\\[1ex ] & = { \ensuremath{\langle \big[{{\ensuremath{\operatorname{t}_{\rightarrow}}}}e^{i\int_{t_0}^{t}dt_1\;\tilde v'_1}\big ] \big[{{\ensuremath{\operatorname{t}_{\leftarrow}}}}e^{-i\int_{t_0}^{t}dt_1\;\tilde v_1}\big ] \rangle}}\nonumber\\ & = 1 + i\int_{t_0}^{t}dt_1\;{\ensuremath{\langle \tilde v'_1-\tilde v_1 \rangle}}\nonumber\\ & \hspace{2ex}+\int_{t_0}^{t}dt_1\int_{t_0}^{t_1}dt_2\ ; { \ensuremath{\langle ( \tilde v'_1-\tilde v_1)\tilde v_2-\tilde v'_2(\tilde v'_1-\tilde v_1 ) \rangle}}. \end{aligned}\ ] ] when plugging in the definition of the perturbation ( interaction picture ) and taking into account that the force operator @xmath7 is hermitian , we obtain @xmath167(t_0,t ) = 1 + i\int_{t_0}^{t}dt_1\;y_{t_1}\nonumber\\ & \hspace{2ex}\times \bigg[{\ensuremath{\langle \tilde f(t_1 ) \rangle } } -2\dot x_0 \int_{t_0}^{t_1}dt_2\;t_2 { \ensuremath{\operatorname{im}}}{\ensuremath{\langle \tilde f(t_1)\tilde f(t_2 ) \rangle}}\bigg]\nonumber\\ & \hspace{2ex}-\int_{t_0}^{t}dt_1\int_{t_0}^{t_1}dt_2\;y_{t_1 } { \ensuremath{\operatorname{re}}}{\ensuremath{\langle \tilde f(t_1)\tilde f(t_2 ) \rangle}}y_{t_2 } \label{eq.:gen_influece_functional}. \end{aligned}\ ] ] performing a cluster expansion @xcite , the influence functional can be expressed in terms of the influence phase @xmath168(t_0,t)\equiv\nonumber\\ & \hspace{2ex}-\ln\mathcal{f}[x+y/2,x - y/2](t_0,t ) = - i\int_{t_0}^{t}dt_1\;y_{t_1}\nonumber\\ & \hspace{2ex}\times \bigg[{\ensuremath{\langle \tilde f(t_1 ) \rangle } } -2\dot x_0 \int_{t_0}^{t_1}dt_2\;t_2 { \ensuremath{\operatorname{im}}}{\ensuremath{\langle \delta\tilde f(t_1)\delta\tilde f(t_2 ) \rangle}}\bigg]\nonumber\\ & \hspace{2ex}+\int_{t_0}^{t}dt_1\int_{t_0}^{t_1}dt_2\;y_{t_1 } { \ensuremath{\operatorname{re}}}{\ensuremath{\langle \delta\tilde f(t_1)\delta\tilde f(t_2 ) \rangle}}y_{t_2 } , \end{aligned}\ ] ] whereby @xmath169 denotes the fluctuation around @xmath170 . this equation can be easily verified by expanding the exponential @xmath171(t_0,t)]$ ] to second order in perturbation theory and comparing with eq . . together with the classical action , we define an effective action functional by @xmath172 ( t_0,t ) { \ensuremath{\mathrel{\mathop:}=}}s_{x+y/2 } - s_{x - y/2 } + i\phi[x , y](t_0,t)\nonumber\\ & = \int_{t_0}^t dt_1\ ; \big [ m \dot x_1 \dot y_1 -v_{\rm osc}(x_1+\frac{y_1}{2})+v_{\rm osc}(x_1-\frac{y_1}{2 } ) \big]\nonumber\\ & \hspace{2ex}+ i\phi[x , y](t_0,t ) , \end{aligned}\ ] ] whereby the reduced density matrix expressed in terms of the new variables , eq . , reads @xmath173 ( t_0,t)}. \end{aligned}\ ] ] in a semiclassical approach we assume small deviations of the off diagonal trajectories @xmath174 from the diagonal ones . thus the potential difference leads in second order in @xmath174 to @xmath175 this approximation is exact for quadratic potentials , as we treat in this paper . furthermore , with the boundary conditions @xmath176 and integration by parts the effective action functional is quadratic in @xmath174 , @xmath172 ( t_0,t ) = -\int_{t_0}^t dt_1\;y_1 \bigg [ m \ddot x_1 + v'_{\rm osc}(x_1 ) -{\ensuremath{\langle \tilde f(t_1 ) \rangle } } \nonumber\\ & \hspace{2ex}+2\dot x_0 \int_{t_0}^{t_1}dt_2\;t_2{\ensuremath{\operatorname{im}}}{\ensuremath{\langle \delta\tilde f(t_1)\delta\tilde f(t_2 ) \rangle } } \bigg]\nonumber\\ & \hspace{2ex}+ i\int_{t_0}^{t}dt_1\int_{t_0}^{t_1}dt_2\;y_1 { \ensuremath{\operatorname{re}}}{\ensuremath{\langle \delta\tilde f(t_1)\delta\tilde f(t_2 ) \rangle}}y_2]\nonumber\\ & \equiv -\int_{t_0}^t dt_1\;y_1 k_1[x ] + \frac{i}{2}\int_{t_0}^{t}dt_1\int_{t_0}^{t_1}dt_2\;y_1 l_{1,2}[x]y_2 \label{eq.:gen_action_functional}. \end{aligned}\ ] ] completing the square of the integral kernel @xcite @xmath177\}$ ] , the resulting path integral describes a stochastic process with langevin equation @xmath178 & = m \ddot x_t + v'_{\rm osc}(x_t ) -{\ensuremath{\langle \tilde f(t ) \rangle}}\nonumber\\ & \hspace{2ex } + 2\dot x_0 \int_{t_0}^{t}dt'\;t'\;{\ensuremath{\operatorname{im}}}{\ensuremath{\langle \delta\tilde f(t)\delta\tilde f(t ' ) \rangle } } = \xi_t \end{aligned}\ ] ] with @xmath26 a gaussian stochastic force . to obtain a selfconsistent equation of motion , we finally replace @xmath179 by @xmath17 . this substitution concerns also the unperturbed hamiltonian which leads to @xmath180 . the langevin equation then reads @xmath181 the term quadratic in @xmath174 in eq . determines the correlation function of the stochastic force , @xmath182 the single dot lesser green s function without coupling in energy space is@xcite @xmath183 here we have used the adiabatic approximation for the center of mass coordinate , thus the coupling to the oscillator simply shifts the level @xmath36 by @xmath184 , so we have to replace @xmath36 by the shifted energy @xmath185 . the greater green s function is correspondingly @xmath186+\gamma_{\rm r}[1-f_{\rm r}(\omega ) ] } { ( \omega-\tilde\varepsilon)^2+\gamma^2/4}\nonumber\\ & = g^<(\omega ) -i\frac{\gamma}{(\omega-\tilde\varepsilon)^2+\gamma^2/4}. \end{aligned}\ ] ] hereby @xmath187 $ ] denote fermi functions with lead index @xmath111 , inverse temperature @xmath83 and chemical potential @xmath86 . the time - dependent green s functions are obtained by fourier transformation as @xmath188 . in the zero temperature limit we have to replace @xmath189 . then for @xmath190 an integration leads to @xmath191{\ensuremath{|t|}}\big\}{\ensuremath{\operatorname{sgn}}}(t)\nonumber\\ & \hspace{2ex}+e^{-\gamma/2{\ensuremath{|t|}}}{\ensuremath{\operatorname{e_1}}}\big\{[i\omega_\alpha{\ensuremath{\operatorname{sgn}}}(t)-\gamma/2]{\ensuremath{|t|}}\big\}{\ensuremath{\operatorname{sgn}}}(t)\nonumber\\ & \hspace{2ex}\pm 2\pi ie^{-\gamma/2{\ensuremath{|t|}}}\theta(\pm\omega_\alpha ) \bigg]\bigg|_{\omega_\alpha=\mu_\alpha-\varepsilon_x } , \end{aligned}\ ] ] where the first exponential integral is defined by @xmath192 for the case @xmath193 the lesser greens function reads @xmath194 . \end{aligned}\ ] ] for finite temperature one has to regard fermi functions instead of the heaviside theta function . by virtue of the residue theorem one finds @xmath195 with the digamma function @xmath88 . the green s functions in the frequency domain according to the hamiltonian in equation ( [ eq.:h_dqd ] ) is derived via the equation of motion method . there the green s function @xmath196 is defined as resolvent of the hamiltonian @xmath197 via @xmath198 denoting the bath states with @xmath199 the previous equation yields @xmath200 taking the matrix elements and inserting the hamiltonian leads to a set of equations , from whom the green s functions are derived . + here , the electron - phonon coupling is described adiabatically , so the interaction part @xmath201 came in by shifting the dot level energies @xmath202 with @xmath203 . finally the dot green s function is given through @xmath204 with the elements @xmath205 \left[\omega - \widetilde{\nu}_{\rm{r } } - \sigma_{\rm{r}}(\omega)\right ] - \left|t_{c}\right|^{2 } } \nonumber \\ g_{\rm{rr}}(\omega ) & = & \frac{\omega - \widetilde{\nu}_{\rm{l } } - \sigma_{\rm{l}}(\omega)}{\left[\omega - \widetilde{\nu}_{\rm{l } } - \sigma_{\rm{l}}(\omega)\right ] \left[\omega - \widetilde{\nu}_{\rm{r } } - \sigma_{\rm{r}}(\omega)\right ] - \left|t_{c}\right|^{2 } } \nonumber \\ g_{\rm{lr}}(\omega ) & = & \frac{t_{c}}{\omega - \widetilde{\nu}_{\rm{l } } - \sigma_{\rm{l}}(\omega ) } \ g_{\rm{rr}}(\omega ) \nonumber \\ g_{\rm{rl}}(\omega ) & = & \frac{t^{\ast}_{c}}{\omega - \widetilde{\nu}_{\rm{r } } - \sigma_{\rm{r}}(\omega ) } \ g_{\rm{ll}}(\omega ) \end{aligned}\ ] ] hereby we introduced the self energy @xmath206 corresponding to the left or the right dot with @xmath207 @xmath208 is the undisturbed green s function for the leads . we derive the associated retarded and the advanced green s function by using the continuation rules @xmath209 the derivation of the lesser / greater green s function is taking usage of the keldysh equation : @xmath210 whereas we assume both dots initially unoccupied . the lesser / greater self energy follows from @xmath211 . \end{aligned}\ ] ] the friction is determined by the imaginary part of the correlation function of the double dot . from equation , we start with expressing the correlation function through green s functions and use their fourier transforms : @xmath212(t ) \nonumber \\ & = 2 \lambda^{2 } \ \sum_{\alpha,\beta } \ \left[2 \delta_{\alpha,\beta } -1\right ] \ \int dt ' \ t ' \ \frac{1}{2i } \ \ * \nonumber \\ & \hspace{0.5 cm } \left [ g_{\beta \alpha}^{<}(t'-t ) g_{\alpha \beta}^{>}(t - t ' ) - g_{\beta \alpha}^{<}(t - t ' ) g_{\alpha \beta}^{>}(t'-t ) \right ] \nonumber \\ & = \frac{\lambda^{2}}{\pi } { \ensuremath{\operatorname{im}}}\sum_{\alpha,\beta } \left[2 \delta_{\alpha,\beta } -1\right ] \int d\omega_{1 } \int d\omega_{2 } g_{\beta \alpha}^{<}(\omega_{1 } ) * \nonumber \\ & \hspace{0.5 cm } g_{\alpha \beta}^{>}(\omega_{2 } ) e^{i ( \omega_{1 } - \omega_{2 } ) t } ( -i ) \frac{\partial}{\partial \omega_{2 } } \int \frac{dt'}{2\pi } e^{-i(\omega_{1}-\omega_{2 } ) t ' } \nonumber \\ & = \frac{\lambda^{2}}{\pi } \sum_{\alpha,\beta } \left[2 \delta_{\alpha,\beta } -1\right ] \int d\omega g^{<}_{\beta \alpha}(\omega ) \frac{\partial}{\partial \omega } g^{>}_{\alpha \beta}(\omega ) . \end{aligned}\ ] ] in the last step a term was identified as the derivative of dirac s delta , this result agrees with the solution in the work by mozyrsky et al.@xcite . + the lesser / greater green s function keldysh equation follow from the keldysh equation and we consider the tunneling rates to be frequency independent , therefore the self energies are ( @xmath213 ) : @xmath214 because of the lengthy calculation , we just outline the calculation for the ll - term @xmath215 . the other terms ( @xmath216 ) can be derived in a similar way . here , the product of retarded and advanced green s functions equates the squared modulus of @xmath217 , so we can write @xmath218 for simplicity in the following we omit the superscript of the retarded greens function and abbreviate the moduli by @xmath219 we obtain for the first term of the friction @xmath220 \nonumber \\ & = & \frac{\gamma^{2}}{4 } \int d\omega \bigg [ { \ensuremath{|g_{\rm{ll}}(\omega)|}}^{2 } \theta(\mu_{\rm{l } } - \omega ) + { \ensuremath{|g_{\rm{lr}}(\omega)|}}^{2 } \theta(\mu_{\rm{r } } - \omega ) \bigg ] * \nonumber \\ & & \bigg [ \left [ \frac{\partial}{\partial \omega } { \ensuremath{|g_{\rm{ll}}(\omega)|}}^{2 } \right ] \theta(\omega - \mu_{\rm{l } } ) + { \ensuremath{|g_{\rm{ll}}(\omega)|}}^2 \delta(\omega - \mu_{\rm{l } } ) + \left[\frac{\partial}{\partial \omega } { \ensuremath{|g_{\rm{lr}}(\omega)|}}^{2 } \right ] \theta(\omega - \mu_{\rm{r } } ) { \ensuremath{|g_{\rm{lr}}(\omega)|}}^{2 } \delta(\omega - \mu_{\rm{r } } ) \bigg ] . \nonumber \\ \end{aligned}\ ] ] some parts of the integration terms can directly be evaluated with the help of the delta and heavyside functions . by choosing the condition @xmath221 , we get rid of a case distinction , which would be necessary for two product terms which included a delta - function . after performing also an integration by parts , we arrive at @xmath222 \bigg ] . \end{aligned}\ ] ] in the following the green s functions , derived in section [ sec.:dqd_green ] , were inserted , whereas we assume equal tunneling rates for the left and the right side , @xmath223 . then a number of integrations by parts is performed to dispose of the derivation in the integral term . so we obtain a closed expression , whereas @xmath224 abbreviates the denominator of the green s function @xmath225^{2}}{n(\mu_{\rm{l}})^{2 } } + \frac{\left| t_{c } \right|^{4}}{n(\mu_{\rm{r}})^{2 } } + \nonumber \\ & & \hspace{-0.5 cm } 2 \left| t_{c } \right|^{2 } \frac{\left[(\mu_{\rm{r } } - \widetilde{\nu}_{\rm{r}})^{2 } + \frac{\gamma^{2}}{16}\right]}{n(\mu_{\rm{r}})^{2 } } - 4 \left| t_{c } \right|^{2 } \int\limits_{\mu_{\rm{r}}}^{\mu_{\rm{l } } } d\omega \frac{(\omega - \widetilde{\nu}_{\rm{r}})}{n(\omega)^{2 } } \bigg ] . \nonumber \\ \end{aligned}\ ] ] in an analogue way the other terms in can be derived , and finally the solution for the friction is @xmath226(t ) & = & \frac{\lambda^{2}}{\pi } \frac{\gamma^{2}}{8 } \bigg [ \frac{\left[(\mu_{\rm{l } } - \widetilde{\nu}_{\rm{r}})^{2 } + \frac{\gamma^{2}}{16}\right]^{2}}{n(\mu_{\rm{l}})^{2 } } + \frac{\left[(\mu_{\rm{r } } - \widetilde{\nu}_{\rm{l}})^{2 } + \frac{\gamma^{2}}{16}\right]^{2}}{n(\mu_{\rm{r}})^{2 } } + \frac{\left| t_{c } \right|^{4}}{n(\mu_{\rm{l}})^{2 } } + \frac{\left| t_{c } \right|^{4}}{n(\mu_{\rm{r}})^{2 } } + 2 \left| t_{c } \right|^{2 } \frac{\left[(\mu_{\rm{r } } - \widetilde{\nu}_{\rm{r}})^{2 } + \frac{\gamma^{2}}{16}\right]}{n(\mu_{\rm{r}})^{2 } } \nonumber \\ & & - 2 \left| t_{c } \right|^{2 } \frac{\left[(\mu_{\rm{l } } - \widetilde{\nu}_{\rm{r}})^{2 } + \frac{\gamma^{2}}{16}\right]}{n(\mu_{\rm{l}})^{2 } } - 4 \left| t_{c } \right|^{2 } \frac{\left[(\mu_{\rm{r } } - \widetilde{\nu}_{\rm{r } } ) ( \mu_{\rm{r } } -\widetilde{\nu}_{\rm{l } } ) - \frac{\gamma^{2}}{16}\right]}{n(\mu_{\rm{r}})^{2 } } - 8 \left| t_{c } \right|^{2 } ( \widetilde{\nu}_{\rm{l } } -\widetilde{\nu}_{\rm{r } } ) \int\limits_{\mu_{\rm{r}}}^{\mu_{\rm{l } } } d\omega \ \frac{1}{n(\omega)^{2 } } \bigg].\nonumber \\ \end{aligned}\ ] ] with the assumption @xmath227 we get : @xmath228 and the integral in equation is given trough @xmath229 in section [ sec.:dqd_aux_int ] . the fixed points of a nonlinear two dimensional system can be investigated with standard methods for linear dynamical system @xcite . + the general solution for a two dimensional linear system @xmath230 is @xmath231 and so determined by the eigenvalues @xmath232 of the matrix a. the constants @xmath233 depend on the initial conditions and @xmath234 are eigenvectors . + the eigenvalues can be obtained from @xmath235 thereby @xmath139 corresponds to the trace and @xmath138 to the determinant of a. these two qualities determine the evolution of the trajectories in the phase plane . + for a fixed point @xmath236 the condition @xmath237 must be fulfilled . various classes of fixed points exist , whereas the determinant @xmath138 decides which kind of point appears . in case of saddle points the determinant is negative and it is positive for spirals or nodes . the difference between a spiral and a node is that for the second one the eigenvalues have no imaginary part . + the trace @xmath139 defines the stability of nodes and spirals , this is caused by the fact that @xmath139 determines the sign of the eigenvalue s real part , for instance with negative real part decaying oscillations occur and the fixed point is stable , see equation . there also exist some borderline cases , whereas the centers are the most significant ones , they occur when the trace is equal to zero . + this analysis can be assigned to a two dimensional nonlinear system @xmath238 . by assuming a small disturbance @xmath239 from a fixed point , we can invest if this disturbance grows or decays by performing a taylor expansion @xmath240 the first term is zero and higher terms ( @xmath241 ) can be neglected because the disturbance is small . so we get a linearised system @xmath242 , containing the jacobi matrix @xmath243 evaluated at the fixed point coordinates . the above explained analysis can be performed for this system . this is valid as long no borderline cases occur , then the higher terms may be more important . @xmath244 , \nonumber\\[2ex ] \mathcal i_2(\mu , a , b ) & = \int_{-\infty}^\mu d\omega \frac{\omega}{\omega^4 + 2 a\omega^2+b^2 } \nonumber\\ & = \frac{1}{2\sqrt{a^2-b^2 } } \bigg [ + \frac12 \ln\big(\mu^2 + a-\sqrt{a^2-b^2}\big ) \nonumber\\ & \hspace{16ex}-\frac12 \ln\big(\mu^2 + a+\sqrt{a^2-b^2}\big ) \bigg ] , \nonumber\\[2ex ] \mathcal i_3(\mu , a , b ) & = \int_{-\infty}^\mu d\omega \frac{\omega^2}{\omega^4 + 2 a\omega^2+b^2 } \nonumber\\ & = \frac{1}{2\sqrt{a^2-b^2 } } \bigg [ -\sqrt{a-\sqrt{a^2-b^2 } } \arctan\big(\frac{\mu}{\sqrt{a-\sqrt{a^2-b^2}}}\big ) \nonumber\\ & \hspace{16ex}+\sqrt{a+\sqrt{a^2-b^2 } } \arctan\big(\frac{\mu}{\sqrt{a+\sqrt{a^2-b^2}}}\big ) \nonumber\\ & \hspace{16ex}-\frac\pi2\bigg ( \frac{1}{\sqrt{\frac{1}{a-\sqrt{a^2-b^2 } } } } -\frac{1}{\sqrt{\frac{1}{a+\sqrt{a^2-b^2 } } } } \bigg ) \bigg ] , \nonumber\\[2ex ] \mathcal i_4(a , b ) & = \int d\omega \frac{1}{\left[\omega^{4 } + 2 a \omega^{2 } + b^{2 } \right]^{2 } } \nonumber\\ & = \frac{1}{8 b^{2}(b^{2}-a^{2 } ) } \bigg [ \frac{2 \omega ( b^{2}-2a^{2}-a \omega^{2})}{b^{2}+2a\omega^{2}+\omega^{4 } } \nonumber \\ & \hspace{16ex } -\frac{(a^{2}-3b^{2}+a \sqrt{a^{2}-b^{2 } ) } \arctan \frac{\omega}{\sqrt{a-\sqrt{a^{2}-b^{2 } } } } } { \sqrt{a^{2}-b^{2 } } \ \sqrt{a-\sqrt{a^{2}-b^{2 } } } } \nonumber \\ & \hspace{16ex } + \frac{(a^{2}-3b^{2 } - a \sqrt{a^{2}-b^{2 } ) } \arctan \frac{\omega}{\sqrt{a+\sqrt{a^{2}-b^{2 } } } } } { \sqrt{a^{2}-b^{2 } } \ \sqrt{a+\sqrt{a^{2}-b^{2 } } } } \bigg ] . \label{eq.aux_definite_integrals } \end{aligned}\ ] ]

there is a highly - compact cluster of blue stars @xcite at the heart of the m31 double nuclear - cluster of old stars . @xcite showed that this blue cluster in turn hosts the @xmath4 super - massive black hole also known to reside in the nucleus @xcite . @xcite argue that the integrated spectrum and spectral - energy distribution of the stars are consistent with their formation in a burst 200 myr ago . young stars are also closely bound to the nuclear black hole in our galaxy @xcite , suggesting that this may not be a rare phenomenon . understanding how apparently young stars were formed deeply interior to the tidal field of a supermassive black hole probes several issues of how stars and gas interact within such an extreme environment ( see @xcite for a review ) . study of the blue cluster in m31 provides an additional context in which to test theories for the formation of the unusual population surrounding the milky way black hole . there is a diversity of ideas for explaining the origin of the m31 blue cluster . @xcite , for example , argue that the cluster is the expected consequence of the non - axisymmetric structure of the surrounding m31 nucleus funneling gas from stellar mass - loss into orbit around the black hole , where it periodically reaches surface densities sufficient to induce collapse and star formation . likewise , @xcite , @xcite , and @xcite have advanced the formation of compact disks of young stars around the milky way black hole by massive central accretion of gas . in contrast , @xcite argue that the blue stars in m31 result from stellar collisions and tidal stripping acting on an old population of stars interacting within the high - velocity orbits and strong tidal - field associated with the black hole . we present new observations of the blue cluster obtained at the maximum angular resolution offered by the _ hubble space telescope _ to provide better constraints on the origin of the stars residing in this unusual environment . the discovery of the blue cluster proceeded in stages . @xcite imaged the nucleus of m31 with the hst faint object camera at 1750 , seeing a double structure similar to that discovered in the optical by @xcite . by the simple expedient of checking the position - angle of their images , however , they discovered that the optically - dimmer peak , designated p2 by lauer et al . , was actually brighter in the uv than p1 . @xcite lacked @xmath5 sufficient to resolve the spatial structure of the uv source at p2 , noting only that it was highly - compact and possibly consistent with a point source , such as low - level emission associated with the weak agn detected in the radio @xcite , or emission from a single pagb star , such as those seen further out from the nucleus @xcite . in retrospect , the discovery of a uv - bright nuclear source explains the observations of @xcite , who found the nucleus to have color gradient becoming _ bluer _ towards its center , using cfht images obtained at 3750 . discovery of the cluster may also have been presaged by an enigmatic reference in @xcite to third - party unpublished `` mt . wilson material '' purporting to show that `` the spectrum of the nucleus is of a peculiar dwarf a type , '' a result supported by no other work prior to @xcite to the best of our literature research . later hst observations by @xcite and @xcite were able to resolve the uv - source ; both papers argued that it is a cluster of stars . wfpc2/pc images at 3000 obtained by @xcite showed that the half - power radius of the blue source is @xmath6 pc . @xcite further combined their u and v band fluxes of the source with the @xcite fluxes at 1750 to conclude that the source is consistent with an a - star spectral energy distribution . @xcite refined this picture further by obtaining hst stis spectroscopy of the nucleus over @xmath7 , and reanalyzing the f300w images presented in @xcite . the spectroscopy shows strong balmer lines , including a strong balmer break , consistent with either a0 main - sequence or giant stars , but not a population of white dwarfs of the same photospheric temperature . the best match to the spectrum is provided by a @xmath8 myr population formed in a single burst . populations of less than half this age would exhibit too much uv - flux to the blue side of the balmer break . @xcite further show that the stars are distributed in a flat disk with an exponential profile of scale - length of @xmath9 pc in rapid rotation about the black hole . the rotational broadening of the stis spectrum implies a black hole mass of @xmath10 a significant upwards revision from earlier determinations . the m31 black hole and its cluster of blue stars mark the center of a much larger nuclear star - cluster of complex structure . the surface brightness of the nuclear stellar system begins to rise above the underlying m31 bulge interior to @xmath11 or @xmath12 pc from the center . the cluster becomes increasingly elliptical at smaller radii , with @xmath13 for @xmath14 but its isophotes remain concentric with the photo - center of the bulge for @xmath15 @xcite . at yet smaller radii the nuclear cluster exhibits the double - peaked structure discovered by @xcite . the optically - brighter peak , p1 , is offset by @xmath16 from the secondary brightness peak , p2 , and a slightly lesser amount from the m31 bulge photo - center . the best explanation for the double morphology is that both peaks result from the line - of - sight projection of an eccentric disk of stars roughly @xmath17 pc in diameter bound to the black hole @xcite . the black hole strongly dominates the potential , thus the stars in the disk follow keplerian orbits , spending most of their time at the orbital apo - center , creating an enhancement of light at p1 . significantly , the p1p2 line twists away from the major axis of the outer nucleus by @xmath18 p1 and p2 are both redder in @xmath19 than the surrounding nucleus @xcite . high spatial - resolution spectroscopy shows that both p1 and p2 comprise old stellar populations with characteristics more like each other , than the underlying bulge @xcite . the `` tremaine disk '' thus appears to be a distinct component embedded in a much larger nuclear - cluster . the initial model of the disk provided by @xcite has been refined by the more detailed analysis presented by @xcite . an interesting feature of the refined model is that the disk requires a central hole in order to generate the apparent minimum of stellar emission that occurs between p1 and p2 . the cluster of blue stars occurs at still smaller radii , @xmath20 and must be closely bound to the black hole , which was first inferred to exist from ground - based spectroscopy @xcite . @xcite carefully registered their high spatial - resolution spectroscopy to the wfpc2 imagery of @xcite , concluding that the m31 black hole was coincident with the cluster and that both were within @xmath21 of the m31 bulge photo - center . the conclusion that the cluster hosts the black hole was further confirmed by @xcite , who used stis spectroscopy to show that the central velocity dispersion in the cluster rises to @xmath22 with organized rotation peaking at @xmath23 building on the analysis of @xcite , they reduced the offset of the black hole from the bulge center to @xmath24 in the `` anti - p1 '' direction along the p1-p2 line . they further demonstrated that this slight offset is balanced by the asymmetry of the tremaine disk the center of mass of the complete nuclear system is coincident with the bulge photo - center to @xmath25 @xcite also designated the cluster as a third component , p3 , of the inner nucleus . this designation includes a subtle redefinition of p2 as it previously had been introduced by @xcite and used by @xcite . the center of the cluster is coincident with the peak of a shallow surface - brightness cusp seen in v - band that falls within @xmath26 of the bulge center ; it was this location that the two lauer et al . papers adopted as the center of p2 . @xcite , however , use p2 to denote the elongated and more diffuse concentration of older stars that extends from the cluster on the side opposite from p1 . in this schema , p1 and p2 explicitly correspond to the apo- and peri - center apses of the tremaine disk , which has a mean peri - center that is _ not _ coincident with the cluster . the present environment of p3 appears to be largely quiescent , despite the immediate proximity of the black hole ( see the extensive discussion of this topic by * ? ? ? @xcite find a weak radio source coincident with p3 , and @xcite find low - level x - ray emission associated with p3 , as well ( however , see * ? ? ? on the other hand , a high spatial - resolution map of emission from ionized gas within in the nucleus finds a few sources within an arcsecond of p1 , but nothing coincident with p2 or p3 @xcite . while central accretion of gas may have created the p3 stars in the past , there is no evidence that any significant reservoir of cold gas exists there at present . under the @xcite description of p3 , the cluster is a disk comprising only a few hundred a - stars . given the scale length of the disk , we concluded that it should begin to show resolution into discrete sources in images with slightly higher angular resolution than were obtained by @xcite , such as could be provided by the hst acs / hrc in the blue . the photometry of the brighter sources might be obtained directly , but in any case , the full image of the cluster could also be analyzed by surface - brightness fluctuation ( sbf ) analysis @xcite to constrain the p3 stellar population . if the cluster indeed comprises main sequence a stars , it would look `` lumpy '' in hrc images . sbf could quantify the amplitude of the lumpiness , and in turn the typical luminosity of the stars that the cluster comprises . we obtained images of the m31 nucleus under go - program 10572 , using the hst acs high resolution channel ( hrc ) , which provided the best angular resolution available on the telescope . images were obtained in filters f330w ( @xmath0 ) ' to avoid confusing it with the standard johnson u - filter , which has a significantly different passband . all magnitudes are on the vega system , expect as noted . ] and f435w ( b ) to best enhance the visibility of the blue cluster against the contrast of the older , redder population of the larger surrounding nucleus . even the fine scale of the hrc ( @xmath27 ) undersamples the point - spread function ( psf ) in the blue , however ; we thus dithered the exposures in a @xmath28 square pattern of 0.5 pixel sub - steps ( respecting the different angular sampling - pitches of the two ccd axes ) . in detail , the b - band images were obtained in a single orbit comprising two dither - sets with a total of eight 298s exposures . the two sets were offset by 3.5 pixels to reject hot - pixels and other fixed defects . the reduced flux and sensitivity of the camera in the @xmath0-band required three orbits to obtain an adequate signal . a complete dither set was obtained in each orbit for a total of 12 672s - exposures . the three sets were offset from each other in an l - shaped pattern of @xmath29 steps , again to reject ccd defects . the total peak signal levels per pixel are @xmath30 in p3 and @xmath31 in p1 for the @xmath0 image ; and @xmath32 in p3 and @xmath33 in p1 for the b image . the goal of the image reduction was to produce a single summed - image for each filter with a @xmath34 finer scale , sufficient to ensure nyquist sampling . because the images were obtained with non - redundant pointings , the image construction was done in stages to eliminate cosmic ray events and other defects . the sub - pixel dither - patterns were executed very precisely , thus each dither set could be simply interlaced to produce an initial up - sampled image at the given pointing , with the images within the set used to recognize and repair ( with interpolation ) cosmic - ray events by comparing any one image to its three neighbors . the two b and three @xmath0 interlaced images were then combined into a single image for each filter . this initial estimate of a nyquist - sampled image was then used in a repeat of the first reduction cycle to go back to and improve the initial detection and repair of cosmic ray events ( and fixed ccd defects ) in the source images . with the second cycle of defect repair completed , the source images could be then combined more precisely to make a nyquist - sampled image directly , using the fourier - method of @xcite . this procedure combines the images in the fourier domain to algebraically remove the aliased power present in the source images ; significantly , it handles departures from the optimal dither pattern and combines the complete data set into the optimal average nyquist - sampled image ( in a least - squares sense ) . crucially , the @xcite method of image reconstruction is used in preference to _ drizzle _ @xcite . the fourier - method requires no parametric choices and avoids the blurring inherent in _ drizzle _ reconstructions . the final reduction step for the @xmath0 and b images was to reproject them to rectilinear sampling , removing the inherent angular shear and the rectangular @xmath27 pixels present in the native hrc image - sampling . this was done with reference to the the stsci field - distortion tables for the hrc , which vary with filter . sinc - function interpolation was used to compute the undistorted images . this is the only appropriate interpolation algorithm to use for nyquist - sampled images ; it does not degrade the resolution in any way . the final pixel scale of the images was selected to be @xmath35 which is 1/4 of the wfpc2/pc pixel scale , to allow for convenient comparison of the present hrc images to the v and i m31 nucleus images presented in @xcite . the nyquist - sampled , rectified , and summed @xmath0 and b images of the m31 nucleus are presented in figures [ fig : u_fig ] and [ fig : b_fig ] . the spatial resolution of the @xmath0 images is @xmath36 fwhm , or 0.15 pc at m31 , putting this among the highest resolution images obtained by _ hst . _ as sharp as this psf width is , however , it is actually significantly broader than a purely diffraction - limited psf at 330 nm , which has @xmath37 fwhm . this is due in part to the blurring associated with the hrc pixel - kernel ; however , psf deconvolution can largely correct for this , as well removing the wings of the psf , yielding nearly diffraction - limited images at both @xmath0 and b. figures [ fig : u_fig ] and [ fig : b_fig ] also show the nuclear images after 40 iterations of lucy - richardson @xcite deconvolution . the psfs required were generated from bright stars present in the outskirts of the images . tests presented in @xcite show that this deconvolution procedure is well - suited to recovering the intrinsic light distribution of the m31 nucleus . the final fwhm after deconvolution is @xmath38 which corresponds to a physical resolution at m31 of 0.11 pc . the uv - bright p3 star - cluster clearly dominates the @xmath0 image . the cluster has a clumpy appearance in this filter , and further breaks up into apparently discrete sources in its outskirts ( about @xmath39 from the center of the cluster ) . p3 is still bright in the b image ; however , it appears to be less clumpy and more concentrated . as we show below , the source that appears to define the center of p3 ( s11 in table [ tab : stars ] ) is somewhat redder than the surrounding cluster , while the discrete sources in the cluster outskirts are especially blue . the center of p3 thus increases in relative prominence going from the @xmath0 to the b band - pass . the center of p3 also remains unresolved . s11 , and a secondary maximum ( s07 in table [ tab : stars ] ) only offset from it by @xmath40 are both extended rather than stellar . both sources likely comprise the blended light of several bright stars . at large radii , both the @xmath0 and b images show the presence of a number of hot stars distributed throughout the nucleus . all the images shown in this paper are limited to radii interior to the nucleus . ] this is the uv - bright stellar population observed by @xcite and @xcite , although these authors did not characterize sources within @xmath41 of the nucleus . the visual impression is that the hot stars outside of p3 appear to be weakly concentrated about the cluster , but show no obvious correlation with the p1 or p2 concentrations of the tremaine disk . p3 is much less prominent at redder band - passes . it is essentially invisible at i - band , and shows up only weakly in v - band , as can be seen in figure [ fig : ubvi_fig ] , which compares the present hrc images to the wfpc2/pc images presented in @xcite . for this comparison , all four images were normalized to have the same surface brightness at the center of the diffuse p1 component . the wfpc2/pc images had already been subsampled to a scale @xmath34 finer than that of the native pc - scale , and were readily shifted , interpolated , and rotated using sinc - function interpolation to a final scale @xmath42 finer than the native pc scale . figure [ fig : color ] shows the @xmath43 b , and v images combined to make a representative color image . the reddish color of the tremaine disk contrasts strongly with the hot blue p3 population . in passing , we note that the grainy texture of p1 , p2 , and the surrounding nucleus in the deconvolved images is due to strong intrinsic sbf . close comparison of the b and v images show them to have correlated sbf patterns , an impressive agreement , given that the images were taken with different hst cameras , with differing psfs and pixel scales . at @xmath1 angular resolution the image `` noise '' characteristics are dominated by the finite number of stars making up the nuclear components , not the number of photons collected by the instruments . the reduced visibility of p3 at v and i highlights the distinction between it and p2 , under the revised schema of @xcite . p2 , which would now be considered to be the stellar emission associated with the _ peri - center _ portion of the tremaine disk , is visible in the v and i images as a highly elongated feature extending beyond p3 in the `` anti - p1 '' direction . the centroid of this emission is clearly displaced from p3 . figures [ fig : u_fig ] and [ fig : b_fig ] are shown as negative prints , while figure [ fig : ubvi_fig ] is given a positive gray - scale to offer some diversity in representing the inner nucleus . the negative gray scale in the first two figures , for example , appears to accentuate the diffuse toroidal morphology of the tremaine disk . to understand the properties of the p3 system , we attempted to isolate it from the surrounding nucleus . since the cluster is much less prominent at v and b or @xmath0 , we used the v image as a proxy for the `` background '' light distribution for both the b and @xmath0 images . the correct scaling required to subtract the background was somewhat problematic , given variations in color among the different nuclear stellar populations . as shown in @xcite , p1 is significantly redder than the outer nucleus , having @xmath44 with the portion of p2 outside p3 in the `` anti - p1 '' direction having an intermediate color . a simple hypothesis is that the tremaine disk , which comprises both p1 and p2 , is redder than the outer portions of the nucleus . p2 appears to be slightly bluer than p1 , since it has lower contrast against the underlying nucleus . the redder color of the tremaine disk as compared to the surrounding nucleus means that there is no simple scaling of the v to the @xmath0 or b images that will yield a completely flat background around p3 . scaling the v image to the outer nucleus in the @xmath0 image over - subtracts the tremaine disk , leaving large negative residuals at p1 , as well as somewhat reduced but still negative residuals in the p2 region immediately surrounding p3 . conversely , scaling the two images by their p1 fluxes , leaves p3 in a significantly positive background . our solution was to scale to the portion of p2 immediately outside the apparent extent of the p3 cluster in the anti - p1 direction . this still yields a positive background in the outer nucleus , and negative residuals in p1 , but does produce a low background surrounding p3 . we also `` clipped out '' residual portion of p3 still present at v , and slightly smoothed the image as well , to avoid biassing the measurement of the central fluxes of p3 in @xmath0 or b , or introducing artificial structure to either image . the isolated p3 cluster before and after deconvolution in the @xmath0 and b filters is shown in figure [ fig : p3 ] . this figure zooms into the cluster , ratifying the visual impression stated at the start of this section that the outskirts of the cluster breakup into individual point sources . the deconvolved images show little diffuse emission outside the very center of the cluster . close comparison of the @xmath0 and b images shows that the sources cover a range in color . there are a number of sources that are readily visible in @xmath43 but nearly invisible in b ( which unfortunately give rise to large errors in the @xmath3 values of the sources ) . the bright source , s11 , at the center of p3 is clearly extended , as is s7 , the source adjacent to it ; however , to the degree of resolution afforded by these observations , most of the other sources appear to be stellar . measuring the fluxes of the p3 sources is difficult with traditional psf - fitting techniques , given the high degree of crowding and the strongly variable background . instead , we used a novel procedure that measures stellar fluxes from aperture photometry on deconvolved images , which has been shown to work extremely well on nyquist - sampled acs / hrc images of m32 @xcite . the methodology is to take the deconvolution of the @xmath0 and b frames to 640 lucy - richardson iterations . this effectively transforms the observed psf into a highly compact `` deconvolved psf '' that essentially concentrates all of its flux to within a radius of @xmath45 aperture photometry is done on the sources in the resulting image , once the background nucleus light is subtracted off . the resulting colors and magnitudes of the sources within 1 pc of p3 are mapped in figure [ fig : stars ] and listed in table [ tab : stars ] . a comparison of this figure to the images in figure [ fig : p3 ] , ratifies the impression that the bluer sources occur on the outskirts of the p3 cluster . interpreting the present photometry is complicated by the large color - terms required to bring the f330w into concordance with the standard johnson u band - pass . based on the color equations presented by @xcite , for johnson @xmath46 we derive the transformation from the f330w to u bandpass as @xmath47 fortunately , the difference between the hrc f435w filter and johnson b only varies by a few hundreds of a magnitude over the large f330w - f435w color range of the observations , and can be directly taken as b , given the larger random errors in the photometry . the analogous transformation for @xmath48 is @xmath49 as can be seen from the equations above , and the transformed values in table [ tab : stars ] , the implied corrections are as large as @xmath50 of the observed @xmath51 colors . while @xcite present calibrating observations over the needed range of @xmath52 the actual corrections will depend on the unknown metalicities and atmospheric properties of the actual p3 stars . we plot the colors and absolute magnitudes of the p3 sources in figure [ fig : cmd ] . we attempted to correct for both foreground and m31 internal dust absorption . the galactic foreground extinction provided by @xcite is @xmath53 mag , which given the reddening tables in @xcite yields @xmath54 and @xmath55 mag . a survey of the literature shows that internal dust extinction in the m31 bulge is small ( e.g. * ? ? ? hui dong ( private communication ) estimates internal @xmath56 for the region of the bulge just outside the nucleus based on stellar - population models of multi - color _ hst _ imagery . the implied total values are then @xmath57 and @xmath58 mag . figure [ fig : cmd ] plots padova @xcite isochrones ( blue ) for single - burst stellar populations of 50 , 100 , and 200 myr . it appears that most , if not all , of the sources are consistent with an age of @xmath59 myr . we note that the brightest source in the cluster , s11 in table [ tab : stars ] , is also the reddest source ; it is extended as well . s11 is at the center of p3 , and we have adopted it is the _ ad hoc _ center of the system for the coordinates given in table [ tab : stars ] and figure [ fig : stars ] . at the same time , evolutionary tracks for some late evolutionary - phases of old stellar populations pass through the u and @xmath3 domain occupied by the p3 stars , as well . these are all due to metal - rich giants that have evolved into `` extreme horizontal branch '' ( ehb ) stars and later move back to the red as their core helium - burning phase is completed . stars that populate the ehb have lost a significant amount of their envelope on the rgb , though the mechanism causing the mass loss is not precisely known ( for a full review see * ? ? ? * ; * ? ? ? these stars do not have a large enough convective envelope to allow a full agb phase . instead , the hottest of these ehb stars will fail to reach the agb track at all , and evolve through the agb - manqu channel . cooler ehb stars will eventually leave the agb track early and become post early - agb stars ( peagb ) . these phases , collectively called hot post - horizontal branch ( hp - hb ) evolution , have been previously found in m31 ( e.g. , * ? ? ? * ; * ? ? ? their surface density has recently been shown to increase towards the center of m31 ( rosenfield et al . , submitted ) more rapidly than the underlying bulge surface brightness , although there appears to be no enhancement associated with the overall nucleus , itself . figure [ fig : cmd ] shows several examples of hp - hb tracks , such as h - burning post - agb ( pagb ) tracks of masses @xmath60 ( red , dashed ) from @xcite with @xmath61 and @xmath62 and agb - manqu and peagb tracks ( red , dotted ) of masses @xmath63 these are from the most recent padova stellar evolution library ( bressan et al . , in prep . ) with @xmath64 and @xmath65 and an @xmath66-enhanced composition typical of elliptical galaxies ( adapted from * ? ? ? tracks were converted to johnson filters following the color transformations of @xcite . the metallicity of the padova tracks was chosen as an extreme case to allow for high helium content , ensuring high mass loss on the rgb . as these tracks are preliminary , other metallicity and mass loss rate combinations could also produce hp - hb stars with similar luminosities and effective temperatures . figure [ fig : cmd ] shows that the canonical pagb tracks are too bright and too hot to explain the p3 sources ; however , a number of the p3 sources , especially the bluest and faintest ones , are consistent with the peagb evolutionary tracks shown ( and the brightest of the agb - manqu tracks , which is the faintest track shown ) . despite this concordance , however , we suggest that only a few of the p3 stars could be in the hp - hb phase , in contrast to being hot newly - formed stars . hp - hb evolution across the u and @xmath3 domain in figure [ fig : cmd ] is very rapid , occupying only @xmath67 yr . the p3 stars are highly concentrated , suggestive of a local population tightly bound to the black hole , but there is no trace of the associated substantial population of giants required to generate and maintain a short - lived uv - phase sufficient to account for the complete uv luminosity of p3 . a separate approach to characterizing the stellar population of p3 is to use surface brightness fluctuations " ( sbf ) as a diagnostic @xcite . in this case we are effectively using the clumpiness " of p3 as an indicator of the typical luminosity of the stars that characterize the variance in surface brightness . for a constant total luminosity of the p3 cluster , the variance about a perfectly smooth distribution will steadily increase as we go from modelling it as the aggregate of a large number of intrinsically faint old stars to a small handful of high luminosity young stars . the procedure for conducting an sbf analysis on p3 requires measuring the power spectrum from the images of the cluster , and comparing it to the power spectra of models of the cluster . measurement of the observed power spectrum begins with subtracting a smooth model of the p3 light distribution ( see @xmath68 ) , and then measuring the power spectrum of the residual image . since the sbf technique uses the shape of the psf power - spectrum as a template , the analysis is done on the background - subtracted image of p3 in the observed , as opposed to the deconvolved domain . in detail , we restricted the analysis to a @xmath69 radius aperture centered on p3 . we also only performed the sbf analysis on the @xmath70 image , given concerns about the accuracy of the technique in the b image , which has a significantly reduced contribution from p3 , but increased background variance ( which can not be subtracted ) from the underlying old - population of the nucleus . the `` raw '' power spectrum of p3 includes a constant noise term that largely represents random photon shot - noise in the image . we subtracted this off by measuring the power a spatial frequencies higher than the psf nyquist scale . we show the resulting power spectrum in figure [ fig : sbf ] . the simulated power spectra comes from models constructed for bursts of 50 , 100 , 200 , and 400 myr , drawing stars at random from the padova @xcite isochrones transformed to @xmath71 the location of the stars was drawn at random from the smooth model of the p3 light distribution until the total @xmath70 luminosity of p3 was reached . the models were then analyzed by the same procedure used for the real data . a few dozen models were generated for each burst age to average out statistical fluctuations in the power spectra . the power spectra in figure [ fig : sbf ] are plotted in terms of the implied absolute @xmath70 magnitude of the flux . the intercept of the curves at zero spatial - frequency gives the luminosity of the @xcite @xmath72 parameter , which characterizes the variance as due stars of that representative brightness . the limit of the vertical scale is set by the luminosity of p3 , itself . the observed sbf power appears to fall between that of the 100 and 200 myr bursts , which is qualitatively ratified by the subjective appearance of the p3 models of the same age . this result is thus in excellent agreement with the p3 cmd analysis presented in the previous section . while we can constrain the age of the p3 population , it appears , however , that we have little sensitivity to the form of the initial mass function ( imf ) , given the limited dynamic range of the luminosity of the sources seen in p3 . @xcite argue that the stars surrounding the milky way black hole have a markedly `` top - heavy '' imf . we generated 200 myr sbf models with their @xmath73 imf to see if this might produce significantly different sbf power as compared to that for the same age derived from the padova isochrones . we only measured a modest increase in sbf power , however , which would cause us to infer a slightly older age for the population by @xmath74 myr ( it takes an older `` top - heavy '' model to match the observed sbf power of p3 , itself ) , a result that is within the error limits of the sbf results already quoted . the sbf power is dominated by the resolved sources near the main - sequence ( ms ) turn - off , which is determined by age alone . a more top - heavy imf qualitatively down - weights the contribution of fainter stars on the ms to the total sbf power , causing the relative sbf variance to increase for the same age and total p3 luminosity . however , given the steep @xmath75 mass - luminosity relation for the ms , we have in the end little leverage on the mass function . a small range in ms mass corresponds to a large change in stellar luminosity , thus suppressing information on the mass function , itself . we further note that the cmd analysis has even less sensitivity to the imf , given the even more limited mass range of the resolved stars . some sample models for p3 are shown in figure [ fig : p3_mod ] . the clumpiness of the models shown can be seen to steadily decrease with increasing burst age . it is notable that the strong sbf pattern completely dominates the appearance of the simulated clusters . further , the location of either the peak brightness , or light - weighted centroid of the entire cluster can be displaced significantly away from the nominal center of the ideal smooth cluster model due to the large random fluctuations in the structure . this is an important caveat to respect when evaluating the precise location of the p3 cluster with respect to the m31 center of mass or the location of the m31 black hole ( see below ) . for the ensemble of 50 myr models , the average location of the peak flux is offset from the true cluster center by @xmath76 and the photo - center is displaced placed by @xmath40 with the typical scatter from model to model being nearly as large as these mean values . for the 100 myr models , the two offsets decrease to @xmath77 and @xmath78 and @xmath1 and @xmath79 for the 200 myr models . the total brightness of p3 was measured by aperture photometry performed on the difference images . the apparent magnitudes of p3 are @xmath80 ( ab ) or @xmath81 ( vega ) in @xmath0 , and @xmath82 ( ab ) and @xmath81 ( vega ) in b , uncorrected for foreground extinction . using the reddening values stated earlier , the the corrected apparent fluxes are @xmath83 @xmath84jy in @xmath0 and @xmath85 @xmath84jy in b. the implied reddening - corrected colors are @xmath86 and @xmath87 ( vega ) , using the @xmath88 measure from @xcite . the implied spectral energy distribution ( sed ) of p3 is plotted in figure [ fig : flux ] , based on the fuv 175 nm flux measured by @xcite , the @xmath0 and v fluxes measured by @xcite , the present @xmath0 and b values . we reduced the fuv measurement from @xmath89 to @xmath90 , for consistency with the present reddening corrections . @xcite assumed @xmath91 based on the total @xmath92 for m31 measured by @xcite . our @xmath93 value implies @xmath94 instead . an @xmath95 sed of an a0 main sequence star scaled to the flux - points is shown for comparison . as noted in @xcite , and shown spectroscopically by @xcite , the sed of p3 is indeed `` a - like . '' the fuv point does fall well below the template , but this portion of the sed is highly sensitive to temperature , and the fuv point could be fitted with a modest adjustment of the template ; this point will also be the most sensitive to the correct value of the extinction . the over all distribution of light within the p3 cluster is confirmed to roughly follow the exponential light distribution derived by @xcite . a @xmath0-band surface brightness profile of p3 is shown in figure [ fig : p3_sb ] , with the exponential model shown as a solid line . s11 was taken as the center of the cluster , but the fit excludes points with @xmath96 given that the inner - most points will still be affected by the residual psf associated with the deconvolution . the exponential form recovered is @xmath97 where @xmath98 is the semimajor axis in arcseconds , and @xmath99 is the surface brightness ( magnitudes / sq . arcsec ) observed with the f330w vega - based zeropoint ( no reddening correction has been applied ) . the implied exponential scale - length is @xmath100 which is marginally more compact than the @xmath101 value derived by @xcite . the average ellipticity of the isophotes is @xmath102 which implies an inclination of @xmath103 if p3 is interpreted to be a disk , again in good agreement with the @xmath104 of @xcite . although s11 was excluded from the fit , given concerns about the residual deconvolved - psf , the brightness profile of s11 actually rises about the exponential model . the core radius of s11 is @xmath105 or @xmath106 pc . if we estimate the volume luminosity - density of s11 from @xmath107 where @xmath108 is the central surface brightness ( @xmath109 in u , corrected for extinction ) , then at s11 we find @xmath110 in the u - band . the present work provides no new dynamical information about the m31 black hole , and can not offer any objective improvement on the conclusion of @xcite or @xcite that the black hole must be close to the center of p3 . at the same time , we advance subjective arguments that the location of the black hole may be coincident with s11 ( see table [ tab : stars ] ) . the central location of this source is suggestive . while we showed in @xmath111 that simulated models of the cluster often have the brightest clump of stars to be significantly offset from the true center of the cluster ( which is where we would expect the black hole to be ) . however , while we do not know the `` true '' center of p3 , case , the location of s11 and the photo - center of p3 at least agree to within @xmath45 s11 is also extended and significantly redder than the other sources , underscoring its unique properties among the other sources . its red color indeed makes it the only source to mark the location of p3 in the v - band . a truly extended source should not exist within the strong tidal field of the black hole , unless it represents the peak of a cusp made of the stars most closely bound to it . at the same time , it s possible that the structure of s11 is only due to a random fluctuation of stars along the line - of - sight . s07 , is also extended , is offset from the center of p3 , but otherwise has no special attributes . the sbf simulations presented in the previous section show that most of the sources in p3 are indeed dominated by a single bright star ; however , very close to the center there can be one or two blends of similarly bright stars within the hrc psf . we began the paper with a history of earlier work on p3 , which already built a strong case that it is a cluster of young stars . we have now partially resolved the cluster into individual stars , and characterized its form with significantly higher spatial resolution than was provided by the wfpc2/pc observations of @xcite ; however , in broad detail the conclusions of that paper , and the later work of @xcite remain unaltered . both a color - magnitude diagram ( cmd ) obtained of the resolved stars , and an sbf analysis performed on the total image , show that p3 plausibly comprises a population formed in a burst 100 to 200 myr ago , consistent with the p3 sed and the spectrum obtained by @xcite . the result may also be supported by the @xcite spectroscopy observation of the inner bulge , which suggest that a burst of star formation occurred within the nucleus about @xmath112 myr ago . one small modification of the earlier results provided by the present work is that that p3 may be slightly younger than the 200 myr age derived by @xcite , but that work really only ruled out an age much younger than 100 myr . another new result is that the color of s11 suggests that a single - age starburst model for the cluster does not capture the full story of p3 . at some level , the large luminosity of p3 blends with the much bluer stars around it to generate an overall a - like sed . it is intriguing that s11 provides the only visible trace of p3 in the v - band . if it does mark the location of the m31 black hole , then it suggests that there may yet be another change in the stellar population of the nucleus within @xmath113 or @xmath114 pc of the black hole . although we have concluded that the p3 stars are young , @xcite and @xcite argue that the uv - bright stars in the bulge of m31 , which of course surrounds the nucleus , represent the final stages of stellar evolution in an old metal - rich stellar population . if there are truly two different populations contributing uv - bright stars to the bulge and nucleus of m31 , it still begs the question of whether or not old remnants may be present in p3 , as well as how far out the putative young stars in p3 may be traced out into the surrounding nucleus or even bulge . we show that peagb tracks traverse the u vs. @xmath3 domain defined by the p3 stars , thus it is possible that a few of the p3 stars are indeed old - evolved stars , rather than being young massive main sequence stars . however , since this is a very rapid evolutionary phase , it is extremely unlikely that the entire p3 cluster could be accounted for by peagb stars . a strong factor in constraining the population of p3 is the compactness of p3 , itself . the uv - stars outside the nucleus do trace the bulge - light at long wavelengths , but yet are not strongly concentrated within the outer nucleus ; their abundance by total light or number is but a small faction of the bulge . requiring any sort of hp - hb stars as the dominant p3 population requires an associated massive population of red giants also closely bound to the m31 black hole to continually generate the short - lived peagb stars . there is no evidence for such a population in the v and i nuclear images of @xcite . this same consideration would appear to also apply to the hypothesis of @xcite , who suggest that the p3 uv - bright stars may be the remnant cores of giants stripped by close encounters with the black hole . the high velocities associated with the black hole are such that stars passing close to the black hole from the bulge , or even the outer nucleus will not dwell in the vicinity of p3 . if the p3 stars are really processed giants , they still must have been initially closely bound to the black hole , which again begs the question of where the progenitor giants are . the only way that this mechanism might work is if the stripping of the p3 stars takes place just as they begin their first ascent up the red giant branch . this might produce long - lived uv - bright stripped - cores , and would account for lack of progenitor giants as well . further investigation of the origin of p3 might be linked to understanding the early - type stars closely bound to the black hole in our own galaxy . the existence of p3 in a second local group galaxy suggests that this is not a rare phenomenon . the present observations offer an additional site to test mechanisms that can form stars within the strong tidal field of a supermassive black hole . we finish by noting that as with the galactic center , it may be possible to measure the proper motions of stars closely bound to the m31 black hole . figure [ fig : stars ] shows that circular velocities in the plane of the sky around the m31 black hole will exceed @xmath115 over the entire extent of p3 . it is possible that proper motions may be detected in about a decade after the present observations were made . we thank karl gebhardt , scott tremaine , jessica lu , and pierre demarque for useful conversations . we thank alessandro bressan and leo girardi for advance use of preliminary padova stellar evolution tracks . we thank the referee for a number of useful suggestions . pete merenfeld ( noao ) kindly prepared the three - color image of the nucleus . alexander , t. 2006 , journal of physics conference series , 54 , 243 allen , d. a. , hyland , a. r. , & hillier , d. j. 1990 , , 244 , 706 bartko , h. , martins , f. , trippe , s. , et al . 2010 , , 708 , 834 bender , r. , et al . 2005 , , 631 , 280 bensby , t. , feltzing , s. , johnson , j. a. , et al . 2010 , , 512 , a41 bertola , f. , bressan , a. , burstein , d. , buson , l. m. , chiosi , c. , & di serego alighieri , s. 1995 , , 438 , 680 bonnell , i. a. , & rice , w. k. m. 2008 , science , 321 , 1060 brown , t. m. , ferguson , h. c. , stanford , s. a. , & deharveng , j .- 1998 , , 504 , 113 chang , p. , murray - clay , r. , chiang , e. , & quataert , e. 2007 , , 668 , 236 crane , p. c. , dickel , j. r. , & cowan , j. j. 1992 , , 390 , l9 del burgo , c. , mediavilla , e. , & arribas , s. 2000 , , 540 , 741 demarque , p. , & virani , s. 2007 , , 461 , 651 dressler , a. 1984 , , 286 , 97 dressler , a. , & richstone , d. o. 1988 , , 324 , 701 forrest , w. j. , shure , m. a. , pipher , j. l. , & woodward , c. e. 1987 , the galactic center , 155 , 153 freedman , w. l. , & madore , b. f. 1990 , , 365 , 186 fruchter , a. s. , & hook , r. n. 2002 , , 114 , 144 garcia , m. r. , williams , b. f. , yuan , f. , et al . 2005 , , 632 , 1042 garcia , m. r. , hextall , r. , baganoff , f. k. , et al . 2010 , , 710 , 755 greggio , l. , & renzini , a. 1999 , , 70 , 691 king , i. r. , stanford , s. a. , & crane , p. 1995 , , 109 , 164 king , i. r. , et al . 1992 , , 397 , l35 kormendy , j. 1988 , , 325 , 128 kormendy , j. & bender , r. 1999 , , 522 , 772 krabbe , a. , et al . 1995 , , 447 , l95 lauer , t. r. 1999 , , 111 , 227 lauer , t. r. , faber , s. m. , ajhar , e. a. , grillmair , c. j. , & scowen , p. a. 1998 , , 116 , 2263 lauer , t. r. et al . 1993 , , 106 , 1436 levin , y. , & beloborodov , a. m. 2003 , , 590 , l33 li , z. , wang , q. d. , & wakker , b. p. 2009 , , 397 , 148 lucy , l. b. 1974 , , 79 , 745 marigo , p. , girardi , l. , bressan , a. , et al . 2008 , , 482 , 883 mcclure , r. d. , & racine , r. 1969 , , 74 , 1000 monachesi , a. , trager , s. c. , lauer , t. r. , freedman , w. , dressler , a. , grillmair , c. , & mighell , k. j. 2011 , , 727 , 55 nieto , j .- l . , macchetto , f. d. , perryman , m. a. c. , di serego alighieri , s. , & lelievre , g. 1986 , , 165 , 189 oconnell , r. w. 1999 , , 37 , 603 peiris , h. v. , & tremaine , s. 2003 , , 599 , 237 redman , r. o. , & shirley , e. g. 1937 , , 97 , 416 richardson , w. h. 1972 , j. opt . a. , 62 , 52 richstone , d. , bower , g. , & dressler , a. 1990 , , 353 , 118 saglia , r. p. , fabricius , m. , bender , r. , et al . 2010 , , 509 , a61 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , , 500 , 525 sirianni , m. , et al . 2005 , , 117 , 1049 tempel , e. , tamm , a. , & tenjes , p. 2010 , , 509 , a91 tonry , j. , & schneider , d. p. 1988 , , 96 , 807 tremaine , s. 1995 , , 110 , 628 vassiliadis , e. , & wood , p. r. 1994 , , 92 , 125 wardle , m. , & yusef - zadeh , f. 2011 , arxiv:1108.2175 rrrccccc s01&@xmath116&@xmath117&@xmath118&@xmath119&@xmath120&@xmath121&@xmath122 + s02&@xmath123&@xmath124&@xmath125&@xmath126&@xmath127&@xmath128&@xmath129 + s03&@xmath130&@xmath131&@xmath132&@xmath133&@xmath134&@xmath135&@xmath136 + s04&@xmath137&@xmath138&@xmath139&@xmath140&@xmath141&@xmath142&@xmath143 + s05&@xmath144&@xmath145&@xmath146&@xmath147&@xmath148&@xmath149&@xmath150 + s06&@xmath151&@xmath152&@xmath153&@xmath154&@xmath155&@xmath156&@xmath157 + s07&@xmath158&@xmath159&@xmath160&@xmath161&@xmath162&@xmath163&@xmath164 + s08&@xmath165&@xmath166&@xmath167&@xmath168&@xmath169&@xmath170&@xmath171 + s09&@xmath172&@xmath173&@xmath174&@xmath175&@xmath176&@xmath177&@xmath178 + s10&@xmath172&@xmath179&@xmath180&@xmath181&@xmath182&@xmath183&@xmath184 + s11&@xmath185&@xmath185&@xmath186&@xmath187&@xmath188&@xmath189&@xmath190 + s12&@xmath191&@xmath192&@xmath193&@xmath194&@xmath195&@xmath196&@xmath197 + s13&@xmath198&@xmath199&@xmath200&@xmath177&@xmath201&@xmath202&@xmath203 + s14&@xmath204&@xmath205&@xmath206&@xmath207&@xmath208&@xmath209&@xmath164 + s15&@xmath210&@xmath211&@xmath212&@xmath213&@xmath214&@xmath215&@xmath216 + s16&@xmath217&@xmath218&@xmath219&@xmath220&@xmath221&@xmath222&@xmath223 + s17&@xmath224&@xmath225&@xmath226&@xmath227&@xmath228&@xmath229&@xmath230 + s18&@xmath231&@xmath232&@xmath233&@xmath234&@xmath235&@xmath236&@xmath237 + s19&@xmath238&@xmath239&@xmath240&@xmath241&@xmath242&@xmath243&@xmath244 + s20&@xmath245&@xmath172&@xmath246&@xmath247&@xmath248&@xmath249&@xmath250 + [ tab : stars ]

magnetic fields are ubiquitous in diverse astrophysical settings on various scales ranging from proto - stellar discs , young stellar objects ( ysos ) , microquasars , quasars , galaxies and grandiose astrophysical jets to clusters of galaxies . in some cases , magnetic fields are relatively weak so that dynamical processes are almost unaffected by their presence . however , there do exist numerous cases where magnetic fields are necessary and important for both dynamics and diagnostics , especially in spiral galaxies and accretion disc systems ( e.g. , sofue et al . 1986 ; beck et al . 1996 ; balbus & hawley 1998 ; tagger & pellat 1999 ; widrow 2002 ) . in general , it is challenging to model magnetic fields realistically because of their complexities ( sofue et al . 1986 ; kronberg 1994 ; caunt & tagger 2001 ; widrow 2002 ) . in the problems involving protostar formation and disc galaxies , shu & li ( 1997 ) presumed the so - called ` isopedic ' magnetic field configuration where the mass - to - magnetic flux ratio in a razor thin disc remains constant . using a singular isothermal disc ( sid ) model ( mestel 1963 ) that is isopedically magnetized , shu et al . ( 2000 ) studied stationary coplanar perturbations and proposed that these global modes are bifurcations to either secularly or dynamically unstable configurations . along a separate yet complementary line , lou ( 2002 ) studied an azimuthally magnetized singular isothermal disc ( msid ) and derived two different global stationary magnetohydrodynamic ( mhd ) perturbation modes . based on the msid model , lou ( 2002 ) explored the manifestation of interlaced optical and magnetic field spiral arms in the outer portion of a disc with a nearly flat rotation curve such as the case of the nearby spiral galaxy ngc 6946 ( beck & hoernes 1996 ; fan & lou 1996 ; lou & fan 1998a , 2002 ; frick et al . 2000 , 2001 ; lou et al . 2002 ) . in contexts of spiral galaxies , it is natural and important to consider a composite disc system consisting of gravitationally coupled gaseous and stellar discs . this is because various physical processes on different scales occur in the gaseous disc yet large - scale gas dynamics and environment are significantly affected by large - scale structures in the stellar disc ( lou & fan 1998b ) . to construct global coplanar perturbation structures in a systematic manner , we started with a composite sid system of two fluid discs without magnetic field ( lou & shen 2003 ) to generalize the work of shu et al . ( 2000 ) for a single sid . in terms of the global axisymmetric stability for such a composite sid system , we provided a more straightforward @xmath12criterion ( shen & lou 2003 ) in contrast to the axisymmetric stability criteria of elmegreen ( 1995 ) and jog ( 1996 ) . to be more general than a background sid profile , we recently constructed global stationary perturbation patterns in a composite system of two fluid scale - free discs without magnetic field ( shen & lou 2004a , b ) to further generalize the work of syer & tremaine ( 1996 ) . with proper adaptations , these global stationary perturbation solutions may be utilized to model large - scale structures spiral galaxies . while a scale - free disc has no characteristic scales , one may impose boundary conditions to describe a modified scale - free disc , e.g. , by cutting a central hole in a disc ( e.g. , zang 1976 ; evans & read 1998a , b ) . with such boundary conditions , discrete eigenfunctions for perturbations may be constructed as growing normal modes . on the other hand , by specifying a phase relation for a postulated reflection of spiral waves from the origin @xmath13 , goodman & evans ( 1999 ) could also define discrete normal modes even for an unmodified gaseous sid . shu et al . ( 2000 ) speculated that the swing amplification process ( e.g. , goldreich & lynden - bell 1965 ; julian & toomre 1966 ; toomre 1977 ; fan & lou 1997 ) across corotation allows a continuum of normal modes ( lynden - bell & lemos 1993 ) while proper ` boundary conditions ' may select from this continuum a discrete spectrum of unstable normal modes ; they also suggested that zero - frequency ( stationary ) disturbances do signal the onset of instability . on the basis of the work of shu et al . ( 2000 ) for a single isopedically magnetized sid and of lou & shen ( 2003 ) for a composite sid system without magnetic field , lou & wu ( 2004 ) analyzed coplanar mhd perturbation structures in a composite sid system with one of the disc being isopedically magnetized ( shu & li 1997 ) . as x - ray emitting hot gases can freely flow along open magnetic field lines anchored at the gaseous disc , global stationary solutions constructed by lou & wu ( 2004 ) set an important preliminary stage for modelling spiral galactic mhd winds . in particular , if activities of star formation , starbursts and supernovae etc . are mainly responsible for producing x - ray emitting hot gases streaming out of the galactic plane from both sides , we would expect stronger winds from the circumnuclear starburst region ( e.g. lou et al . 2001 ) and along spiral arms than those winds from the remainder of the galactic disc plane . both sid and msid models offer valuable physical insight for disc dynamics and belong to a wider family of scale - free discs ( syer & tremaine 1996 ) where physical variables in the axisymmetric background equilibrium scale as powers of radius @xmath14 ( e.g. , the rotation curve @xmath15 ) . it turns out that coplanar perturbations in such scale - free discs in a proper range of @xmath8 can be treated globally and analytically ( lemos et al . 1991 ; syer & tremaine 1996 ) without invoking the usual wkbj or tight - winding approximation to solve the poisson equation for density wave perturbations ( lin & shu 1964 ; binney & tremaine 1987 ; bertin & lin 1996 ) . in other words , it becomes feasible to include azimuthal magnetic fields in a scale - free disc and to construct global coplanar mhd perturbation configurations that are stationary in an inertial frame of reference . such analytical solutions , still idealized and limited , are making important further steps and are extremely valuable for bench - marking numerical codes and for initializing numerical mhd simulations . for an azimuthally magnetized scale - free disc with an infinitesimal thickness , the axisymmetric background equilibrium without radial flow is characterized by following features : a surface mass density @xmath0 , a rotation curve @xmath2 , an azimuthal magnetic field @xmath1 and a barotropic equation of state @xmath16 where @xmath5 is the vertically integrated gas pressure with constant @xmath17 for a warm disc and @xmath18 as the barotropic index . to maintain a radial force balance at all radii , there exists an explicit relationship among @xmath7 , @xmath8 , @xmath9 and @xmath6 such that only one parameter is independent . as will be seen in section 2 , it is convenient to use @xmath8 as this independent parameter ( syer & tremaine 1996 ; shen & lou 2004b ) . in order to satisfy relevant constraints , the prescribed range of @xmath8 turns out to be @xmath10 that includes the previous special case of @xmath19 ( i.e. , the msid case ) studied by lou ( 2002 ) , lou & zou ( 2004a ) and lou & wu ( 2004 ) . in section 2 , we introduce coplanar mhd perturbations and present the linearized perturbation equations . we construct and analyze global stationary aligned and logarithmic spiral perturbation configurations in section 3 and as an example of illustration , we discuss specifically the case of @xmath11 with the background azimuthal magnetic field being force - free . in section 4 , we derive phase relationships between perturbation enhancements of surface mass density and magnetic field for both aligned and logarithmic spiral cases , and discuss the problem of angular momentum conservation . finally , we summarize our results in section 5 . for the convenience of reference , relevant technical details are collected in the appendices . in the magnetofluid approximation , the disc is taken to be razor - thin ( i.e. , we use vertically integrated magnetofluid equations and neglect vertical derivatives of physical variables ) and large - scale stationary aligned and spiral coplanar disturbances develop in a background mhd rotational equilibrium of axisymmetry ( syer & tremaine 1996 ; lou 2002 ; lou & zou 2004a ; lou & wu 2004 ) . the background magnetic field is taken to be azimuthal to avoid the magnetic field winding dilemma ( lou & fan 1998a ) . for the sake of simplicity at this stage , non - ideal effects such as viscosity , resistivity and thermal diffusion etc . are ignored for large - scale perturbations . using cylindrical coordinates @xmath20 , we have the following two - dimensional nonlinear mhd equations for a razor - thin disc geometry : the mass conservation equation @xmath21 where @xmath22 is the surface mass density , @xmath23 is the radial bulk flow velocity , @xmath24 is the specific angular momentum in the vertical @xmath25-direction and @xmath26 is the azimuthal linear velocity ; the radial component of the momentum equation @xmath27\ , \end{split}\ ] ] where @xmath5 is the vertically integrated gas pressure ( sometimes referred to as two - dimensional pressure ) , @xmath28 is the total gravitational potential including that of an axisymmetric dark matter halo in contexts of disc galaxies , and @xmath29 and @xmath30 are the azimuthal and radial components of the coplanar magnetic field @xmath31 , respectively ; the azimuthal component of the momentum equation @xmath32\ ; \end{split}\ ] ] the poisson integral equation @xmath33^{1/2}}\ , \ ] ] where @xmath34 is defined as the ratio of the potential arising from the disc to that arising from the entire system including a massive dark matter halo that is presumed not to respond to coplanar perturbations in the disc plane ( syer & tremaine 1996 ; shu & li 1997 ; shu et al . 2000 ; lou & shen 2003 ; shen & lou 2004a , b ; lou & zou 2004a ) with @xmath35 for a full disc and @xmath36 for partial discs ; the divergence - free condition of magnetic field @xmath37 @xmath38 the radial component of the magnetic induction equation @xmath39 the azimuthal component of the magnetic field induction equation @xmath40 and the barotropic equation of state @xmath41 where constants @xmath17 and @xmath18 . among the three equations ( [ divergfree ] ) , ( [ radiaminduct ] ) and ( [ azimuminduct ] ) , we can freely choose two independent ones . by barotropic equation of state ( [ barotropic ] ) , the sound speed @xmath42 is defined by @xmath43 where subscript @xmath44 indicates the background equilibrium . an isothermal sound speed corresponds to a barotropic index @xmath45 . we now proceed to derive properties of an axisymmetric rotational mhd equilibrium characterized by physical variables associated with a subscript @xmath44 . in our notations , such a background equilibrium has the following power - law scalings : surface mass density of @xmath0 , a radial velocity profile @xmath46 and @xmath47 , a purely azimuthal magnetic field with @xmath48 and @xmath49 . substitutions of these power - law scalings into equations @xmath50 lead to the following condition for the radial force balance , namely @xmath51 where the alfvn speed @xmath52 in a thin disc is defined by @xmath53 and the poisson integral gives @xmath54 with the numerical factor @xmath55 explicitly defined by @xmath56 and @xmath57 being the standard gamma function . this expression ( [ p0 ] ) can also be included in a more general form of @xmath58 as in expression ( [ pm ] ) derived later with the limiting result of @xmath59 as @xmath60 . to satisfy the radial force balance equation ( [ sfeqn ] ) at all radii ( i.e. , scale - free ) would necessarily require ) in general and for the special case of @xmath11 , we have @xmath61 for the equilibrium magnetic field to be force - free ( e.g. , low & lou 1990 ) . the non - force - free msid case studied by lou ( 2002 ) has @xmath62 , @xmath19 , @xmath63 and @xmath45 . ] @xmath64 which immediately leads to the explicit expressions of indices @xmath7 , @xmath9 and @xmath6 all in terms of @xmath8 @xmath65 in this formulation , @xmath66 remains finite only if @xmath67 which implies @xmath68 . for the total gravity force arising from the equilibrium surface mass densities to be finite , a larger @xmath69range @xmath70 is allowed . the @xmath69range is also constrained by @xmath10 where the first inequality is required for @xmath18 and the second inequality is imposed such that the central point mass will not diverge . by these considerations , the plausible range of @xmath8 falls in @xmath71 ( see syer & tremaine 1996 and shen & lou 2004a , b ) . according to equations ( [ sfeqn ] ) and ( [ pararelation ] ) , we can explicitly write all physical variables of the background equilibrium state in power - law scalings of @xmath14 in terms of parameter @xmath8 , namely @xmath72 } { 2\pi g(2\beta{\cal p}_0)r^{1 + 2\beta}}\ , \ \ \qquad a^2=\frac{a^2}{(1 + 2\beta)r^{2\beta}}\ , \ \ \qquad c_a^2\equiv q^2a^2=\frac{q^2a^2}{(1 + 2\beta)r^{2\beta}}\ , \\ & v_0=\frac{ad}{r^{\beta}}\ , \ \ \qquad j_0\equiv rv_0=\frac{ad}{r^{\beta-1}}\ , \ \ \qquad \omega\equiv\frac{j_0}{r^2}=\frac{ad}{r^{1+\beta}}\ , \ \ \qquad\kappa\equiv [ ( 2\omega / r)d(r^2\omega)/dr]^{1/2}=[2(1-\beta)]^{1/2}\omega\ . \end{split}\ ] ] by introducing a reference radius @xmath73 , the constant @xmath74 is actually related to a scaled sound speed @xmath75 , the constant @xmath76 $ ] is a scaled rotational mach number , the constant @xmath77 is the ratio of the alfvn speed @xmath52 to the sound speed @xmath42 , @xmath78 and @xmath79 are the equilibrium angular speed and the epicyclic frequency , respectively ( lou 2002 ; shen & lou 2004a , b ) . the first expression of equation ( [ propequilib ] ) puts an upper limit on the magnetic field strength , namely , @xmath80 . alternatively , once a magnetic field parameter @xmath81 is chosen , the scaled rotational mach number @xmath82 must meet the following physical requirement @xmath83 for a positive equilibrium surface mass density @xmath84 . note that for @xmath85 , inequality ( [ restrt ] ) is automatically satisfied without actually restricting @xmath81 and the relevant physical requirement is simply @xmath86 . not all stationary solutions of @xmath87 can satisfy condition ( [ restrt ] ) . this then implies limits on allowed magnetic field strength for a known background rotational mhd equilibrium . it is also important to note that for the axisymmetric background rotational mhd equilibrium under consideration , the magnetic force arising from the azimuthal magnetic field is radially inward when @xmath88 , radially outward when @xmath89 and is zero when @xmath11 [ i.e. , the azimuthal magnetic field is force - free ( lou & fan 1998a ) and the equilibrium density distribution reduces to that of the single unmagnetized disc case ( lemos et al . 1991 ; syer & tremaine 1996 ) ] . these different possibilities can be physically understood in terms of the competition between the magnetic pressure and tension forces . in order to see this more specifically for the background , we can write @xmath90 $ ] referring to equation ( [ radialmomeneqn ] ) . the first term in the square brackets stands for the magnetic pressure force and the second term stands for the magnetic tension force for a purely azimuthal magnetic field @xmath91 in the cylindrical geometry . as @xmath92 scales @xmath93 , we have @xmath94 . therefore for @xmath95 or equivalently @xmath88 , the magnetic tension force dominates and the total magnetic force is radially inward ; for @xmath96 or equivalently @xmath89 , the magnetic pressure force dominates and the total magnetic force is radially outward . the two magnetic forces balance each other for @xmath61 ( or @xmath11 ) . on the basis of the full mhd equations @xmath97 , we readily derive the linearized coplanar mhd perturbation equations as @xmath98 @xmath99 -\frac{1}{\sigma_0}\int\frac{dzb_{\theta}}{4\pi r } \frac{\partial(rb_0)}{\partial r}\ , \end{split}\ ] ] @xmath100 @xmath101^{1/2}},\ ] ] @xmath102 @xmath103 @xmath104 where we use subscript @xmath105 to denote associations with small disturbances in physical variables and @xmath106 stands for the coplanar magnetic field perturbation . as the background rotational mhd equilibrium is stationary and axisymmetric , these perturbed physical variables can be decomposed in terms of fourier harmonics with the periodic dependence @xmath107 where @xmath108 is the angular frequency and @xmath109 is an integer for azimuthal variations . more specifically , we write @xmath110 where @xmath111 , @xmath112 , @xmath113 , @xmath114 , @xmath115 and @xmath116 are all radial variations of the corresponding perturbed physical variables and can be complex in general for possible radial oscillations . without loss of generality , we take @xmath117 in our analysis . with fourier decomposition ( [ fourier ] ) , it is then straightforward to cast coplanar mhd perturbation equations @xmath118 into the following forms of @xmath119 @xmath120\ , \end{split}\ ] ] where @xmath121 is a short - hand notation , and @xmath122 @xmath123 @xmath124 @xmath125 @xmath126 where equation ( [ perturb07 ] ) can be derived by combining equations ( [ perturb05 ] ) and ( [ perturb06 ] ) . we now rearrange the time - dependent coplanar mhd perturbation equations by taking @xmath127 from equations ( [ perturb01])@xmath128([perturb07 ] ) ; and the special case of @xmath129 will be analyzed in details at the end of this subsection . from equations ( [ perturb05 ] ) and ( [ perturb06 ] ) , we readily obtain @xmath130 a substitution of equation ( [ zr ] ) into the radial and azimuthal components of the momentum equation ( [ perturb02 ] ) and ( [ perturb03 ] ) yields @xmath131 \bigg(\frac{u}{\omega - m\omega}\bigg)\ , \end{split}\ ] ] and @xmath132 respectively . in order to construct global solutions without the wkbj approximation , we are mainly interested in stationary configurations with zero pattern speed @xmath133 , also referred to as neutral @xmath134 modes ( e.g. , syer & tremaine 1996 ; shu et al . 2000 ; lou 2002 ; shen & lou 2004a , b ; lou & zou 2004a ) . by setting @xmath134 and taking @xmath127 , equations ( [ perturb01 ] ) , ( [ zr ] ) , ( [ radial ] ) and ( [ azimuthal ] ) can be readily reduced to @xmath135 @xmath136 @xmath137\frac{\hbox{i}u}{\omega}\ , \end{split}\ ] ] @xmath138 equations ( [ stationary1 ] ) through ( [ stationary4 ] ) together with equation ( [ perturb04 ] ) are the basic mhd perturbation equations for constructing stationary non - axisymmetric @xmath127 configurations of both aligned and unaligned logarithmic spiral cases . note that aligned and spiral stationary global solutions both involve propagations of fast and slow mhd density waves ( lou 2002 ) . we next present the basic coplanar mhd perturbation equations for stationary axisymmetric @xmath129 configurations with or without radial propagations . starting from equations ( [ perturb01 ] ) , ( [ perturb06 ] ) , ( [ perturb07 ] ) , ( [ radial ] ) and ( [ azimuthal ] ) by setting @xmath129 and @xmath139 , we obtain @xmath140\hbox{i}u\ , \\ & \omega j=\frac{r\kappa^2}{2\omega}\hbox{i}u\ , \\ & r=0\ , \qquad\qquad \omega z=\frac{\partial}{\partial r}(b_0\hbox{i}u)\ . \end{split}\ ] ] this set of equations is applicable as we proceed to investigate stationary axisymmetric coplanar mhd perturbations ( lou & zou 2004a , b ) . for coplanar hydrodynamic perturbations or coplanar mhd perturbations in an isopedically magnetized sid ( shu et al . 2000 ) , there is no essential difference for changing the order of limiting procedure for @xmath129 and @xmath141 . however , for msid with a coplanar azimuthal background magnetic field , the results would be different by changing the order of limiting procedure for @xmath129 and @xmath141 ( lou 2002 ; lou & fan 2002 ; lou & zou 2004a ) . to solve the poisson equation connecting the perturbed surface mass density and the perturbed gravitational potential , we may consistently assume disturbances to carry proper scale - free forms ( lynden - bell & lemos 1993 ) for aligned and logarithmic spiral perturbations . for aligned perturbations , @xmath111 contains only an amplitude variation in @xmath14 of perturbed surface mass density and does not involve phase variation in @xmath14 so that the maximum density perturbations at different radii line up in the azimuth . for spiral perturbations in comparison , in addition to an amplitude variation in @xmath14 , @xmath111 also involves a phase variation in @xmath14 such that a spiral pattern emerges . for aligned global coplanar mhd perturbations , we select those perturbations that carry the same power - law variation in @xmath14 of the background equilibrium , namely @xmath142 where @xmath143 is a small constant coefficient and the numerical factor @xmath58 is given explicitly by @xmath144 where @xmath145 ( qian 1992 ; syer & tremaine 1996 ; shen & lou 2004a , b ) . in the limit of @xmath60 , we simply have @xmath146 consistent with earlier results on sids and msids ( shu et al . 2000 ; lou 2002 ; lou & shen 2003 ; lou & zou 2004a ; lou & wu 2004 ) . in fact , a more general class of self - consistent potential - density pairs satisfying the poisson integral ( [ perturb04 ] ) is also available ; see footnote 3 in shen & lou ( 2004b ) . for aligned neutral modes of axisymmetry ( @xmath129 ) , we can start from equation ( [ m0 ] ) by setting @xmath134 ; this implies @xmath147 and gives no constraints on @xmath148 . as @xmath149 , the scale - free condition requires @xmath150 with @xmath151 and @xmath152 . it turns out that such stationary coplanar mhd perturbations are simply alternative states for the axisymmetric background equilibrium with a proper rescaling factor ( shu et al . 2000 ; lou 2002 ) . as this is somewhat trivial , we now turn our attention mainly to cases of non - axisymmetric @xmath127 stationary aligned coplanar mhd perturbations . for the potential - density pair @xmath149 and @xmath153 , we have @xmath154 and @xmath152 according to equations @xmath155 . it then follows that @xmath156}{2m\omega r^2}\hbox{i}u\ , \\ & m\omega j+r(1-\beta)\omega \hbox{i}u=\frac{ma^2}{r\sigma_0}\bigg(\frac{r\sigma_0}{a^2}2\pi g { \cal p}_m-1\bigg)rs+\frac{c_a^2(1 - 4\beta)}{2\omega r}\hbox{i}u\ , \\ & z=-\bigg(\frac{1}{2}-2\beta\bigg)\frac{b_0\hbox{i}u}{m\omega r}\ , \qquad\qquad\qquad r=\frac{b_0u}{\omega r}\ . \end{split}\ ] ] rearranging the first three equations of ( [ alignedmge1 ] ) , we obtain a set of three algebraic equations , namely @xmath157 where coefficients @xmath158 , @xmath159 , @xmath160 ( @xmath161 ) are explicitly defined by @xmath162\ , \quad\qquad b_2\equiv m\omega+\frac{[(1 - 4\beta)^2 - 2m^2]c_a^2}{2m\omega r^2}\ , \quad\qquad c_2\equiv\frac{2\omega}{r}\ , \\ & a_3\equiv -mr\bigg[2\pi g{\cal p}_m(\beta ) -\frac{a^2}{\sigma_0 r}\bigg]\ , \qquad\qquad b_3\equiv r(1-\beta)\omega-\frac{(1 - 4\beta)c_a^2}{2\omega r}\ , \qquad\qquad c_3\equiv m\omega\ . \end{split}\ ] ] for nontrivial solutions of @xmath163 , the determinant of algebraic equations ( [ suj ] ) must vanish , namely @xmath164 using coefficient definitions ( [ abc ] ) in condition ( [ aligned0 ] ) , we obtain a more informative form of stationary dispersion relation @xmath165}{[m^2\omega^2r^2 -c_a^2(m^2 - 1/2 - 8\beta^2 + 4\beta ) ] [ ( 1 + 2\beta)\omega^2r^2-c_a^2(1/2 - 2\beta)]}\ \\ & \qquad\qquad\qquad -\frac{4\beta a^2 - 8\beta\pi g{\cal p}_m\sigma_0r + ( 1 - 4\beta)c_a^2}{2[m^2\omega^2r^2 -c_a^2(m^2 - 1/2 - 8\beta^2 + 4\beta)]}\ . \end{split}\ ] ] note that equation ( [ aligned1 ] ) takes a simpler form in the limit of @xmath60 ( i.e. , a flat rotation curve of a sid ) and reduces to equation ( 3.1.9 ) of lou ( 2002 ) as expected . by using background mhd equilibrium variables ( [ propequilib ] ) in condition ( [ aligned1 ] ) , we obtain the stationary dispersion relation for global aligned @xmath127 coplanar mhd configurations in the form of a quadratic equation in terms of @xmath166 , namely @xmath167 where the three coefficients @xmath168 ( @xmath161 ) are defined by @xmath169q^2 -{\cal a}_m(1-f { \cal c}{\cal p}_m)\ , \\ & c_0\equiv -\frac{1 - 4\beta}{4(1 + 2\beta ) } \bigg[\frac{-2m^2+(1 - 4\beta)(1 - 2\beta)}{1 + 2\beta } f{\cal c}{\cal p}_m+1 - 4\beta\bigg]q^4 -\frac{-2m^2+(1 - 4\beta)(1 - 2\beta)}{2(1 + 2\beta ) } ( 1-f{\cal c}{\cal p}_m)q^2\ , \end{split}\ ] ] with auxiliary parameters explicitly defined by @xmath170\ . \end{split}\ ] ] which are exactly the same as those defined in shen & lou ( 2004b ) for aligned cases of global perturbations in a purely hydrodynamic composite disc system . here , the effect of coplanar magnetic field is associated with the @xmath81 parameter in definitions ( [ alignc2c1c0 ] ) . the determinant of quadratic equation ( [ aligned ] ) , @xmath171 , depends on various parameters and can become negative under some circumstances ( i.e. , no physical solutions for @xmath166 ) . for @xmath172 , the two real solutions of @xmath166 to equation ( [ aligned ] ) are @xmath173 as a first check of necessary consistency , we consider the limiting case of vanishing magnetic field , that is , @xmath174 . then , equation ( [ aligned ] ) simply reduces to @xmath175 which gives a non - trivial solution @xmath176 this @xmath87 solution ( [ alignednomsol ] ) is exactly the same result for the case of a single scale - free gas disc without magnetic field ( see syer & tremaine 1996 and subsection 3.1 of shen & lou 2004b ) . as a second check of necessary consistency , we consider another limiting case of @xmath60 which corresponds to the msid case with a flat rotation curve studied by lou ( 2002 ) . one can readily show that equation ( [ aligned ] ) reduces to equation ( 3.2.11 ) of lou ( 2002 ) as expected . in contrast to the case of a single fluid disc studied by syer & tremaine ( 1996 ) without azimuthal magnetic fields , there are now two branches of @xmath87 solutions of algebraic equation ( [ aligned ] ) in general . a @xmath87 solution is considered to be physical when it satisfies both the condition @xmath86 and inequality ( [ restrt ] ) . when the magnetic field becomes sufficiently weak , among these two branches of @xmath87 solutions , @xmath177 is the counterpart of the single disc case and @xmath178 is additional due to the coupling between the surface mass density and magnetic field . we recall that in a composite system of one stellar disc and one gaseous disc coupled through the mutual gravitational interaction , there are also two branches of @xmath87 solutions ( lou & fan 1998b , 2000 ; lou & shen 2003 ; shen & lou 2004b ) . in a composite stellar - gaseous disc system , the two different classes of perturbation modes correspond to either in - phase or out - of - phase of surface mass density perturbation enhancements ( see also lou & wu 2004 ) . in a single azimuthally magnetized disc , the two coplanar mhd modes @xmath177 and @xmath178 will be distinguished by either in - phase or out - of - phase perturbation enhancements of the surface mass density and the azimuthal magnetic field in the wkbj or tight - winding approximation . in the wkbj regime , the @xmath177 and @xmath178 branches correspond to stationary fast and slow mhd density waves , respectively ( fan & lou 1996 ; lou & fan 1998a ) . although the phase relationships for perturbation enhancements of the surface mass density and the azimuthal magnetic field become @xmath179 for open spiral structures ( lou 2002 ; lou & fan 2002 ) , we shall still refer to @xmath177 and @xmath178 solutions as global stationary fast and slow mhd density waves , respectively . we shall come back to this discussion of ` phase relationship ' in more details in section 4 . as an example of illustration , we now focus on the specific case of @xmath11 with inequality ( [ restrt ] ) automatically satisfied . in such a magnetized disc system , the rotation curve scales as @xmath180 , the surface mass density scales as @xmath181 , the barotropic index is @xmath182 and the azimuthal magnetic field scales as @xmath183 ( i.e. , the background equilibrium magnetic field is force - free ) . for this specific case , the determinant @xmath184 of algebraic equation ( [ aligned ] ) remains always positive for both full and partial discs ( see proofs in appendix a ) . as seen from @xmath87 solutions ( [ solalign ] ) , the sign of @xmath185 determines whether @xmath177/@xmath178 is the upper / lower branch or reverse . for @xmath186 , we always have @xmath187 and therefore @xmath177 and @xmath178 remain to be the higher and lower branches , respectively . the situation becomes somewhat involved for the @xmath188 case when the sign of @xmath185 is dependent on the value of @xmath189 parameter . we show in fig . 1 the @xmath166 solution branches for @xmath188 and @xmath190 cases , respectively , versus the gravitational potential ratio @xmath189 . for the @xmath188 case , there exists a diverging point at @xmath191 $ ] for @xmath192 solution branch where @xmath193 . the value of this critical @xmath194 is numerically determined to be @xmath195 . more specifically , we have @xmath196 for @xmath197 , while we have @xmath187 for @xmath198 . this sign variation of @xmath185 is indicated by the relative locations of @xmath177 and @xmath178 solution branches as shown in panel @xmath199 of fig . in other words , for @xmath200 , @xmath177 remains negative and the only physical solution is @xmath178 . on the other hand for @xmath198 , both @xmath177 and @xmath178 solution branches become physically plausible . for @xmath11 , the explicit form of algebraic equation ( [ aligned ] ) becomes @xmath201d^4 -\bigg[\bigg(\frac{2{f}{\cal c}{\cal p}_m}{3 } + 1\bigg)m^2q^2+\bigg(m^2+\frac{3}{4}\bigg)(1-{f } { \cal c}{\cal p}_m)\bigg]d^2+\frac{2}{3}(1-{f } { \cal c}{\cal p}_m)m^2q^2=0\ , \ ] ] where auxiliary parameters @xmath202 , @xmath203 , @xmath204 , @xmath205 and @xmath206 are all evaluated at @xmath11 . it is fairly straightforward to show that @xmath207 for all @xmath127 when @xmath11 ( shen & lou 2004b ) . as @xmath208 , we then have @xmath209 . in our formulation , stationary surface mass density perturbations are characterized by @xmath210 . for aligned coplanar mhd perturbations , we took @xmath211 where @xmath212 is a real exponent . in section 3.1 , we have chosen @xmath213 that carries the same power - law dependence as the equilibrium disc does . for a complex @xmath212 , a perturbation pattern would appear spiral , that is , a logarithmic spiral in the form of @xmath214 $ ] where @xmath215 and @xmath216 are real and imaginary parts of @xmath212 . to ensure the gravitational potential arising from this perturbed surface mass density being finite , we should require @xmath217 ( e.g. , qian 1992 ) . in fact , there exists a more general class of self - consistent potential - density pairs as indicated in footnote 7 of shen & lou ( 2004b ) . however , for the analysis of axisymmetric stability problem in section 3.2.1 , equation ( [ spirperturb ] ) is used such that dispersion relation ( [ m0u ] ) derived later on is real on both sides ( e.g. , lemos et al . 1991 ; syer & tremaine 1996 ; shu et al . 2000 ) . to be specific here , we consistently take the logarithmic spiral surface mass density perturbations and the resulting gravitational potentials as ( kalnajs 1971 ; lemos et al . 1991 ; shu et al . 2000 ; lou 2002 ; lou & shen 2003 ; shen & lou 2004b ; lou & zou 2004a ) @xmath218 where @xmath143 is a small constant , @xmath219 is a parameter related to the radial wavenumber and the kalnajs function @xmath220 is defined by @xmath221 ( kalnajs 1971 ) . as @xmath220 is an even function of @xmath219 , a consideration of @xmath222 would suffice . in our convention of notations , @xmath223 and @xmath224 correspond to leading and trailing logarithmic spiral waves , respectively , and @xmath225 relates the parameter @xmath219 to the radial wavenumber @xmath226 . we note that @xmath220 decreases monotonically with increasing @xmath223 , and @xmath227 for @xmath127 ; for @xmath129 , @xmath228 is positive and can be greater than 1 when @xmath219 becomes sufficiently small . the choice of unaligned perturbations is not unique , and the perturbation potential - density pairs may involve @xmath8 parameter such that the background plus coplanar mhd perturbations are altogether scale - free ( e.g. , syer & tremaine 1996 ) . as an example of illustration , we here presume coplanar logarithmic spiral perturbations defined by the density - potential pair ( [ spirperturb ] ) in the following analysis . for the specific @xmath11 case , the background surface mass density and the perturbed surface mass density bear the same radial scaling @xmath229 . we start from the axisymmetric coplanar mhd perturbation equation ( [ m0 ] ) with @xmath139 . for radial oscillations with a logarithmic potential - density pair , we presume a surface mass density @xmath230 , a gravitational potential @xmath231 and thus a radial speed @xmath232 according to the first of equation ( [ m0 ] ) . after proper rearrangements , the set of equations ( [ m0 ] ) can be combined to form a single equation in terms of @xmath112 , namely @xmath233u\\ & = \frac{a^2}{r^2}\bigg\{\bigg[1-{f}{\cal c } { \cal n}_0\bigg(1+d^2-\frac{1 - 4\beta}{2 + 4\beta}q^2\bigg)\bigg ] \bigg(\frac{1}{4}+\xi^2\bigg ) + q^2\bigg(\frac{1}{4}+\xi^2-\beta\bigg)\bigg\}u\ , \end{split}\ ] ] where @xmath234 , @xmath235 $ ] and @xmath236 is extremely small . by inspection , equation ( [ m0u ] ) bears similar features of the dispersion relation in the classic wkbj regime involving magnetic field . as the right - hand side of equation ( [ m0u ] ) is real , axisymmetric instabilities first set via neutral @xmath134 modes . it should be emphasized that the scale - free condition is met only for neutral modes . with @xmath237 , the marginal stability curves are then given by @xmath238\}(\xi^2 + 1/4)+q^2(\xi^2-\beta+1/4)}{{f } { \cal c}{\cal n}_0(\xi^2 + 1/4)+(1 + 2\beta)(-2 + 2\beta)}\ .\ ] ] as a check of necessary consistency , the axisymmetric marginal stability curve for an msid when @xmath19 ( @xmath239 ) then gives @xmath240 ( \xi^2 + 1/4)}{{f}{\cal n}_0(\xi^2 + 1/4)-2}\ .\ ] ] one can readily show ) is automatically satisfied for @xmath85 and we need only to examine the range of @xmath88 . when the denominator of the right - hand side of equation ( [ msc ] ) is negative , the numerator must also be negative for @xmath86 . it then follows that @xmath241 and inequality ( [ restrt ] ) is met . when the denominator of the right - hand side of equation ( [ msc ] ) is positive , it is easy to show that inequality ( [ restrt ] ) holds . as solution ( [ mscbeta0 ] ) is a special case of solution ( [ msc ] ) , inequality ( [ restrt ] ) is thus met . ] that stationary dispersion relations ( [ msc ] ) and ( [ mscbeta0 ] ) automatically satisfy inequality ( [ restrt ] ) for physical solutions @xmath86 . in the msid case with different values of @xmath77 parameter for magnetic field strengths , one readily obtains the marginal stability curves that separate two distinct unstable regimes , namely , the collapse regime for large radial spatial scales and the ring fragmentation regime for relatively short radial wavelengths ( lemos et al . 1991 ; shen & lou 2003 , 2004a ) . there exists a critical @xmath242 where @xmath87 diverges and this @xmath243 is determined by @xmath244 . for @xmath245 , we have @xmath246 and thus an enhancement of @xmath77 tends to suppress the collapse boundary . for @xmath247 , we have @xmath248 and thus an enhancement of @xmath77 tends to raise the ring fragmentation boundary . in short , an enhancement of ring magnetic field tends to reduce dangers of both collapse and ring fragmentation instabilities in the msid case . while the enhancement of ring magnetic field tends to suppress ring fragmentation instabilities in general , it only suppresses collapse instabilities for @xmath88 and tends to aggravate collapse instabilities for @xmath89 ( see fig . 2 for details ) . it has been shown that the magnetic force is radially inward and outward for @xmath88 and @xmath89 , respectively . why does the trend of variations for the collapse - stability appear seemingly paradoxical ? this situation can be understood because the unperturbed background is in an mhd radial force balance and the surface mass density , the rotation speed and the magnetic field are coupled explicitly through condition ( [ propequilib ] ) . furthermore , for fast mhd density waves with @xmath129 ( lou & fan 1998a ) , the gas pressure and magnetic pressure together is associated with the radial wavenumber squared , while the surface mass density and self - gravity are associated linearly with the radial wavenumber ( see the first line of dispersion relation [ m0u ] with small @xmath236 ) . for @xmath88 , the surface mass density becomes smaller for stronger magnetic field strength , while for @xmath89 , the situation reverses . for the ring fragmentation instability at relatively large radial wavenumbers , the dominant magnetic pressure effect tends to stabilize the disc . for the jeans collapse instability at relatively small radial wavenumbers , the dominant self - gravity effect is proportional to the background surface mass density ; the roles of magnetic field for two different situations can then be readily understood . it is the coupling effect between the surface mass density and the magnetic field of the background that gives rise to this collapse feature . the marginal stability curves for scale - free discs with @xmath249 and without magnetic fields are @xmath250 that are consistent with those for the single - disc case of shen & lou ( 2004b ; e.g. , see their subsection 3.2.1 ) . the marginal axisymmetric stability curves generally consist two branches , namely , the collapse branch and the ring - fragmentation branch ( lemos et al . 1991 ; shu et al . 2000 ; lou 2002 ; lou & shen 2003 ; shen & lou 2003 ) . the lowest value of @xmath87 for stability is the maximum of the collapse branch and the highest value of @xmath87 for stability is the minimum of the ring - fragmentation branch . we note that the maximum of the collapse branch is always located at @xmath251 for @xmath252 , while for @xmath253 the maximum of the collapse branch may locate at @xmath223 with @xmath251 corresponding to a local minimum of @xmath87 . it was proven in appendix c of shen & lou ( 2004b ) that @xmath251 is always a local extremum for @xmath87 . variations of the stable range for @xmath87 with parameters @xmath254 , @xmath8 and @xmath77 are shown in panels ( a ) and ( b ) of fig . finally we note that the axisymmetric destabilization involves only the fast mhd wave branch because the slow mhd wave branch is negative and thus unphysical . for non - axisymmetric ( @xmath255 ) perturbations , both fast and slow mhd density waves are possible and stationary global configurations may represent transitions from stable to unstable situations ( e.g. , shu et al . 2000 ) . here , the stability of our slow mhd density waves differs from those models involving finite disc thickness where either vertical or horizontal weak magnetic fields may induce velikhov - chandrasekhar - balbus - hawley instability and magneto - rotational instabilities ( mri ) through slow mhd modes ( chandrasekhar 1961 ; balbus & hawley 1998 ; kim & ostriker 2000 ; lou , yuan & fan 2001 ) . it would be interesting to recall the axisymmetric stability analysis in the wkbj approximation . the local dispersion relation for fast mhd waves in an azimuthal magnetic field is ( fan & lou 1996 ) @xmath256 where @xmath226 is the radial wavenumber and other notations have the conventional meanings . equation ( [ drwkbj ] ) gives the marginal stability curve for axisymmetric @xmath129 perturbations as @xmath257 by using scale - free disc equilibrium condition ( [ propequilib ] ) and inserting @xmath258 for large @xmath226 and @xmath219 into dispersion relation ( [ stability ] ) , we readily obtain the corresponding marginal stability curve in the wkbj regime as @xmath259\xi } { f{\cal c}\xi+(1 + 2\beta)(-2 + 2\beta)}\ , \ ] ] which bears a strong resemblance to expression ( [ msc ] ) of our global analysis , especially in view of the asymptotic expression for @xmath220 , namely @xmath260 when @xmath261 . another equivalent way of this comparison is to note the parallel terms in the wkbj condition ( [ stability ] ) and in the first line of the corresponding global condition ( [ m0u ] ) , also in reference to the above asymptotic expression for @xmath220 . a specific illustrating example of such a comparison is shown in fig . 3 with parameters @xmath19 ( @xmath239 ; flat rotation curve ) , @xmath35 ( full disc ) and @xmath262 . in the short - wavelength limit ( i.e. , @xmath263 ) , the wkbj approximation is well justified and the results are qualitatively consistent , while in the long - wavelength of jeans collapse regime , the wkbj analysis differs from the global exact treatment significantly . , @xmath35 and @xmath262 . the marginal stability curves for the long - wavelength jeans collapse and the short - wavelength ring fragmentation regimes are denoted . for @xmath263 ( i.e. , the short - wavelength limit ) , the wkbj approximation is well applicable , while for small @xmath219 the results in the wkbj approximation deviates from those of the exact global treatment considerably ( see shen & lou 2003 , 2004b ) . ] in parallel with the aligned @xmath127 cases , we now construct unaligned logarithmic spiral solutions from coplanar mhd perturbation equations ( [ stationary1])@xmath128([stationary4 ] ) . for the density - potential pair @xmath230 and @xmath264 , we then have radial flow speed @xmath265 and @xmath266component specific angular momentum @xmath267 . coplanar mhd perturbation equations @xmath155 can therefore be written in the forms of @xmath268\hbox{i}u\ , \\ & m\omega j+r(1-\beta)\omega \hbox{i}u = m\bigg(2\pi g{\cal n}_m -\frac{a^2}{r\sigma_0}\bigg)rs+\frac{c_a^2(1 - 4\beta ) } { 2\omega r}\hbox{i}u\ , \\ & z=-\hbox{i}\xi\frac{b_0iu}{m\omega r}\ , \qquad\qquad\qquad\qquad r=\frac{b_0u}{\omega r}\ . \end{split}\ ] ] the first three equations of set ( [ spiralmge1 ] ) immediately lead to @xmath269 where coefficients @xmath158 , @xmath159 and @xmath160 ( @xmath161 ) are explicitly defined by @xmath270c_a^2}{m\omega r^2}\ , \qquad c_2\equiv \frac{2\omega}{r}\ , \\ & a_3\equiv -mr\bigg(2\pi g{\cal n}_m-\frac{a^2}{\sigma_0 r}\bigg)\ , \qquad\qquad b_3\equiv r(1-\beta)\omega-\frac{(1 - 4\beta)c_a^2}{2\omega r}\ , \qquad\qquad c_3\equiv m\omega\ . \end{split}\ ] ] for nontrivial solutions of the set of homogeneous algebraic equations ( [ sujspiral ] ) , the determinant must vanish @xmath271 by using the set of expressions ( [ abcspiral ] ) and applying equilibrium conditions ( [ propequilib ] ) , we obtain the final dispersion relation for stationary logarithmic spiral @xmath127 mhd configurations in the form of a quadratic equation in terms of @xmath166 , namely @xmath272 where the coefficients @xmath185 , @xmath273 and @xmath274 are explicitly defined by @xmath275q^2 -{\cal a}_m(1-{f}{\cal c}{\cal n}_m)\ , \\ & c_0\equiv -\frac{(1 - 4\beta ) } { 4(1 + 2\beta ) } \bigg[\frac{(-2m^2 - 2\xi^2 - 2\beta+1/2)}{(1 + 2\beta ) } { f}{\cal c}{\cal n}_m+1 - 4\beta\bigg]q^4 -\frac{(-2m^2 - 2\xi^2 - 2\beta+1/2)}{2(1 + 2\beta ) } ( 1-{f}{\cal c}{\cal n}_m)q^2\ , \end{split}\ ] ] with auxiliary parameters defined by @xmath276\ . \end{split}\ ] ] that are the same as those in shen & lou ( 2004b ) for unaligned logarithmic spiral cases . for a positive determinant @xmath277 , the two real solutions to quadratic equation ( [ spiral ] ) are given by @xmath278 similar to @xmath127 cases for aligned perturbations in the limit of vanishing magnetic field ( i.e. , @xmath279 ) , equation ( [ spiral ] ) reduces to @xmath280 that leads to a nontrivial solution @xmath281 solution ( [ spiralnomsol ] ) is exactly the same result for the single - disc case discussed in section 3.2 of shen & lou ( 2004b ) . another limiting regime of @xmath60 has been analyzed by lou ( 2002 ) for global stationary perturbation structures in a single msid . for stationary spiral mhd configurations , we here analyze again the @xmath11 case as a specific example of illustration . the background equilibrium ring magnetic field is force - free and inequality ( [ restrt ] ) is satisfied . the determinant @xmath184 of quadratic equation ( [ spiral ] ) remains always positive as shown in appendix a and hence two real solutions of @xmath282 to equation ( [ spiral ] ) exist . in parallel with the aligned @xmath11 case , we solve quadratic equation ( [ spiral ] ) for given values of parameter @xmath283 and display in diagrams the two branches of solutions @xmath177 and @xmath178 ( @xmath166 ) versus @xmath219 variation . as @xmath284 for @xmath188 and when @xmath283 becomes sufficiently small , there is a diverging point for @xmath192 solution where @xmath193 . since @xmath285 increases monotonically with increasing @xmath219 for @xmath222 and attains its minimum value at @xmath251 ( @xmath285 is an even function of @xmath219 ; see appendix c of shen & lou 2004b ) , the existence of such a critical point @xmath243 for a diverging @xmath166 requires @xmath286 for the specific @xmath11 case , this inequality ( [ criticalf ] ) implies a critical @xmath283 value of @xmath287 $ ] below which there exists one divergent point of @xmath243 where @xmath177 diverges . numerically , we have determined @xmath288 [ note that this @xmath289 is of the same value as that of the aligned @xmath188 case because @xmath290 . for selected @xmath254 values in the range of @xmath291 , the corresponding critical @xmath243 at which @xmath292 can be numerically computed ; for example , we have @xmath293 for @xmath294 . it becomes much simpler for cases of @xmath186 because @xmath185 remains always positive and there is no diverging point for @xmath87 solutions . therefore , @xmath177 and @xmath178 remain always as the upper and lower branches , respectively . for @xmath188 and @xmath190 cases with different parameters of @xmath283 and @xmath77 , we show in four panels @xmath295 of fig . 4 relevant curves of @xmath166 versus @xmath219 variation to illustrate the basic features . to be more explicit for the case of @xmath11 in quadratic equation ( [ spiral ] ) , we write @xmath296d^4 -\bigg[(m^2+\xi^2)\bigg(\frac{2}{3 } { f}{\cal c}{\cal n}_m+1\bigg)q^2\\ & \hbox { } \qquad\qquad\qquad\qquad\qquad\qquad\qquad + \bigg(m^2+\xi^2+\frac{3}{4}\bigg ) ( 1-{f}{\cal c}{\cal n}_m)\bigg]d^2 + \frac{2}{3}(m^2+\xi^2)(1-{f}{\cal c}{\cal n}_m)q^2=0\ , \end{split}\ ] ] where auxiliary parameters @xmath297 , @xmath203 , @xmath204 , @xmath205 and @xmath206 are all evaluated for the special value of @xmath11 . for multi - wavelength observations of nearby spiral galaxies ( sofue et al . 1986 ; kronberg 1994 ; beck et al . 1996 ) , it is possible to identify large - scale spatial phase relationships of spiral patterns using various observational diagnostics ( e.g. , mathewson et al . 1972 ; visser 1980a , b ; neininger 1992 ; beck & hoernes 1996 ; frick et al . 2000 , 2001 ; ferrire 2001 ; lou et al . 2002 ) . for this reason , we derive below spatial phase relationships for various coplanar mhd perturbation variables , keeping in mind of potential applications to magnetized spiral galaxies ( e.g. , fan & lou 1996 ; lou & fan 1998a ) . here we systematically examine the spatial phase relationships among enhancements of gas surface mass density , magnetic field and velocity perturbations by analysing the relationships among @xmath298 , @xmath148 , @xmath73 , @xmath299 and @xmath300 for aligned and logarithmic spiral cases , separately . starting from the set of homogeneous algebraic equations ( [ suj ] ) for the aligned cases , we obtain the relationships among @xmath298 , @xmath301 and @xmath299 in the following , namely @xmath302 where @xmath158 , @xmath159 and @xmath160 ( @xmath161 ) are defined by expressions ( [ abc ] ) . by using definitions ( [ abc ] ) and equations ( [ alignedmge1 ] ) , we obtain @xmath303 for the aligned cases , where the common dimensionless real factor @xmath304 is defined below by @xmath305 similarly , by using the set of homogeneous algebraic equations ( [ sujspiral ] ) , definitions ( [ abcspiral ] ) and equations ( [ spiralmge1 ] ) , we readily derive for the cases of global stationary logarithmic spiral configurations @xmath306\ , \end{split}\ ] ] where the common dimensionless complex factor @xmath307 is defined below by @xmath308 the two sets of algebraic equations ( [ relsurjz ] ) and ( [ relsurjzspiral ] ) are the most general phase relationships for aligned and logarithmic spiral cases , respectively . once a @xmath8 value is specified and a particular @xmath87 for a stationary solution inserted , we immediately obtain the phase relationship between any two perturbation variables . as an example of illustration , we examine again the special @xmath11 case and determine the relevant phase relationships in the next subsection . when @xmath11 , the results of phase relations derived in the last subsection 4.1 can be simplified considerably as all terms involving the factor @xmath309 vanish . meanwhile , inequality ( [ restrt ] ) is automatically satisfied . the only requirement for a physical @xmath87 solution is simply @xmath86 . as already noted in subsections 3.1.3 and 3.2.3 , there are two branches of @xmath87 solutions . we insert relevant physical @xmath87 solutions in equations ( [ relsurjz ] ) and ( [ relsurjzspiral ] ) to determine phase relationships among perturbation variables . in particular , we are interested in the phase relationship between the perturbed surface mass density and the perturbed azimuthal magnetic field . we now analyze below this mass - magnetic field phase relationships for aligned and spiral cases , separately . some illustrative examples are shown in fig . 5 and fig . 6 to provide more direct visual impressions . for aligned @xmath11 cases , we readily find @xmath310 which means the azimuthal magnetic field remains unchanged . from equation ( [ relsurjz ] ) , we derive @xmath311 meanwhile , equation ( [ align0d25 ] ) can be rewritten in the form of @xmath312 } = \frac{d^2\{d^2-(1-f{\cal c}{\cal p}_m)/ [ f{\cal c}{\cal p}_m+3(m^2 - 3/2)/(2m^2 + 3/2)]\ } } { d^2-(1-f{\cal c}{\cal p}_m)/(f{\cal c}{\cal p}_m+3/2 ) } \ , \ ] ] that contains two distinct @xmath87 solutions according to subsection 3.1.3 , namely , the plus - sign solution @xmath177 and the minus - sign solution @xmath178 in the solution form ( [ solalign ] ) , respectively . as shown in appendix b , we emphasize that the plus - sign solution @xmath177 ( if positive and thus physical ) makes @xmath313 and the minus - sign solution @xmath178 makes @xmath314 on the right - hand side of equation ( [ q2aligned ] ) , respectively . therefore , the two branches of @xmath87 solutions ( if positive and thus physical ) will give either @xmath315 for the plus - sign @xmath177 solution or @xmath316 for the minus - sign @xmath178 solution . for spiral cases with @xmath11 , we readily obtain from equations ( [ relsurjzspiral ] ) and ( [ thetas ] ) that @xmath317(3 + 2\hbox{i}\xi)\ .\ ] ] following the same procedure , we can show that the two branches of @xmath87 solutions , namely , the plus - sign solution @xmath177 and the minus - sign solution @xmath178 given by equation ( [ solspiral ] ) , if positive and thus physical , correspond to the factor @xmath318 being positive and negative , respectively ( see appendix b for a detailed proof ) . we now examine the phase relationship between coplanar perturbation enhancements of surface mass density and magnetic field . in the short - wavelength limit ( i.e. , @xmath263 with a logarithmic spiral pattern tightly wound ) , we have @xmath319 and @xmath320 where the plus- and minus - signs correspond to the @xmath177 and @xmath178 solutions , respectively . the phase relations between perturbation enhancements of surface mass density and magnetic field are in - phase for the @xmath177 solution ( if positive and physical ) and out - of - phase for the @xmath178 solution . for @xmath186 , @xmath177 remains always greater than @xmath178 . in contexts of magnetized spiral galaxies , these solution properties in the wkbj regime are consistent with the physical identifications of the @xmath177 mode with a stationary fast mhd density wave and the @xmath178 mode with a stationary slow mhd density wave , respectively ( fan & lou 1996 ; lou & fan 1998a ; lou & fan 2002 ) . it is now possible to model magnetized spiral galaxies in terms of coplanar stationary fast and slow mhd density waves with more general rotation curves including the case of a flat rotation curve for a magnetized singular isothermal disc ( shu et al . 2000 ; lou 2002 ; lou & fan 2002 ; lou & zou 2004a , b ; lou & wu 2004 ) . in the long - wavelength limit ( i.e. , @xmath321 with a logarithmic spiral pattern widely open ) , we have @xmath322 and @xmath323 where the minus- and plus - signs correspond to the @xmath177 and @xmath178 solutions , respectively . therefore , the coplanar perturbation enhancement of magnetic field is either ahead of or lagging behind the perturbation enhancement of surface mass density by a phase difference of @xmath324 . for intermediate radial wavelengths , the phase difference between perturbation enhancements of surface mass density and magnetic field in a logarithmic spiral pattern can be readily determined for given specific conditions . on the basis of our model analysis here , it appears that except for global stationary fast mhd density waves in the extreme wkbj regime , magnetic and gas spiral arms tend to be phase shifted relative to each other for a global stationary logarithmic spiral pattern for a much broader set of rotation curves including the special case of a flat rotation curve . one can readily show that for aligned global stationary mhd perturbation configurations , there is no net radial angular momentum flux as aligned stationary mhd density waves do not propagate radially ( shu et al . 2000 ; lou 2002 ) . we here briefly discuss the process of angular momentum transfer in association with the stationary logarithmic spiral mhd modes obtained in this paper . according to the analyses of lynden - bell & kalnajs ( 1972 ) and fan & lou ( 1999 ) , the total angular momentum flux carried by coplanar mhd density waves should contain three separate contributions , namely , the advective transport flux @xmath325 defined by @xmath326 the gravity torque flux @xmath327 defined by @xmath328\bigg[\frac{\partial\re(\phi_1 ) } { \partial r}\bigg]\\ ] ] and the magnetic torque flux @xmath329 defined by @xmath330 the perturbed three - dimensional gravitational potential @xmath331 associated with a fourier component of a coplanar logarithmic spiral perturbation in surface mass density @xmath332 is then @xmath333 where @xmath334 is the cylindrical bessel function of order @xmath109 with an argument @xmath23 ( e.g. , binney & tremaine 1987 ) . in fact , integral ( [ 3dphi1 ] ) gives a simple form of kalnajs function in the disc plane at @xmath335 with the real part of @xmath331 explicitly given by @xmath336 it is fairly straightforward to show that @xmath325 and @xmath329 are independent of @xmath14 where the phase relationships for stationary logarithmic spiral cases derived in section 4.1 have been use . while mathematically tedious , one can further show that @xmath327 is also independent of @xmath14 ( see appendix c ) . therefore , the total angular momentum flux @xmath337 remains constant at all radii , a natural result of angular momentum conservation given that perturbation surface mass density falls off as @xmath181 . therefore , in order to maintain a stationary logarithmic spiral pattern , there must be a source or a sink at the disc center . the perturbation energy flux , however , is zero at all radii for the stationary coplanar mhd configurations [ see equation ( 4.17 ) of fan & lou ( 1999 ) in the wkbj regime ] . the mass flux is also zero at all radii since the first - order mass flux @xmath338 , and the second - order mass flux @xmath339 [ where @xmath340 is the second - order perturbed radial velocity , this equation holds due to the second - order mass conservation ( fan & lou 1999 ) ] . therefore our model result differs from that of spruit ( 1987 ) for stationary logarithmic spiral shocks , where only the gravity of the central object is considered . in spruit s model , there is a net mass accretion through the disc and an effective viscosity parameter `` @xmath7 '' ( shakura & sunyaev 1973 ) may be estimated . physically , the central source of angular momentum required to sustain global stationary logarithmic spiral patterns may be plausibly identified with a fast spinning supermassive black hole ( smbh ) in the nucleus of a spiral galaxy . the kerr spacetime associated with such a spinning smbh excites and sustains mhd density waves in the surrounding disc through a long - range gravitational interaction . for trailing spiral patterns , a net flux of angular momentum is transport radially outward in the disc . conceptually , this process should cause the spinning smbh to gradually slow down , entirely analogous to the spin - down of a rotating neutron star by emitting electromagnetic waves that carry a net angular momentum away . in this paper , we have explored and analyzed stationary perturbation structures in coplanarly magnetized razor - thin scale - free discs in a more general manner . to be specific , we start from a rotational and magnetized background equilibrium of axisymmetry that is dynamically self - consistent with a surface mass density @xmath341 , a rotation curve @xmath2 , a purely azimuthal ( ring ) magnetic field @xmath1 with a vertically integrated barotropic equation of state in the form of @xmath16 . the radial force balance at all radii in an mhd disc implies several simple relationships among these power indices @xmath7 , @xmath8 , @xmath9 and @xmath6 , as stated explicitly in equation ( [ pararelation ] ) . without loss of generality , we can simply use @xmath8 as an independent parameter to specify properties of an important class of radial variations for rotational mhd background equilibria . the allowed range of @xmath8 falls within the interval @xmath71 ( syer & tremaine 1996 ; shen & lou 2004a , b ) with the special case of @xmath19 corresponding to a magnetized singular isothermal disc ( msid ) system ( shu et al . 2000 ; lou 2002 ; lou & fan 2002 ; lou & zou 2004a , b ; lou & wu 2004 ) . for clarity and convenience of our analysis , we introduced several dimensionless parameters . the first parameter is the partial disc parameter @xmath34 for the ratio of the gravitational potential ( @xmath342 ) arising from the disc to that ( @xmath28 ) arising from the entire system ( syer & tremaine 1996 ; shu et al . 2000 ; lou 2002 ) including an axisymmetric halo mass distribution that is involved in the background equilibrium but is presumed to be unresponsive to coplanar mhd disc disturbances ( a massive dark matter halo is an example in mind ) ; for a partial disc system , we have @xmath343 , while for a full disc system , we have @xmath344 by conventional definitions . the second parameter is an effective rotational mach number @xmath82 defined as the ratio of the disc rotation speed to the sound speed yet with an additional scaling factor @xmath345 ; @xmath82 is constant at all radii . the third parameter @xmath77 is a measure for the background ring magnetic field strength defined as the ratio of the azimuthal alfvn speed to the sound speed ; and @xmath77 remains constant at all radii . for @xmath89 , there is no constraint on @xmath77 and the net lorentz force of the background azimuthal magnetic field is radially outward ( the magnetic pressure force is stronger than the magnetic tension force ) to resist the self - gravitation ; for @xmath11 , the background azimuthal magnetic field is force - free ( the magnetic pressure and tension forces cancel each other exactly ) ; for @xmath88 , the net lorentz force of the background azimuthal magnetic field points radially inward ( the magnetic pressure force is weaker than the magnetic tension force ) to aggravate the self - gravity and thus a restriction [ i.e. , inequality ( [ restrt ] ) ] is imposed on the magnetic field strength in order to guarantee that the surface mass density be positive in equation ( [ propequilib ] ) . inequality ( [ restrt ] ) provides a necessary physical criterion for plausible @xmath87 solutions . with such a rotational mhd background equilibrium chosen , we introduce coplanar mhd disturbances to construct global stationary perturbation structures as viewed in an inertial frame of reference . perturbation variables are expressed in terms of fourier components for either aligned or logarithmic spiral pattern forms ( syer & tremaine 1996 ; shu et al . 2000 ; lou 2002 ; lou & shen 2003 ; shen & lou 2003 ; lou & zou 2004a ) . we derive analytical solutions for stationary coplanar mhd perturbation configurations in both aligned and logarithmic spiral cases . we now summarize key results below . 1 . cases of aligned coplanar mhd perturbations 0.2 cm + for aligned cases , we choose perturbation variables that bear the same power - law dependence in radius @xmath14 as the background mhd equilibrium does . for the special aligned case of @xmath129 , the stationary coplanar mhd perturbations actually describe alternative equilibria to the axisymmetric background equilibrium with proper rescaling factors ( shu et al . 2000 ; lou 2002 ; lou & zou 2004a ) and are therefore somewhat trivial . + for aligned cases of @xmath127 , we derive a quadratic equation ( [ aligned ] ) in terms of @xmath166 . the resulting positive @xmath87 is the rotational parameter required for sustaining a stationary nonaxisymmetric mhd perturbation configuration . other parameters involved in quadratic equation ( [ aligned ] ) include @xmath8 , @xmath109 , @xmath283 and @xmath77 . for a general combination of @xmath8 , @xmath109 , @xmath283 and @xmath77 , the determinant @xmath184 of quadratic equation ( [ aligned ] ) may not be always positive definite . on the other hand , for specific values of @xmath8 , equation ( [ aligned ] ) always gives two different real solutions ( see appendix a ) . once we prescribe parameters @xmath8 , @xmath109 , @xmath283 and @xmath77 , the two solutions of @xmath87 can be readily determined . for physical solutions , @xmath87 must be positive and satisfy inequality ( [ restrt ] ) simultaneously . + as a specific case study , we examined the case of @xmath11 . here the background azimuthal magnetic field remains force - free and the @xmath87 solution is free from constraint ( [ restrt ] ) . it has been further shown in appendix a that for this @xmath11 case , equation ( [ aligned ] ) always has two distinct real solutions , the plus - signed solution @xmath177 and the minus - signed solution @xmath178 , as contained in equation ( [ solalign ] ) . all these aspects contribute to a significant simplification of the relevant computational procedures . we derive two branches of @xmath87 solutions and need to require only @xmath346 . there is yet a qualitative difference between the @xmath188 case and the @xmath186 cases . for @xmath188 , there exists one critical value @xmath347 ; for @xmath283 below @xmath289 , @xmath177 and @xmath178 are negative and positive , respectively , while @xmath283 above @xmath289 , both branches of @xmath87 solutions are positive and thus physical . for @xmath186 in comparison , both branches of @xmath87 solutions remain always positive and thus physical . moreover , @xmath177 remains always to be the upper branch for azimuthal stationary fast mhd density waves and represents the hydrodynamic counterpart of the unmagnetized single disc case ; complementarily , @xmath178 remains always to be the lower branch for azimuthal stationary slow mhd density waves ( fan & lou 1996 ; lou & fan 1998a ) and is caused by the very presence of the azimuthal magnetic field ( detailed computations reveal that @xmath178 must be smaller than @xmath348 which seems to be too small for galactic applications ) . when a disc rotates sufficiently fast , it could only support the upper @xmath177 stationary mhd configurations . an increase of magnetic field parameter @xmath77 will raise both physical branches of @xmath87 solutions as displayed in fig . the variation trend of the upper @xmath177 branch appears to be more sensitive . therefore for a much stronger magnetic field , a sufficiently rapidly rotating disc may sustain stationary coplanar mhd configurations in terms of the upper @xmath349 branch . + 0.2 cm 2 . cases of logarithmic spiral configurations for coplanar mhd perturbations 0.2 cm + for cases of stationary spiral perturbation structures , we choose coplanar mhd perturbations in terms of kalnajs logarithmic spirals ( kalnajs 1971 ; lemos et al . 1991 ; shu et al . 2000 ; lou 2002 ; lou & fan 2002 ; lou & shen 2003 ; lou & zou 2004a ; lou & wu 2004 ) . for the special @xmath129 case with radial propagations , it happens that inequality ( [ restrt ] ) is automatically satisfied and the stationary coplanar mhd configurations with @xmath350 obtained here represent marginal stability curves ( syer & tremaine 1996 ; shu et al . 2000 ; shen & lou 2003 , 2004b ; lou & zou 2004b ) . only when the rotational parameter @xmath87 falls within a specific finite range can the magnetized disc be stable against axisymmetric disturbances at all wavelengths . a disc with too slow a rotation speed will succumb to jeans instability in the collapse regime corresponding to long wavelength perturbations , while a disc with too fast a rotation speed will suffer the ring - fragmentation instability corresponding to relatively short wavelength perturbations ( safronov 1960 ; toomre 1964 ; lemos et al . 1991 ; lou & fan 1998a , 2000 ; shu et al . 2000 ; shen & lou 2003 , 2004a ; lou & zou 2004b ) . in comparison to a hydrodynamics disc system , the enhancement of azimuthal magnetic field tends to suppress ring - fragmentation instabilities as expected from the perspective of the mhd @xmath351 parameter ( lou & fan 1998a ; lou 2002 ; lou & zou 2004a , b ) , while an enhancement of azimuthal magnetic field tends to suppress jeans collapse instabilities for @xmath88 , to aggravate jeans collapse instabilities for @xmath89 and to bear no effect on jeans collapse instabilities when @xmath11 . we have provided intuitive interpretations for the roles of magnetic field earlier . + in parallel with the aligned cases , we derive quadratic equation ( [ spiral ] ) in terms of @xmath166 for logarithmic spiral cases of @xmath127 . in addition to the four parameters @xmath8 , @xmath109 , @xmath283 and @xmath77 , we now have one more parameter @xmath219 for the radial wavenumber . while the determinant @xmath184 of quadratic equation ( [ spiral ] ) may not be always positive in general , we prove in appendix a that @xmath352 for the specific @xmath11 case . once the five parameters @xmath8 , @xmath109 , @xmath283 , @xmath77 and @xmath219 are specified , the @xmath87 solutions can be readily obtained . for physical solutions , @xmath87 must be non - negative and also satisfy inequality ( [ restrt ] ) . + as an example of illustration , we study the specific @xmath11 case inequality ( [ restrt ] ) being satisfied automatically . not surprisingly , the aligned cases and the logarithmic spiral cases parallel with each other very well , because of a mere change of wave propagation direction . for the special spiral case of @xmath188 , there again exists the same critical @xmath347 for @xmath283 value as in the aligned @xmath188 case ; for @xmath353 , there is one critical point of @xmath243 where @xmath354 diverges . more specifically , we have unphysical @xmath355 for @xmath356 , while we have physical @xmath357 for @xmath247 . there is no divergent point for @xmath87 solution branch and we always have @xmath358 ( @xmath359 when @xmath174 ) . the solution structures become simpler for @xmath186 cases when both @xmath87 branches of solutions are non - negative and there is no divergent point of @xmath354 . moreover , @xmath177 remains always to be the upper branch for stationary spiral fast mhd density waves and is the hydrodynamic counterpart for those in the unmagnetized single disc case , while @xmath178 remains always to be the lower branch for stationary spiral slow mhd density waves ( fan & lou 1996 ) and is caused by the very presence of the magnetic field ( detailed computations reveal @xmath178 must be smaller than @xmath348 ) . + 0.2 cm 3 . phase relationships among coplanar mhd perturbation variables 0.2 cm + so far , we have constructed global stationary coplanar mhd configurations for both aligned and logarithmic spiral cases in magnetized discs with a range of rotation curves . as expected , the inclusion of azimuthal magnetic field gives rise to one additional distinct physical solution for stationary slow mhd density waves to the problem ( fan & lou 1996 ; lou & fan 1998a ; lou 2002 ; lou & zou 2004a ) ; the alfvnic modes are excluded by our restricted consideration for coplanar mhd perturbations . the two @xmath87 branches of solutions are at least mathematically reasonable and may both have applications to astrophysical disc systems such as magnetized disc galaxies and so forth . these devised mhd disc problems are highly idealized with yet precious global analytical perturbation solutions , and conceptually still belong to a subclass of more general time - dependent mhd perturbation solutions . for applications to magnetized spiral galaxies , it would be of considerable interest to discuss pertinent physical aspects and identify relevant observational diagnostics . + for physical solutions of @xmath86 , the @xmath177 branch would be applicable to mhd discs with relatively fast rotation while the @xmath178 branch would be applicable to mhd discs with relatively slow rotation in general . for spatial phase relationships , the two branches of @xmath86 solutions give different results . we take the specific @xmath11 case as an example of illustration . for corresponding aligned cases , the plus - solution @xmath177 will lead to @xmath315 and the minus - solution @xmath178 will lead to @xmath316 ; there are no azimuthal magnetic field perturbations for aligned @xmath11 cases . for tightly wound logarithmic spiral cases , perturbations of both radial and azimuthal magnetic field are approximately in phase with the perturbation enhancement of the surface mass density for the @xmath177 branch , while they are approximately out of phase for the @xmath178 branch . for open logarithmic spirals , the radial and azimuthal magnetic field perturbations are either ahead of or lagging behind the enhancement of the surface mass by a significant phase difference . these phase relationships together with those associated with flow perturbations ( visser 1980a , b ) are valuable to interpret the spatial phase shifts between optical arms and magnetic arms in lopsided , barred or other spiral galaxies ( mathewson et al . 1972 ; neininger 1992 ; beck & hoernes 1996 ; beck et al . 1996 ; fan & lou 1996 ; lou & fan 1998 , 2000 , 2002 ; frick et al . 2000 , 2001 ; lou et al . + 0.2 cm 4 . angular momentum transfer in steady logarithmic spirals for coplanar mhd perturbation configurations 0.2 cm + stationary logarithmic spiral configurations of mhd density waves carry constant angular momentum flux either outward or inward associated with the advective transport , the gravity torque and the magnetic torque ( lynden - bell & kalnajs 1972 ; goldreich & tremaine 1978 ; fan & lou 1999 ) . therefore , the net angular momentum flux in the entire disc system is conserved and there must be either a source or a sink at the disc center . for the @xmath177 branch with @xmath86 , we find that the total angular momentum flux is inward for leading ( i.e. , @xmath223 ) spiral mhd density waves and is outward for trailing ( i.e. , @xmath224 ) spiral mhd density waves . for the @xmath178 branch in comparison , the total angular momentum flux is outward for leading ( i.e. , @xmath223 ) spiral mhd density waves and is inward for trailing ( i.e. , @xmath224 ) spiral mhd density waves . as a real mhd disc system typically rotates at a relatively large @xmath87 and inequality ( [ restrt ] ) can also rule out small @xmath87 for @xmath360 , the lower @xmath178 branch might be rarely useful in galactic applications . in either case , however , the net mass accretion rate is zero since if we focus on materials between two concentric circles , at any time the influent angular momentum from the inner ring is equal to the effluent angular momentum from the outer ring such that materials inside this belt neither gain nor lose any angular momentum and therefore no net mass accretion occurs . this research has been supported in part by the asci center for astrophysical thermonuclear flashes at the university of chicago under department of energy contract b341495 , by the special funds for major state basic science research projects of china , by the tsinghua center for astrophysics , by the collaborative research fund from the national natural science foundation of china ( nsfc ) for young outstanding overseas chinese scholars ( nsfc 10028306 ) at the national astronomical observatory , chinese academy of sciences , by nsfc grant 10373009 at the tsinghua university , and by the yangtze endowment from the ministry of education through the tsinghua university . affiliated institutions of y.q.l . share this contribution . 9 balbus s. a. , hawley j. f. , 1998 , rev . , 70 , 1 beck r. , brandenburg a. , moss d. , shukurov a. , sokoloff d. , 1996 , ara&a , 34 , 155 beck r. , hoernes p. , 1996 , nat , 379 , 47 binney j. , tremaine s. , 1987 , galactic dynamics . princeton university press , princeton , new jersey caunt s. e. , tagger m. , 2001 , a&a , 367 , 1095 chandrasekhar s. , 1961 , hydrodynamic and hydromagnetic stability . dover , new york elmegreen b. g. , 1995 , mnras , 275 , 944 evans n. w. , read j. c. a. , 1998a , mnras , 300 , 83 evans n. w. , read j. c. a. , 1998b , mnras , 300 , 106 fan z. h. , lou y .- q . , 1996 , nat , 383 , 800 fan z. h. , lou y .- q . , 1999 , mnras , 307 , 645 ferrire k. m. , 2001 , rev . mod . phys . , 73 , 1031 frick p. , beck r. , shukurov a. , sokoloff d. , ehle m. , kamphuis j. , 2000 , mnras , 318 , 925 frick p. , beck r. , berkhuijsen e. m. , patrickeyev i. , 2001 , mnras , 327 , 1145 goodman j. , evans n. w. , 1999 , mnras , 309 , 599 goldreich p. , lynden - bell d. , 1965 , mnras , 130 , 125 goldreich p. , tremaine s. , 1978 , icarus , 34 , 240 jog c. j. , 1996 , mnras , 278 , 209 julian w. h. , toomre a. , 1966 , apj , 146 , 810 kalnajs a. j. , 1971 , apj , 166 , 275 kim w .- , ostriker e. c. , 2000 , apj , 540 , 372 kronberg p. p. , 1994 , reports on progress in physics , 57 , 325 lemos j. p. s. , kalnajs a. j. , lynden - bell d. , 1991 , apj , 375 , 484 lin c. c. , shu f. h. , 1964 , apj , 140 , 646 lou y .- q . , 2002 , mnras , 337 , 225 lou y .- q . , fan z. h. , 1998a , apj , 493 , 102 lou y .- q . , fan z. h. , 1998b , mnras , 297 , 84 lou y .- q . , fan z. h. , 2000 , mnras , 315 , 646 lou y .- q . , fan z. h. , 2002 , mnras , 329 , l62 lou y.q . , walsh w. m. , han j. l. , fan z. h. , 2002 , apj , 567 , 289 lou y. q. , yuan c. , fan z. h. , 2001 , apj , 552 , 189 lou y. q. , yuan c. , fan z. h. , leon s. , 2001 , apj , 553 , l35 lou y .- q . , shen y. , 2003 , mnras , 343 , 750 ( astro - ph/0304270 ) lou y .- q . , wu y. , 2004 , mnras , submitted lou y .- q . , zou y. , 2004a , mnras , 350 , 1220 ( astro - ph/0312082 ) lou y .- q . , zou y. , 2004b , mnras , submitted low b. c. , lou y .- q . , 1990 , apj , 352 , 343 lynden - bell d. , kalnajs a. j. , 1972 , mnras , 157 , 1 lynden - bell d. , lemos j. p. s. , 1993 , preprint ( astro - ph/9907093 ) mathewson d. s. , van der kruit p. c. , brouw w. n. , 1972 , a&a , 17,468 mestel l. , 1963 , mnras , 126 , 553 neininger n. , 1992 , a&a , 263 , 30 qian e. , 1992 , mnras , 257 , 581 rdiger g. , kitchatinov l. l. , 2000 , astro . nachr . , 3 , 181 safronov v. s. , 1960 , ann . , 23 , 979 shakura n. i. , sunyaev r. a. , 1973 , a&a , 24 , 337 shen y. , lou y .- q . , 2003 , mnras , 345 , 1340 ( astro - ph/0308063 ) shen y. , lou y .- q . , 2004a , chjaa , in press ( astro - ph/0404190 ) shen y. , lou y .- q . , 2004b , mnras , 353 , 249 ( astro - ph/0405444 ) shu f. h. , laughlin g. , lizano s. , galli d. , 2000 , apj , 535 , 190 shu f. h. , li z .- y . , 1997 , apj , 475 , 251 sofue y. , fujimoto m. , wielebinski r. , 1986 , ara&a , 24 , 459 spruit h. c. , 1987 , a&a , 184 , 173 syer d. , tremaine s. , 1996 , mnras , 281 , 925 tagger m. , pellat r. , 1999 , a&a , 349 , 1003 toomre a. , 1964 , apj , 139 , 1217 toomre a. , 1977 , ara&a , 15 , 437 visser h. c. d. , 1980a , a&a , 88 , 149 visser h. c. d. , 1980b , a&a , 88 , 159 widrow l. m. , 2002 , rev . phys . , 74 , 775 zang t. a. , 1976 , phd thesis , mit , cambridge ma for @xmath127 , we first examine the solution criterion of quadratic equation ( [ aligned ] ) for the aligned case . the determinant @xmath184 of quadratic equation ( [ aligned ] ) is @xmath361 where the three coefficients @xmath362 ( @xmath363 ) are defined explicitly by @xmath364 ^ 2\\ & \qquad\qquad\qquad\qquad\qquad + \frac{(1 - 4\beta)}{(1 + 2\beta ) } \bigg[\frac{-2m^2+(1 - 4\beta)(1 - 2\beta)}{(1 + 2\beta)}{f } { \cal c}{\cal p}_m+1 - 4\beta\bigg]{\cal h}_m\ , \\ c_1\equiv\ & 2(1-{f}{\cal c}{\cal p}_m)\bigg\{{\cal a}_m \bigg[\frac{(3 - 4\beta)m^2-(1 - 4\beta)(1 - 6\beta+4\beta^2)}{2(1 + 2\beta ) } { f}{\cal c}{\cal p}_m+m^2-(1 - 4\beta)(2-\beta)\bigg]\\ & \qquad\qquad\qquad\qquad\qquad + \frac{-2m^2+(1 - 4\beta)(1 - 2\beta)}{(1 + 2\beta)}{\cal h}_m\bigg\}\ , \\ c_0\equiv\ & { \cal a}_m^2(1-{f}{\cal c}{\cal p}_m)^2\ , \end{split}\ ] ] respectively . for the quadratic expression ( [ a01 ] ) in terms of @xmath81 , we further introduce another determinant @xmath365 such that @xmath366 [ ( 1 + 8\beta)m^2-(1 - 4\beta)(1 + 4\beta-8\beta^2)]\ .\ ] ] the determinant @xmath184 of quadratic equation ( [ aligned ] ) in terms of @xmath87 can be either positive or negative in general for different parameter combinations of @xmath8 , @xmath109 , @xmath77 and @xmath283 . it is fairly straightforward to obtain @xmath87 solutions by assigning values of all relevant parameters . in particular , for one important case of @xmath19 ( i.e. , the msid case with a flat rotation curve and an isothermal equation of state ) , the condition that @xmath172 for @xmath127 and @xmath367 when @xmath188 holds for all values of @xmath81 in a full msid with @xmath344 and the condition that @xmath352 for @xmath127 holds in a partial msid with @xmath343 ( lou 2002 ) . for the magnetic force - free case of @xmath11 with a declining rotation curve with increasing @xmath14 and a polytropic index @xmath182 , we have @xmath352 for @xmath127 and an arbitrary @xmath368 in both full and partial magnetized discs . we now proceed to prove this statement rigorously . for @xmath11 of the aligned case , we explicitly have @xmath369 ^ 2\ , \qquad c_1\equiv 2(1-{f}{\cal c}{\cal p}_m)m^2 [ { \cal a}_m(2{f}{\cal c}{\cal p}_m/3 + 1)-4{\cal h}_m/3]\ , \qquad c_0\equiv { \cal a}_m^2(1-{f}{\cal c}{\cal p}_m)^2\ , \ ] ] where @xmath209 and @xmath370 for @xmath11 and @xmath365 introduced above simply reduces to @xmath371 it is apparent that for @xmath372 , we have @xmath373 . as @xmath374 , it follows immediately that @xmath352 in definition ( [ a01 ] ) . on the other hand for @xmath375 , we have @xmath376 and therefore @xmath352 by definition ( [ a01 ] ) . for @xmath11 in the logarithmic spiral case , we again have the determinant @xmath352 for quadratic equation ( [ spiral ] ) , where in parallel with definitions ( [ a01 ] ) and ( [ a02 ] ) of the aligned case , the relevant parameters in the spiral case are @xmath377 ^ 2\ , \ c_1=2(1-{f}{\cal c } { \cal n}_m)(m^2+\xi^2)\bigg[{\cal a}_m\bigg(\frac{2}{3}{f } { \cal c}{\cal n}_m+1\bigg)-\frac{4}{3}{\cal h}_m\bigg]\ , \ c_0={\cal a}_m^2(1-{f}{\cal c}{\cal n}_m)^2\ , \ ] ] with @xmath378 and @xmath370 for @xmath11 . and @xmath365 simply reduces to @xmath379 hence , as the aligned @xmath11 case , it is apparent that for @xmath372 , we have @xmath373 . as @xmath374 , it follows immediately that @xmath352 . on the other hand for @xmath380 , we have @xmath376 and therefore @xmath352 . in this appendix , we first provide a proof that for the aligned case of @xmath11 , the physical @xmath87 solutions of the plus @xmath357 and the minus @xmath381 determined from quadratic equation ( [ q2aligned ] ) correspond to the factor @xmath382 being positive and negative , respectively . according to appendix a , the determinant @xmath184 for both aligned and spiral cases remains always positive for the case of @xmath11 . we start from equation ( [ q2aligned ] ) by casting it in the following compact form of @xmath383 where @xmath166 and in reference to equation ( [ q2aligned ] ) , @xmath384 , @xmath385 and @xmath386 are constants defined explicitly by @xmath387 } \ .\ ] ] it is apparent that @xmath385 is always positive . while @xmath384 and @xmath386 remain always positive for @xmath186 , their signs depend on the value of @xmath283 for the special case of @xmath188 . in fact , both @xmath384 and @xmath386 diverge at a critical value of @xmath283 for the @xmath188 case as already discussed in subsection 3.1.3 . let us first analyze the simpler cases of @xmath186 . as @xmath388 and @xmath389 , we have two @xmath87 solutions @xmath390^{1/2}}{2 } & > \frac{\wp_2+\wp_3+|\wp_2-\wp_3|}{2}\ge \wp_2\ ; \\ y_2=\frac{\wp_1+\wp_3-[(\wp_1+\wp_3)^2 - 4\wp_2\wp_3]^{1/2}}{2 } & = \frac{\wp_1+\wp_3-[(\wp_1+\wp_3 - 2\wp_2)^2 + 4\wp_2(\wp_1-\wp_2)]^{1/2}}{2}\\ & < \frac{\wp_1+\wp_3-|\wp_1+\wp_3 - 2\wp_2|}{2}\le \wp_2\ . \end{split}\ ] ] we next examine the @xmath188 case . the situation becomes somewhat involved since @xmath391 , @xmath392 when @xmath291 where the critical @xmath283 value @xmath393\simeq 0.6842 $ ] is determined from the condition @xmath292 ( see subsection 3.1.3 ) . on the other hand , we have @xmath394 when @xmath395 in general . for the aligned case and @xmath178 are consistent with the general solution form ( [ solalign ] ) for the aligned case . ] and for @xmath291 , we therefore have two @xmath87 solutions satisfying inequalities @xmath396^{1/2}}{2}<0\ ; \\ 0<y_2&=\frac{\wp_1+\wp_3+[(\wp_1+\wp_3)^2 - 4\wp_2\wp_3]^{1/2}}{2 } \\ & = \frac{\wp_1+\wp_3+[(\wp_1+\wp_3 - 2\wp_2)^2 + 4\wp_2(\wp_1-\wp_2)]^{1/2}}{2}\\ & < \frac{\wp_1+\wp_3+|\wp_1+\wp_3 - 2\wp_2|}{2}=\wp_2\ , \end{split}\ ] ] and for @xmath395 the situation remains the same as ( [ b2 ] ) . for the spiral case of @xmath11 in parallel , we can repeat this same procedure of analysis . as before , we cast quadratic equation ( [ beta0d25s ] ) in the form of @xmath397 } \\ & \qquad\qquad\qquad\qquad\qquad\qquad\qquad\qquad = \frac{d^2\ { d^2-(1-{f}{\cal c}{\cal n}_m)/ [ { f}{\cal c}{\cal n}_m+3(m^2 - 3/2)/(2m^2 + 2\xi^2 + 3/2)]\ } } { d^2-(1-{f}{\cal c}{\cal n}_m)/ ( { f}{\cal c}{\cal n}_m+3/2)}\ , \end{split}\ ] ] which yields two distinct @xmath87 solutions according to the analysis in subsection 3.2.3 . the following treatment is identical to that for the aligned @xmath11 case described in the first part of this appendix . we here provide a proof to show that the angular momentum flux associated with stationary coplanar mhd perturbations due to the gravity torque @xmath327 , the advective transport @xmath325 and the magnetic torque @xmath329 remains constant at all radii for the entire range @xmath10 . we start from expression ( [ rmphi ] ) for @xmath398 to derive the following two expressions in a straightforward manner : @xmath399j_m(k^{\prime}r ) \int_0^{\infty}r_a^{\prime\ -1/2}\cos(m\theta-\xi\ln r_a^{\prime})j_m(k^{\prime}r_a^{\prime})dr_a^{\prime}\ . \end{split}\ ] ] then in integral ( [ gravangm ] ) , we first integrate the kernel over @xmath400 to yield @xmath401dkdk^{\prime}\ , \ ] ] where @xmath402dr_adr_a^{\prime}\ .\ ] ] we next integrate over the vertical coordinate @xmath403 to yield the final result @xmath404dkdk^{\prime}\ .\ ] ] by introducing the following integral transformation @xmath405 we reduce expression ( [ final ] ) to @xmath406dxdx^{\prime}\ , \ ] ] where @xmath407dr_bdr_b^{\prime}={\cal n}_mx^{-1/2}x^{\prime\ -1/2}\sin\bigg[\xi\ln\bigg(\frac{x}{x^{\prime}}\bigg)\bigg]\ , \ ] ] where @xmath297 is the kalnajs function ( kalnajs 1971 ) . with these technical preparations , we finally determine the angular momentum flux associated with coplanar stationary mhd perturbations due to the gravity torque as @xmath408 where @xmath409dxdx^{\prime}\ , \ ] ] which is clearly independent of radius @xmath14 . in the tight - winding regime or the wkbj limit ( i.e. , @xmath410 ) , we may directly apply the result of fan & lou ( 1999 , see their eqn . 4.13 ) by noting @xmath411 such that @xmath412 where @xmath413 has been defined by equation ( [ thetas ] ) ; here , @xmath414 is inward for leading ( @xmath223 ) spiral waves and outward for trailing ( @xmath224 ) spiral waves . in general , @xmath415 can be calculated via numerical integrations and the specific result is displayed in fig . c1 . for fairly open mhd density waves ( @xmath416 ) with @xmath417 , the angular momentum flux associated with the gravity torque is therefore inward for leading ( @xmath223 ) spiral waves and outward for trailing ( @xmath224 ) spiral waves ; with increasing @xmath418 , @xmath415 falls off rapidly and oscillates around zero with the envelope profile closely fitting the wkbj result better and better . versus @xmath219 , shown in heavy solid line . we plot @xmath419 in light dashed line for comparison . for increasing @xmath418 , @xmath415 oscillates around zero and its envelope profile fits the @xmath419 curve with an increasing accuracy . ] meanwhile , we compute the other two perturbation contributions to the angular momentum flux , namely , those from the advective transfer and the magnetic torque . for the part of advective transport , we start from equation ( [ adveangm ] ) and make use of phase relationship ( [ relsurjzspiral ] ) to explicitly derive @xmath420r^{-1/2}\cos(\xi\ln r - m\theta)\\ & \qquad\qquad\qquad\qquad\qquad\qquad\qquad\ -(\omega\sigma/\sigma_0)[(1/2+\beta)\im(\theta_s ) -\xi\re(\theta_s)]r^{-1/2}\sin(\xi\ln r - m\theta)\ . \end{split}\ ] ] it immediately follows that the advective torque is given by @xmath421=m\pi^2g\sigma^2\xi^a\ , \ ] ] where @xmath422 } { { f}{\cal c}[1+d^2-(1 - 4\beta)q^2/(2 + 4\beta)]}\ , \ ] ] that is independent of radius @xmath14 . by specifying parameters @xmath8 , @xmath109 , @xmath283 , @xmath77 and @xmath219 , one finds that for the @xmath177 branch @xmath423 for @xmath223 and @xmath424 for @xmath224 . therefore , the advective angular momentum flux is radially inward for leading stationary spiral density waves and radially outward for trailing stationary spiral density waves for the @xmath177 branch of @xmath87 solutions . for the @xmath178 branch in comparison , the situation switches and the advective angular momentum flux is outward for leading stationary spiral density waves and inward for trailing stationary spiral density waves . practically , to achieve the @xmath178 branch of stationary logarithmic spiral configurations requires a much lower @xmath87 which would be unusual for real magnetized disc systems . an example of illustration for @xmath190 , @xmath344 and @xmath11 is shown in panel ( a ) of fig . c2 . for the momentum flux associated with the magnetic torque , we start from equation ( [ magnangm ] ) and make use of phase relationship ( [ relsurjzspiral ] ) to first derive @xmath425 \ r^{-3/2}\sin(\xi\ln r - m\theta ) + [ mb_0\sigma\im(\theta_s)/\sigma_0]\ r^{-3/2 } \cos(\xi\ln r - m\theta)\ , \\ & \re(b_\theta)=[\xi b_0\sigma\re(\theta_s ) /\sigma_0]\ r^{-3/2}\sin(\xi\ln r - m\theta ) + [ \xi b_0\sigma\im(\theta_s)/\sigma_0 ] \ r^{-3/2}\cos(\xi\ln r - m\theta)\ . ] it immediately follows that the momentum flux @xmath329 associated with the magnetic torque is given by @xmath426 where @xmath427 } \ , \ ] ] which is again independent of radius @xmath14 ; here , @xmath329 is inward for leading spiral density waves ( i.e. , @xmath223 ) and outward for trailing spiral density waves ( i.e. , @xmath224 ) . an example of illustration for @xmath190 , @xmath344 and @xmath11 is shown in panel ( b ) of fig . c2 . by comparing the three contributions to the total angular momentum flux transport , we reach the following conclusions . first , the contribution from the advective transport part tends to be more dominant over those from the magnetic torque and the gravity torque . secondly , for the @xmath177 branch for stationary fast mhd density waves that has a counterpart in the unmagnetized case , leading spiral density waves ( @xmath223 ) transport angular momentum inward and trailing spiral density waves ( @xmath224 ) transport angular momentum outward . thirdly , for the @xmath178 branch for stationary slow mhd density waves caused by the very presence of the azimuthal magnetic field , leading spiral density waves ( @xmath223 ) carry angular momentum outward and trailing spiral density waves ( @xmath224 ) carry angular momentum inward . note that all these results are only valid for stationary logarithmic spiral mhd density waves . there might be some reasons to rule out the @xmath178 branch in galactic contexts . for example , the @xmath178 branch is of lower @xmath87 which can not be sustained in real galactic discs . furthermore , for rotation curve parameter @xmath360 , the @xmath178 branch may be ruled out by inequality ( [ restrt ] ) .

it has been pointed out in previous publications @xcite that various models in quantum field theory ( qft ) , including qed , can be reformulated , using mediating - field green functions , in such a way that exact few - particle eigenstates of the resulting partially truncated hamiltonian can be obtained . this approach was then applied to two - body eigenstates in the scalar yukawa ( wick - cutkosky ) theory @xcite . we implement such an approach to qed in this paper . the lagrangian of two fermion fields , @xmath0 and @xmath1 , interacting electromagnetically , is @xmath2 the corresponding euler - lagrange equations of motion are the coupled dirac - maxwell equations , @xmath3 @xmath4 and @xmath5 where @xmath6 the equations ( [ 1.2 ] ) - ( [ 1.4 ] ) can be decoupled in part by using the well - known @xcite formal solution of the maxwell equation ( [ 1.4 ] ) , namely @xmath7 where @xmath8 is a green function ( or photon propagator in qft terminology ) , defined by @xmath9 and @xmath10 is a solution of the homogeneous ( or `` free field '' ) equation ( [ 1.4 ] ) with @xmath11 . we recall , in passing , that equation ( [ 1.7 ] ) does not define the covariant green function @xmath8 uniquely . for one thing , one can always add a solution of the homogeneous equation ( _ eq . _ ( [ 1.7 ] ) with @xmath12 . this allows for a certain freedom in the choice of @xmath13 , as is discussed in standard texts ( e.g. refs . @xcite ) . in practice , the solution of eq . ( [ 1.7 ] ) , like that of eq . ( [ 1.4 ] ) , requires a choice of gauge . however , we do not need to specify one at this stage . substitution of the formal solution ( [ 1.6 ] ) into equations ( [ 1.2 ] ) and ( [ 1.3 ] ) yields the partially reduced " equations , @xmath14 and @xmath15 these are nonlinear coupled dirac equations for two different fermion fields . to our knowledge no exact ( analytic or numeric ) solutions of equations ( [ 1.8 ] ) and ( [ 1.9 ] ) for classical fields have been reported in the literature , even for the case of a single fermion field ( say @xmath16 ) , though approximate ( perturbative ) solutions have been discussed by various authors , particularly barut and his co - workers ( see refs . @xcite and citations therein ) . however , our interest here is in the quantized field theory . the partially reduced equations ( [ 1.8])-([1.9 ] ) are derivable from the stationary action principle @xmath17 = \delta \int { \cal l}_{r}\ ; d^4x = 0,\ ] ] with the lagrangian density @xmath18 provided that the green function is symmetric in the sense that @xmath19 one can proceed to do conventional covariant perturbation theory using the reformulated qed lagrangian ( [ 1.11 ] ) . the interaction part of ( [ 1.11 ] ) has a somewhat modified structure from that of the usual formulation of qed . thus , there are two interaction terms . the last term of ( [ 1.11 ] ) is a `` current - current '' interaction which contains the photon propagator sandwiched between the fermionic currents . as such , it corresponds to feynman diagrams without external photon lines . the terms containing @xmath20 corresponds to diagrams that can not be generated by the term containing @xmath21 , particularly diagrams involving external photon lines ( care would have to be taken not to double count physical effects ) . however , we shall not pursue perturbation theory in this work . rather , we shall consider an approach that allows one to write down some unorthodox but exact eigenstates of a truncated model , in which terms involving @xmath22 are ignored . the paper is organized as follows . in section 2 we quantize the system using the canonical equal time formalism in the schrdinger picture . in section 3 an unconventional empty " vacuum state is used to construct exact one- , two- , and three - fermion eigenstates of the hamiltonian , truncated to exclude states with free ( physical ) photons . in section 4 we show that the resulting two - fermion equation is the breit equation in the coulomb gauge , but that it is the eddington - gaunt equation in the lorentz gauge . in section 5 we demonstrate that the breit equation can be obtained in the lorentz gauge , provided that higher - order retardation effects are taken into account . the reduction of the breit equation to radial form is described briefly in section 6 . for states of zero total angular momentum ( @xmath23 ) , four coupled radial equations are shown to arise . the analytical structure of their solutions is studied in section 7 . perturbative @xmath24 corrections to the rydberg spectrum of @xmath23 states are obtained in section 8 . in the case of equal rest masses the @xmath25 state equations have no unusual singularities and can be solved numerically . some of these results are presented and discussed in section 9 . the remainder of the paper is devoted to the study of @xmath26 states . in the section 10 the set of eight coupled radial first - order differential equations is reduced to four first - order ones and then to two second - order schrdinger - like equations . they are solved perturbatively in section 11 and @xmath24 relativistic corrections to the non - relativistic mass spectrum are obtained . a summary and concluding remarks are given in section 12 . we consider this theory in the canonical , equal - time formalism . to this end we write down the hamiltonian density corresponding to the lagrangian ( [ 1.11 ] ) @xmath27 where we have not written out the hamiltonian density for the free @xmath28 field . equal - time quantization corresponds to the imposition of anticommutation rules for the fermion fields , namely @xmath29 and all others vanish . in addition , if @xmath30 , there are the usual commutation rules for the @xmath20 field , and commutation of the @xmath20 field operators with the @xmath31 and @xmath32 field operators . the hamiltonian ( [ 2.1 ] ) contains an interaction term that is nonlocal in time , which can complicate the transition to a quantized theory . we shall avoid this problem by working in the schrdinger picture with @xmath33 in the expressions for the field operators and currents , that is @xmath34 , @xmath35 , etc . in equation ( [ 2.1 ] ) . this corresponds to neglecting higher order retardation effects . thereupon we obtain the result @xmath36 where @xmath37 for example , in the lorentz gauge ( @xmath38 ) , we have @xmath39 thus , in the schrdinger picture , the third term of the hamiltonian density ( [ 2.1 ] ) takes on the form @xmath40 in the remainder of this paper , we shall consider the simplified model without the interaction terms in ( [ 2.1 ] ) that contains @xmath41 . such a model is suitable for describing few - fermion states interacting via virtual photon exchange , but without decay or annihilation involving free ( physical ) photons . in short , in all that follows we consider the field theory based on the hamiltonian density of ( [ 2.1 ] ) but with @xmath42 . an attractive feature of this model is that exact few - fermion eigenstates of the hamiltonian can be obtained . we consider now the model for which the hamiltonian , in the schrdinger picture with @xmath33 , is given by the expression @xmath43 where @xmath44 @xmath45 and @xmath46 , where @xmath47 is given in equation ( [ 2.6 ] ) . note , again , that the terms in @xmath48 have been suppressed , so that processes in which free ( physical ) photons are emitted or absorbed are not accommodated . the hamiltonian @xmath49 has the same structure as the coulomb - qed ( cqed ) hamiltonian , that is the hamiltonian of qed in the coulomb gauge , but with the `` transverse - photon '' part ( that contains @xmath50 ) turned off . indeed @xmath49 would beidentical to @xmath51 if the indeces @xmath52 and @xmath53 took on only the value @xmath54 in equation ( [ 2.6 ] ) ( as it is , @xmath55 in eq.([2.6 ] ) ) . it has been shown earlier @xcite that exact two - fermion eigenstates of @xmath51 can be written down if we use an unconventional ( or `` empty '' ) vacuum state , @xmath56 , defined by @xmath57 the same is true of the present more realistic model , as we point out below . the unconventional empty vacuum definition ( [ 3.4 ] ) means that @xmath58 is interpreted as a ( free ) dirac - particle annihilation operator , while @xmath59 is , correspondingly , a dirac - particle creation operator . by `` dirac - particle '' we mean one described by the full dirac spinor , including positive and negative frequency components . ( recall that in the conventional approach , _ i.e. _ using a dirac `` filled negative energy sea '' vacuum which is annihilated by the positive frequency component of @xmath31 , it is only the negative - frequency component of @xmath31 that is an antiparticle creation operator , and the positive - frequency component of @xmath60 that is the particle creation operator ) . with these conventions , we write the normal - ordered hamiltonian @xmath61,\end{aligned}\ ] ] where @xmath62 and @xmath63 , etc . the normal ordering is achieved by using the anticommutation rules ( [ 2.2 ] ) as usual ; but note that it is not identical to the conventional normal ordering because of the unconventional empty vacuum that is being used , and the unconventional definition of @xmath31 and @xmath32 as annihilation operators and of @xmath64 as creation operators . to underscore this unconventional procedure we use the notation @xmath65 rather than @xmath66 in equation ( [ 3.5 ] ) . we note that the state defined by @xmath67 is a @xmath68 c - number coefficient vector , is an eigenstate of @xmath65 ( eq . ( [ 3.5 ] ) ) provided that @xmath69 satisfies the equation @xmath70 which is the usual time - independent one - particle dirac equation ( with positive and negative energy solutions ) , so that @xmath69 is a dirac spinor . therefore , we refer to @xmath71 as a one - dirac - fermion state . similarly , the two - dirac - fermion state , @xmath72 ( summation on @xmath73 implied ) is an eigenstate of @xmath65 ( equation ( [ 3.5 ] ) ) provided that the @xmath74 eigenmatrix @xmath75 satisfies the equation @xmath76^t + q_1 q_2 g_{\mu \nu}({\bx } - { { \mbox{\boldmath$y$ } } } ) { \tilde \gamma}^\mu f({\bx},{{\mbox{\boldmath$y$ } } } ) ( { \tilde \gamma}^\nu)^t = e\ , f({\bx},{{\mbox{\boldmath$y$}}}),\ ] ] where @xmath77 , @xmath78 , and the superscript @xmath79 indicates the transpose of the matrix in question . the detailed form of the interaction matrix @xmath80 depends on the choice of gauge . this equation ( [ 3.9 ] ) is a two - fermion dirac - like , or breit - like , equation with positive and negative energy solutions , and is , in this respect , different from those obtained in the conventional approach @xcite-@xcite in which the negative - energy solutions do not arise . we note , in passing , that if the interaction is turned off in eq . ( [ 3.9 ] ) ( _ i.e. _ @xmath81 ) , then the solution can be written as @xmath82 where @xmath83 and @xmath84 are solutions of the one - body dirac eigenvalue equation ( [ 3.7 ] ) . this indicates that in @xmath85 $ ] , the index @xmath86 corresponds to particle 1 ( with coordinates @xmath87 ) while @xmath88 corresponds to particle 2 ( with coordinates @xmath89 ) . in the rest frame of the two - fermion system ( _ i.e. _ when @xmath90 is taken to be an eigenstate of the momentum operator for this qft , with eigenvalue 0 ) , equation ( [ 3.9 ] ) reduces to an analogous equation in the single relative co - ordinate @xmath91 : @xmath92^t + q_1 q_2 g_{\mu \nu}({\br } ) { \tilde \gamma}^\mu f({\br } ) ( { \tilde \gamma}^\nu)^t = e \ , f({\br}).\ ] ] it can , therefore , be reduced to a set of ordinary , coupled , first - order differential equations for states of given @xmath93 . such equations can , at the very least , be solved numerically . this is a straightforward , though somewhat tedious , problem @xcite which we address below . the structure of the hamiltonian @xmath65 ( eq . ( [ 3.5 ] ) ) is such that generalizations to systems of more than two fermions are readily obtained . thus , the three fermion state , corresponding to a system like @xmath94 , defined by @xmath95 is an exact eigenstate of @xmath65 with eigenvalue @xmath96 , provided that the @xmath97 coefficient functions @xmath98 satisfy the three - body dirac - like equation , @xmath99_{\alpha_1 \alpha}\ ; f_{\alpha \alpha_2 \alpha_3}({\bx}_1,{\bx}_2,{\bx}_3 ) + \left[h_{m_1}({\bx}_2)\right]_{\alpha_2 \alpha}\;f_{\alpha_1 \alpha \alpha_3}({\bx}_1,{\bx}_2,{\bx}_3 ) \nn \\ { } + \left[h_{m_2}({\bx}_3 ) \right]_{\alpha_3 \alpha}\ ; f_{\alpha_1 \alpha_2 \alpha}({\bx}_1,{\bx}_2,{\bx}_3 ) + q_1 q_2 g_{\mu \nu } ( { \bx}_3 - { \bx}_1)\ ; ( { \tilde \gamma}^\mu)_{\alpha_1 \alpha}\;({\tilde \gamma}^\nu)_{\alpha_3 \beta}\ ; f_{\alpha \alpha_2 \beta}({\bx}_1,{\bx}_2,{\bx}_3 ) \nn \\ \label{3.12 } { } + q_1 q_2 g_{\mu \nu } ( { \bx}_3 - { \bx}_2 ) \ ; ( { \tilde \gamma}^\mu)_{\alpha_2 \alpha } \ ; ( { \tilde \gamma}^\nu)_{\alpha_3 \beta } \ ; f_{\alpha_1 \alpha \beta}({\bx}_1,{\bx}_2,{\bx}_3 ) \\ { } + q_1 ^ 2 g_{\mu \nu } ( { \bx}_1 - { \bx}_2 ) \ ; ( { \tilde \gamma}^\mu)_{\alpha_1 \alpha } \ ; ( { \tilde \gamma}^\nu)_{\alpha_2 \beta } \ ; f_{\alpha \beta \alpha_3 } ( { \bx}_1,{\bx}_2,{\bx}_3 ) \nn \\ { } = e \ ; f_{\alpha_1 \alpha_2 \alpha_3}({\bx}_1,{\bx}_2,{\bx}_3 ) , \nn\end{aligned}\ ] ] where summation on repeated spinor indices is implied . in the rest frame of the three - body system equation ( [ 3.12 ] ) reduces to one in two independent vectors only . nevertheless , the reduction of the equation for states of given @xmath93 is more formidable than in the two - body case . even then one is left with the full complexity of a relativistic three - body system . we shall not consider solutions of the three - body equation ( [ 3.12 ] ) in this paper . at this stage we must specify the choice of green s function , that is a choice of gauge . we could use any gauge , in principle , but we shall use the coulomb gauge , as this avoids nonphysical degrees of freedom and the need to take account of auxiliary conditions ( such as @xmath100 in the lorentz gauge ) . we use the relation @xmath101 and note that @xmath102 in the coulomb gauge @xcite . therefore , if we use eq . ( [ 2.3 ] ) and the identity ( @xcite , eq . 39.8 ) @xmath103 we obtain the coordinate - space representation of the coulomb - gauge green function @xmath104 where @xmath105 and @xmath106 . consequently , in the coulomb gauge , equation ( [ 3.10 ] ) becomes @xmath107^t + v(r ) \left [ f(\br ) - { 1 \over 2 } \left ( \arrowal f(\br ) \ , \cdot \arrowal^t + { 1 \over r^2 } \br \ , \cdot \arrowal f(\br ) \ , \br \ , \cdot \arrowal^t \right ) \right ] = e\ , f(\br ) , \ ] ] where @xmath108 . keeping in mind the notation , as explained below eq . ( [ 3.9a ] ) , we see that equation ( [ 4.5 ] ) is nothing other than the breit equation @xcite , written in the rest frame of the two - fermion system . if we had used the lorentz - gauge form of the green function ( [ 2.5 ] ) , equation ( [ 3.10 ] ) would take on the form @xmath109^t + v(r ) \left [ f(\br ) - \arrowal f(\br ) \ , \cdot \arrowal^t \right ] = e\ , f(\br ) , \ ] ] which is recognized to be the eddington - gaunt equation @xcite . the gaunt equation , unlike the breit equation , does not contain even lowest - order retardation effects ( see , also , section 5 , following ) . therefore , it will yield energy eigenvalues that will differ from those of the breit equation already at @xmath24 . note , also , that equations ( [ 4.5 ] ) and ( [ 4.6 ] ) , with the terms involving @xmath110 left out , become identical to the coulomb - qed model discussed earlier @xcite . although we shall use the coulomb gauge in this paper , it is instructive to see how , in the lorentz gauge , one needs to keep retardation effects at least to lowest non - vanishing order in order to achieve the same results . the green function of the dalembert equation in the lorentz gauge has the form : @xmath111 the reduced lagrangian @xmath112 in this case reads ( with @xmath113 ) : @xmath114 where the potential of electromagnetic interparticle interaction is @xmath115 j^\mu(\tp,\bxp ) .\ ] ] thus , the lagrangian ( [ 2 ] ) is nonlocal in time . because of this the standard hamiltonization procedure is not applicable . in order to employ the canonical hamiltonian formalism it is necessary to convert this lagrangian to single - time form . we shall do so by employing a procedure which takes into account the retardation effects approximately . using the substitution @xmath116 in ( [ 3 ] ) and expanding the current @xmath88 in a taylor series in @xmath117 , we obtain the result @xmath118 which reduces the potential ( [ 3 ] ) to the form : @xmath119 where @xmath120 . the functions @xmath121 and @xmath122 satisfy the relation @xmath123 the terms of odd power in @xmath117 in the expansion ( [ 4 ] ) vanish , because @xmath124 is an even function of @xmath117 ( _ c.f . _ ( [ 6 ] ) , ( [ 7 ] ) ) . as a result , the interaction lagrangian @xmath125 ( up to surface terms ) takes the following single - time form : @xmath126 where @xmath127 ( hereafter we omit the common time argument @xmath128 ) . the quantized theory based on the lagrangian @xmath129 ( the coulomb qed mentioned in section 3 ) was discussed in @xcite . it takes into account the relativistic kinematics of the fermion fields exactly , but describes their electromagnetic interaction with the transverse - photon " part turned off . the terms @xmath130 and @xmath131 can be treated as first - order corrections to @xmath132 , thus providing the approximate single - time form @xmath133 for the reduced nonlocal lagrangian @xmath112 ( _ c.f . _ eq . ( [ 1.11 ] ) ) . other terms following from the expansion are corrections of higher order . they will not be considered in the present paper . the lagrangian @xmath134 leads to euler - lagrange equations which are second - order in time derivatives , because of the term @xmath131 . thus , it describes the system with twice as many degrees of freedom as @xmath132 does , because @xmath60 are no longer the conjugate momenta of @xmath31 . this changes completely the dynamical content of the fields @xmath31 and @xmath32 . since the second - order time derivatives occur in small terms only , they should be eliminated by means of the euler - lagrange equations of a lower - order approximation . but the resulting field equations are then not necessarily the euler - lagrange equations of a known lagrangian . thus the transition to the hamiltonian and to a canonical quantum description becomes unclear . to avoid this difficulty it is tempting to eliminate the time derivative of the charge density @xmath135 directly in @xmath131 by taking into account the conservation law @xmath136 this conservation law is a consequence of the euler - lagrange equations , which follow from the reduced lagrangian @xmath112 as well as from the truncated one @xmath132 . however , the direct use of the equations of motion ( or their consequences ) in the lagrangian is not a correct procedure : it changes the equations of motion themselves . this fact was first emphasized in the case of the golubenkov - smorodinskii lagrangian @xcite and then subsequently discussed in the literature @xcite-@xcite . instead , one can use the method of double zero " , used in refs . @xcite . in our case this consists of the following modification of the lagrangian : @xmath137 where @xmath138 it is easy to see that the term @xmath139 possesses the property : @xmath140 so that it does not change the variational problem to the accuracy required . on the other hand , it cancels those terms of @xmath134 which are quadratic in time derivatives of the fields . thus the modified lagrangian @xmath141 yields equations of motion , which are first order in the time derivatives of the particle fields @xmath142 . next , we perform the following transformation of the field variables : @xmath143 where @xmath144 the transformation ( [ 17 ] ) can be regarded as an approximate @xmath145 gauge transformation , which however is not canonical due to the dependence of @xmath146 on the fields . this transformation removes time derivatives from the interaction part of @xmath147 . to the accuracy required , and up to surface terms , lagrangian @xmath141 can be written in the form @xmath148 where the notations @xmath149 , @xmath150 , _ etc . _ mean that the fields @xmath31 , @xmath32 are replaced by @xmath151 , @xmath152 . integrating the last term of @xmath141 by parts , omitting surface terms , and using equations ( [ 6 ] ) and ( [ 8 ] ) , we reduce the lagrangian @xmath141 to the form @xmath153 where @xmath154 this formulation allows us to treat the variables @xmath155 and @xmath156as the canonical conjugates of @xmath151 and @xmath152 , respectively . that is , we impose the anticommutation relations ( [ 2.2 ] ) on the underscored fields , and not on the original ones , when performing the quantization . thereafter , since the hamiltonian @xmath157 is formally identical with the coulomb gauge hamiltonian ( _ c.f . _ ( [ 2.1 ] ) with ( [ 4.4 ] ) ) , calculations like those of section 4 lead to the breit equation ( [ 4.5 ] ) . although two fermion equations have been around since the 1920s , their full reduction to radial form is of more recent vintage ( see , for example @xcite , @xcite ) . the reduction of equation ( [ 4.5 ] ) to radial form is essentially the same as presented in ref . @xcite , hence all the details shall not be repeated here . as shown in @xcite , we note that eq.([4.5 ] ) has the schrdinger equation as a non - relativistic limit , and the dirac equation as a one - body limit if one of @xmath158 . for the two - fermion state @xmath159 , eq . ( [ 3.8 ] ) , to be simultaneously an eigenstate of @xmath160 , @xmath161 , and parity , the bispinor " @xmath162 $ ] must be of the form @xmath163,\ ] ] for @xmath164 parity eigenstates , and @xmath165,\ ] ] for @xmath166 parity eigenstates . the @xmath167 angular `` bispinor harmonics '' @xmath168 , @xmath169 , @xmath170 , @xmath171 , for given total angular momentum quantum numbers @xmath172 and @xmath173 , are @xmath174,\ ] ] @xmath175,\ ] ] @xmath176,\ ] ] and @xmath177 . \nonumber\end{aligned}\ ] ] we note that @xmath178 is antisymmetric and @xmath179 are symmetric matrices . furthermore @xmath180 , @xmath181 and @xmath182 correspond to opposite parity because @xmath183 and @xmath184 , @xmath178 have @xmath185 whereas @xmath182 have @xmath186 . these four bispinor harmonics form an orthonormal set , in the sense that @xmath187 , where @xmath188 and the integrations are taken over the entire solid angle . the eight radial functions in the bispinors ( [ 6 - 1 ] ) and ( [ 6 - 2 ] ) are solutions of the coupled radial equations that are obtained by substituting ( [ 6 - 1 ] ) , ( [ 6 - 2 ] ) into equation ( [ 4.5 ] ) and equating the coefficients of the four independent bispinor harmonics . we make use of the following identities in carrying out the radial reduction : @xmath189 where @xmath190 is any @xmath167 bispinor harmonic , @xmath191 a radial function , and @xmath192 , and @xmath193 . in addition , we note the following useful properties of the above bispinors harmonics : @xmath194 where @xmath195 it is evident from equations ( [ 6 - 1 ] ) and ( [ 6 - 2 ] ) that , in general , eight coupled radial equations are obtained , for arbitrary @xmath26 . for the @xmath196 states , namely the @xmath197 and @xmath198 states , only two linearly independent bispinor harmonics arise , namely @xmath178 and @xmath199 ( equations ( [ 6 - 3 ] ) and ( [ 6 - 6 ] ) ) , and so @xmath200 in equations ( [ 6 - 1 ] ) and ( [ 6 - 2 ] ) . ( here , as elsewhere , we give in brackets the nonrelativistic limit designation , @xmath201 , corresponding to the @xmath93 state in question . ) thus there is only one set of four coupled radial equations for each of @xmath202 and @xmath203 states : @xmath204 @xmath205 @xmath206 @xmath207 where @xmath208 , @xmath209 , the potential @xmath210 ( @xmath211 ) , and @xmath212 is the eigenenergy ( two - particle bound - state mass ) to be determined , while @xmath213 ( @xmath214 ) for the @xmath197 states and @xmath215 ( @xmath216 ) for the @xmath198 states . as shown in ref . @xcite , eqs . ( [ 7 - 1 ] ) - ( [ 7 - 4 ] ) have the expected schrdinger nonrelativistic limit , and the dirac one - body limit . we note that the case with @xmath217 corresponds to the simplified model without transverse - photon interactions , that is the ` coulomb qed ' model of ref . similarly for @xmath217 , if the potential is @xmath218 , and the sign of the potential is reversed in ( [ 7 - 2 ] ) and ( [ 7 - 3 ] ) we recover the @xmath219 radial equation of the yukawa model discussed in ref . @xcite , for which the inter fermion interaction is via a ( massive or massless ) scalar mediating field . we should point out that equations ( [ 7 - 1 ] ) - ( [ 7 - 4 ] ) , like the dirac equations , have both positive and negative energy solutions . indeed , in this two - body case , there are solutions of four types : @xmath220 , @xmath221 , @xmath222 and @xmath223 , as can be seen most easily from the @xmath224 case . of these , two are positive energy and two are negative energy solutions . since we do not have analytic solutions for the eigenenergies of the present qed case , it is useful to illustrate this phenomenon on the scalar yukawa ( or wick - cutkosky " ) model , in which scalar particles interact via a massive or massless mediating field . for such a scalar model , analytic expressions for the two - body bound state energy eigenvalues are available in the massless - exchange case @xcite : @xmath225 where @xmath226 is the effective dimensionless coupling constant , analogous to the fine structure constant of qed , and @xmath227 is the principal quantum number . the @xmath228 in eq . ( [ 7 - 4a ] ) correspond to two segments of a distorted semicircle . the upper branch of this distorted semicircle corresponds to the upper ( positive ) sign in ( [ 7 - 4a ] ) . it begins from @xmath229 at @xmath224 ( indeed , @xmath230 for @xmath231 ) , and decreases to a @xmath232 at the critical value of @xmath233 , beyond which @xmath96 ceases to be real , and the wave functions cease to be normalizable . the lower branch , by contrast , begins from @xmath234 at @xmath235 and rises monotonically to meet the upper branch at the same critical point @xmath236 . these @xmath237 type bound state eigenenergies do not have the correct balmer limit , since for this branch @xmath238 for @xmath239 , but @xmath240 for @xmath241 . thus , this mixed energy " @xmath242 bound - state spectrum must be regarded as unphysical . there are also negative - energy solutions of the @xmath243 and @xmath244 type , but they are not bound - states , since the potential effectively reverses sign for the negative - energy solutions ( as happens also in the dirac - coulomb case ) . the same type of behaviour of the energy spectrum is observed in another analytically solvable case , namely a fermion and a scalar particle interacting via massless scalar quantum exchange @xcite thus , we expect that the energy eigenvalue spectrum of ( [ 7 - 1 ] ) - ( [ 7 - 4 ] ) will be qualitatively similar to that of the scalar exchange models just described . we have not been able to determine solutions to the coupled radial equations ( [ 7 - 1])-([7 - 4 ] ) in terms of common analytic functions . it is of interest , therefore , to consider the properties and general behaviour of the solutions before commencing with numerical solutions . in analogy with the scalar model just described , and with the coulomb qed case @xcite we expect that , as @xmath226 increases , the eigenenergy spectrum @xmath245 of equations ( [ 7 - 1])-([7 - 4 ] ) will have a qualitative behaviour similar to that of the dirac spectrum , namely that @xmath245 decreases monotonically from @xmath246 until @xmath226 hits a critical value @xmath247 , beyond which @xmath245 ceases to be real . it is possible to infer the value of @xmath247 by considering the ultra - relativistic limit , @xmath248 , in which case we can neglect the masses @xmath249 and @xmath250 , and seek solutions of ( [ 7 - 1])-([7 - 4 ] ) with @xmath251 . ( this approach , when applied to the one - body dirac - coulomb case , yields the correct critical values @xmath252 . ) in this ultra - relativistic approximation , equations ( [ 7 - 1])-([7 - 4 ] ) have the solutions @xmath253 , @xmath254 , @xmath255 ( i.e. @xmath256 ) with @xmath257 , which gives @xmath258 for all @xmath219 states . note however , that this result does not mean that the value of the two - fermion rest mass @xmath96 at @xmath247 is necessarily the same for the @xmath259 and @xmath260 states ( certainly , such is not the case in the one - body limit ) . note , also that the result , @xmath261 for @xmath262 states is independent of the masses , that is we expect it to be the same for all finite @xmath263 . the value @xmath261 is different , and somewhat larger , than the known one - body limit ( dirac - coulomb ) value of @xmath264 for @xmath265 states . also , this value is much smaller that the value @xmath266 , which is obtained for the two - fermion coulomb qed case ( where @xmath217 ) for @xmath262 states . for the coulomb potential @xmath267 , where @xmath268 , it is often convenient to rescale the radial variable , that is to let @xmath269 , where @xmath270 is a suitable scale parameter . for example , the radial functions @xmath271 , @xmath128 , @xmath272 , @xmath273 have the large @xmath274 ( negligible @xmath275 and @xmath276 ) behaviour @xmath277 , etc . , for positive energy @xmath23 bound states , where @xmath270 is given by @xmath278[e^2 - m_-^2]}\over { 4e^2 } } \qquad { \rm or } \qquad { 1\over { a^2 } } = m^2 - \left({e\over 2}\right)^2 \quad { \rm if } \quad m_1=m_2 .\ ] ] eq . ( [ 7 - 5 ] ) implies that @xmath270 is positive only for @xmath279 , which means that the bound state spectrum must lie in this domain ( _ c.f . _ ( [ 7 - 4a ] ) ) . from this , and in analogy with the scalar model results , we can infer that the critical value @xmath280 lies between @xmath281 and @xmath282 , and likely closer to the former rather than the latter . with the rescaling @xmath269 , equations ( [ 7 - 1])-([7 - 4 ] ) become modified slightly , in that @xmath274 is replaced by @xmath283 in all of them . for purposes of numerical integration of the radial equations the scale parameter @xmath270 can be chosen to be anything that is convenient , be it that given in eq . ( [ 7 - 5 ] ) , or @xmath284 or @xmath285 , or whatever . for a power series analysis of the radial equations it is useful to make the replacement @xmath286 , etc . assuming solutions of the form @xmath287,\ ] ] @xmath288,\ ] ] @xmath289,\ ] ] @xmath290,\ ] ] we find , upon substitution into the radial equations for @xmath291 , @xmath292 , @xmath293 , @xmath294 and equating coefficients of powers of @xmath295 , that the coefficients @xmath296 must satisfy the following recursion relations : @xmath297 @xmath298 @xmath299 @xmath300 where @xmath301 . if @xmath302 then ( [ 7 - 10])-([7 - 13 ] ) are the recursion relations for the power series representations of the functions @xmath303 , etc . , rather than for @xmath304 , etc . for @xmath305 , and bearing in mind that @xmath306 , ( [ 7 - 10])-([7 - 13 ] ) yield four coupled homogeneous equations for the parameters @xmath307 , @xmath308 , @xmath309 , @xmath310 , which have non - trivial ( and non - singular ) solutions only if @xmath311 for the @xmath23 states , for any values of @xmath249 , @xmath250 , whereupon @xmath312 the condition ( [ 7 - 14 ] ) implies that the radial equations have real bound state solutions of the form ( [ 7 - 6])-([7 - 9 ] ) only for @xmath313 , for any values of @xmath249 and @xmath250 . this , in turn , implies that @xmath314 for the @xmath315 states for any ( finite ) values of @xmath249 and @xmath250 , in agreement with the ultra relativistic limit discussed above . this condition for bound states , @xmath316 , is additional to the one that follows from eq . ( [ 7 - 5 ] ) , namely that @xmath317 . the recursion relations ( [ 7 - 10])-([7 - 13 ] ) , with ( [ 7 - 14 ] ) and ( [ 7 - 15 ] ) , can be used to generate the power series form of the solutions of eqs . ( [ 7 - 6])-([7 - 9 ] ) . these series converge in the domain @xmath318 , as discussed below and in @xcite . such a series can be used , for example , as a starting procedure for the numerical integration of the radial equations ( [ 7 - 1])-([7 - 4 ] ) . unlike in the dirac case , the recursion relations ( [ 7 - 10])-([7 - 13 ] ) do not admit power series solutions of the form ( [ 7 - 6])-([7 - 9 ] ) , which terminate at the same power , say @xmath319 , so that @xmath320 . in particular , the ground state solution is not of the simple form @xmath321 as it is for the two radial dirac equations . this is perhaps to be expected , since in the dirac case there are only two functions , say @xmath291 and @xmath292 , and four unknowns to be determined , namely @xmath322 , @xmath323 , @xmath270 and @xmath96 . since the two coupled radial dirac equations yield four equations ( the coefficients of @xmath324 and of @xmath325 ) , it is not surprising that a solution is obtained . in the present case , we have four coupled radial equations ( [ 7 - 1])-([7 - 4 ] ) , which yield eight equations ( the coefficients of @xmath324 and of @xmath325 ) to be satisfied by the six unknowns of the proposed solutions ( [ 7 - 16 ] ) , namely @xmath308 , @xmath309 , @xmath310 , @xmath323 , @xmath270 and @xmath96 . thus the system is overdetermined and no solution of the form ( [ 7 - 16 ] ) is possible . this situation persists for any solution of the form ( [ 7 - 6])-([7 - 9 ] ) where the polynomials all terminate at the same degree . therefore , we shall solve the radial equations ( [ 7 - 1])-([7 - 4 ] ) numerically . equations ( [ 7 - 1 ] ) - ( [ 7 - 4 ] ) are not independent . indeed , elementary manipulations of these equations , namely subtracting eq . ( [ 7 - 4 ] ) from eq . ( [ 7 - 1 ] ) and similarly eq . ( [ 7 - 3 ] ) from eq . ( [ 7 - 2 ] ) show that [ 7 - 6a ] v(r ) = e - m_+ - ( 1- ) v(r ) s(r ) , u(r ) = e - m_- - ( 1-)v(r ) t(r ) . thus , the number of equations can be reduced from four to two . we introduce the auxiliary functions @xmath326 and @xmath327 . then , adding eqs . ( [ 7 - 1 ] ) to eq . ( [ 7 - 4 ] ) and eq . ( [ 7 - 2 ] ) to eq . ( [ 7 - 3 ] ) , and using eq . ( [ 7 - 6a ] ) yields the equations [ 7 - 7a ] f^ ( r ) = k r f(r ) + w_g ( r ) g(r ) , -g^ ( r ) = k r g(r ) + w_f ( r ) f(r ) , where [ 7 - 8a ] w_g ( r ) = 1 2 , [ 7 - 9a ] w_f ( r ) = 1 2 , and where @xmath328 for the present qed case in the coulomb gauge , for the @xmath259 states ( for which @xmath214 ) , @xmath329 while @xmath330 and @xmath331 . in this case we see that @xmath332 is singular at @xmath333 , where @xmath334 is the reduced mass and @xmath335 is the reduced bohr radius . this singular point is quite close to the origin ( in units of the reduced bohr radius ) for small @xmath226 . the appearance of this singularity may signal difficulties in the numerical determination of eigensolutions of the equations ( [ 7 - 7a ] ) by standard shooting " methods . for the @xmath198 states ( for which @xmath216 ) , the singularity at @xmath336 occurs in @xmath337 , but only if @xmath338 . thus for the equal - mass @xmath260 states , equations ( [ 7 - 7a ] ) have only the usual @xmath339 singularities at the origin , and are amenable to solution by standard methods , as discussed below . equations ( [ 7 - 7a ] ) can be written in the matrix form [ 8 - 1 ] h |= | h = , = , and where @xmath340 . if @xmath341 is replaced by @xmath342 and @xmath337 by @xmath343 , where @xmath344 , then eq . ( [ 7 - 7a ] ) , or ( [ 8 - 1 ] ) , is equivalent to the radial schrdinger equation . the first - order correction to the non relativistic energy @xmath345 is then given by [ 8 - 2 ] = _ nr | h- h_nr| _ nr = f _ nr | - _ nr + w_f ^nr - w_f | f_nr + g _ nr | - _ nr + w_g ^nr - w_g | g_nr if we expand the coefficients @xmath346 and @xmath337 ( eqs . ( [ 7 - 8 ] ) and ( [ 7 - 9 ] ) ) in powers of @xmath347 , and keep only the lowest - order terms , we obtain [ 8 - 2a ] - _ nr + w_f ^nr - w_f ( _ nr - v_)^2 [ 8 - 2b ] - _ nr + w_g ^nr - w_g - ( 1 - 2 ) _ nr + 1 2 v_+ ^2 v _ , where @xmath348 . this leads to the following @xmath24 correction to the non - relativistic energy for the @xmath23 states : @xmath349 \nonumber \\ - \left ( 1 - 2 { \mu \over { m_+ } } \right ) \epsilon _ { \rm nr } \ , \langle \langle 1 \rangle \rangle + { 1 \over 2}\left ( ( 1 + \eta ) + ( 1 - \eta ) \delta^2 \right ) \langle \langle v \rangle \rangle . \label{8 - 3}\end{aligned}\ ] ] we use the notation @xmath350 but @xmath351 . for the @xmath352 states ( for which @xmath213 and @xmath353 ) this formula gives for the lowest order relativistic correction the result [ 8 - 4 ] ( 0 ^ - ( n ^1s_0 ) ) = ^4 \ { 1 n^4 ( 3 8 - 1 8 m_+ ) + 1 n^3 ( -1 2 + 2 m_+ ) } which becomes @xmath354 in the equal mass @xmath241 case . this does not agree with the known positronium value of @xmath355 @xcite . this is not surprising , since the breit equation , without modification , is known to give the incorrect fine structure for hydrogen and positronium . brown and ravenhall @xcite argue that the reason for this , is the mixing of positive and negative energy one - particle states ( which arises , in our formalism , because of our use of the `` empty '' vacuum ( [ 3.4 ] ) ) . this difficulty of the unmodified breit equation is discussed in various works ( _ e.g. _ refs . the modification that is needed to bring the result into agreement with the observed fine structure of h or ps is to subtract off the expectation value of the operator @xcite [ 8 - 7 ] h^= ^2 ( 3 - 2 _ 1 _ 2 + _ 2r ) _ ir = _ i / r , where , in this equation , we use the notation of @xcite . the expectation value of ( [ 8 - 7 ] ) ( with respect to the non - relativistic eigenfunctions ) is [ 8 - 8 ] h^_nr = ^4 1 -^2 f_s = ^4 4 f_s m_+ , where @xmath356 for the singlet @xmath357 states , while @xmath358 for the triplet @xmath359 states with @xmath360 , but @xmath361 for the triplet states with @xmath23 ( see , also , section 10 below ) . the expression ( [ 8 - 8 ] ) gives the value @xmath362 for the equal - mass ground state , which , when subtracted from the ` breit ' value of @xmath363 gives the expected positronium result @xmath364 . more generally for arbitrary masses , if we subtract ( [ 8 - 8 ] ) from the expression ( [ 8 - 3 ] ) , we obtain the corrected " result [ 8 - 9 ] _ c ( n 0 ^ - ( ^1s_0 ) ) = ^4 \ { 1 n^4 ( 3 8 - 1 8 m_+ ) + 1 n^3 ( -1 2 - 2 m_+ ) } . this same result ( [ 8 - 9 ] ) was obtained previously for the @xmath365 state by darewych and horbatsch @xciteb , who used a perturbative approximation on variationally derived equations . somewhat surprisingly , the unequal - mass @xmath24 corrections for arbitrary states seem to have been worked out fully only relatively recently . we refer to the work of connell @xcite , who used a quasipotential formalism based on the work of todorov @xcite , and of hersbach @xcite , who used a formalism based on a relativistic generalization of the lippmann - schwinger equation due to de groot and ruijgrok @xcite . our corrected expression ( [ 8 - 9 ] ) agrees with the results of these authors . ( the @xmath24 corrections for hydrogen and muonium quoted in standard references are expansions in @xmath263 ( _ e.g. _ @xcite , @xcite . ) ) for the @xmath366 states ( for which @xmath215 and @xmath367 ) equation ( [ 8 - 3 ] ) gives [ 8 - 5 ] ( n 0^+ ( ^3p_0 ) ) = ^4 \ { 1 n^4 ( 3 8 - 1 8 m_+ ) + 1 n^3 ( - 1 2 - 2 3 m_+ ) } , which _ does _ agree , in the equal mass case , with the ps values for all the @xmath368 states , as well as with the unequal - mass expressions of connell @xcite and hersbach @xcite for these states . this agreement implies that the correction " @xmath369 vanishes for the @xmath370 states , as indeed it does . we might note , in passing , that formula ( [ 8 - 3 ] ) gives the correct @xmath24 results for the coulomb qed ( @xmath371 ) case @xcite , for which the @xmath146 coefficients are non - singular for @xmath372 . in the case of equal masses @xmath373 the radial equations ( [ 7 - 7a ] ) for @xmath374 states are free of singularities . thus the boundary value problem is well posed , and it can be solved by means of a standard numerical ode - solving procedure . we solved it by the shooting " method using the maple runge - kutta programme . the corresponding perturbative spectrum ( _ c.f . _ ( [ 8 - 5 ] ) ) , @xmath375 agrees with the orthopositronium spectrum since the contribution of the extra terms ( _ c.f . _ ( [ 8 - 7 ] ) ) ( caused by positive - negative energy mixing ) vanishes in this case . in the table 1 the numeric and perturbative results are presented for the lowest - energy @xmath260 state ( _ i.e. _ , @xmath23 , @xmath376 , @xmath365 ) for different values of @xmath377 . . values of @xmath378 for the @xmath365 , @xmath203 state ( @xmath379 ) . [ cols= " < , < , < " , ] note that the critical value of the two - body mass , @xmath380 , increases with @xmath227 , in contrast to the scalar model , for which @xmath381 for all @xmath227 . figure 1 is a plot of the unnormalized reduced radial wave functions @xmath303 , @xmath382 and @xmath383 ( see eqns . ( [ 7 - 1])-([7 - 4 ] ) ) in the case of equal massless , @xmath379 , for the lowest - energy @xmath384 states , when @xmath385 . these wave functions are qualitatively similar to those obtained for these states in the coulomb qed case @xcite . the `` large '' component @xmath303 is nodeless while the `` small '' components @xmath382 and the `` doubly small '' one @xmath383 have one node . the node at the origin , @xmath386 , is a consequence of our use of reduced radial wave functions @xmath303 , etc . , rather than the actual @xmath387 , etc . indeed , the wave functions behave at small @xmath274 as follows : @xmath388 where @xmath389 ( see eq . ( [ 7 - 16 ] ) ) . thus the matrix wave function @xmath390 is singular , @xmath391 , as happens also in the one - body dirac equation with a coulomb potential . nevertheless , @xmath390 is normalizable for all @xmath226 up to and including @xmath392 , at which point @xmath393 , as can be seen in figures 2 and 3 . figure 3 represents the excited @xmath394 state for the critical coupling strength @xmath261 . in this case , the wavefunction @xmath303 has one node while @xmath395 and @xmath383 have two nodes . this behaviour differs from that found in cqed @xcite for the same case ( where the number of nodes was two , one and three , respectively ) . in the @xmath396 case ( figure 4 ) each of the functions @xmath303 , @xmath395 and @xmath383 gets one more node . this tendency likely continues for higher values of the quantum number @xmath227 . for states with @xmath397 , the eigenstate problem reduces to a set of eight first - order differential equations for the functions @xmath398 and the energy @xmath96 ( _ c.f . _ eqns . ( [ 6 - 1 ] ) and ( [ 6 - 2 ] ) ) . it is convenient to present this set in the following matrix form . let us introduce the 8-dimensional vector - function : @xmath399.\ ] ] then the set of radial equations reads : @xmath400 where the 8@xmath4018 matrix @xmath402 has the following structure : @xmath403/r.\ ] ] here @xmath404 is the unit matrix , @xmath405 is diagonal , @xmath406,\ ] ] @xmath208 , and the form of 8@xmath4018 matrices @xmath407 , @xmath408 , and @xmath409 depends on the parity @xmath410 : @xmath411,{\label{r7}}\\ { \mbox{\ssb g } } & = & \left[\mbox{\scriptsize$ \begin{array}{cccccccc } 0 & 0 & -(j{+}1)a & -jb & -(j{+}1)a & -jb & 0 & 0\\ 0 & 0 & ( j{+}1)b & -ja & -(j{+}1)b & ja & 0 & 0\\ -(j{+}1)a & ( j{+}1)b & 0 & 0 & 0 & 0 & -(j{+}1)a & -(j{+}1)b\\ -jb & -ja & 0 & 0 & 0 & 0 & -jb & ja\\ -(j{+}1)a & -(j{+}1)b & 0 & 0 & 0 & 0 & -(j{+}1)a & ( j{+}1)b\\ -jb & ja & 0 & 0 & 0 & 0 & -jb & -ja\\ 0 & 0 & -(j{+}1)a & -jb & -(j{+}1)a & -jb & 0 & 0\\ 0 & 0 & -(j{+}1)b & ja & ( j{+}1)b & -ja & 0 & 0 \end{array}$ } \right],{\label{r8}}\\ { \mbox{\ssb s } } & = & \left[\mbox{\scriptsize$ \begin{array}{cccccccc } 0 & 0 & 0 & 0 & 0 & 0 & 2 & 0 \\ 0 & 0 & 0 & 0 & 0 & 0 & 0 & -1 \\ 0 & 0 & 0 & 0 & -b^2&-ab & 0 & 0\\ 0 & 0 & 0 & 0 & -ab & -a^2 & 0 & 0\\ 0 & 0 & -b^2 & -ab & 0 & 0 & 0 & 0\\ 0 & 0 & -ab & -a^2 & 0 & 0 & 0 & 0\\ 2 & 0 & 0 & 0 & 0 & 0 & 0 & 0\\ 0 & -1 & 0 & 0 & 0 & 0 & 0 & 0 \end{array}$ } \right]{\label{r9}}\end{aligned}\ ] ] for @xmath412 , and @xmath413,{\label{r10}}\\ { \mbox{\ssb g } } & = & \left[\mbox{\scriptsize$ \begin{array}{cccccccc } 0 & 0 & ( j{+}1)a & -(j{+}1)b & ( j{+}1)a & ( j{+}1)b & 0 & 0\\ 0 & 0 & jb & ja & jb & -ja & 0 & 0\\ ( j{+}1)a & jb & 0 & 0 & 0 & 0 & ( j{+}1)a & jb\\ -(j{+}1)b & ja & 0 & 0 & 0 & 0 & ( j{+}1)b & -ja\\ ( j{+}1)a & jb & 0 & 0 & 0 & 0 & ( j{+}1)a & jb\\ ( j{+}1)b & -ja & 0 & 0 & 0 & 0 & -(j{+}1)b & ja\\ 0 & 0 & ( j{+}1)a & ( j{+}1)b & ( j{+}1)a & -(j{+}1)b & 0 & 0\\ 0 & 0 & jb & -ja & jb & ja & 0 & 0 \end{array}$ } \right],{\label{r11}}\\ { \mbox{\ssb s } } & = & \left[\mbox{\scriptsize$ \begin{array}{cccccccc } 0 & 0 & 0 & 0 & 0 & 0 & -b^2 & -ab\\ 0 & 0 & 0 & 0 & 0 & 0 & -ab & -a^2\\ 0 & 0 & 0 & 0 & 2 & 0 & 0 & 0\\ 0 & 0 & 0 & 0 & 0 & -1 & 0 & 0\\ 0 & 0 & 2 & 0 & 0 & 0 & 0 & 0\\ 0 & 0 & 0 & -1 & 0 & 0 & 0 & 0\\ -b^2 & -ab & 0 & 0 & 0 & 0 & 0 & 0\\ -ab & -a^2 & 0 & 0 & 0 & 0 & 0 & 0 \end{array}$ } \right]{\label{r12}}\end{aligned}\ ] ] for @xmath414 , where @xmath415 , @xmath416 and @xmath417 are defined in eq . ( [ 6.14 ] ) . due to the properties @xmath418 the `` radial '' hamiltonian @xmath419 is a hermitian operator with respect to the inner product @xmath420 where the subscript 8 " denotes the dimensions of the vector - functions @xmath421 . in the subsequent reduction of the set ( [ r4 ] ) one can use the fact that @xmath422 ( for either parity ) . thus one can reduce the number of differential equations from 8 to 4 . we perform this reduction in a way that ensures , as far as possible , the hamiltonian structure of the equations . first of all we perform the orthogonal transformation : @xmath423 where @xmath424,\quad p=(-)^{j\pm1},{\label{r17}}\\ { \mbox{\ssb e } } & = & \frac{1}{\sqrt{2}}\left[\mbox{\scriptsize$ \begin{array}{cccccccc } 0 & 0 & 1 & 0 & 1 & 0 & 0 & 0\\ -a & b & 0 & 0 & 0 & 0 & -a & b\\ 0 & 0 & 0 & -1 & 0 & 1 & 0 & 0\\ -b & -a & 0 & 0 & 0 & 0 & b & a\\ b & a & 0 & 0 & 0 & 0 & b & a\\ a & -b & 0 & 0 & 0 & 0 & -a & b\\ 0 & 0 & -1 & 0 & 1 & 0 & 0 & 0\\ 0 & 0 & 0 & 1 & 0 & 1 & 0 & 0 \end{array}$ } \right],\quad p=(-)^{j}{\label{r18}}\end{aligned}\ ] ] it preserves the inner product ( [ r15 ] ) and reduces the equations ( [ r4 ] ) to the form : @xmath425 it is convenient at this stage to express the 8-dimensional vectors and matrices in terms of 4-dimensional blocks : @xmath426 , \qquad \tilde{{\mbox{\ssb v}}}(r ) \equiv \tilde{{\mbox{\ssb u}}}(r ) - e{{\mbox{\ssb i}}}= \left [ \begin{array}{cc } \tilde{{\mbox{\ssb v}}}_{11}(r)&\tilde{{\mbox{\ssb v}}}_{12}(r)\\ \tilde{{\mbox{\ssb v}}}_{21}(r)&\tilde{{\mbox{\ssb v}}}_{22}(r ) \end{array } \right ] , \qquad { \it etc.}\ ] ] then the matrix @xmath427 takes the form @xmath428 , \qquad \tilde{{\mbox{\ssb h}}}_{11 } = 2\left [ \begin{array}{cc } { \mbox{\ssb j } } & { \mbox{\ssb 0}}\\ { \mbox{\ssb 0 } } & { \mbox{\ssb j } } \end{array}\right ] , \qquad { \mbox{\ssb j } } = \left [ \begin{array}{cc } 0 & -1\\ 1 & 0 \end{array}\right],\ ] ] and the set of eqs . ( [ r19 ] ) becomes @xmath429 the set ( [ r23 ] ) is purely algebraic . it permits us to express @xmath430 in the terms of @xmath431 : @xmath432 substitution of ( [ r24 ] ) into ( [ r22 ] ) yields the closed set of four first - order differential equations : @xmath433 where @xmath434.\end{aligned}\ ] ] here the upper sign corresponds to @xmath435 , and the lower sign corresponds to @xmath414 . the operator @xmath436 in eq . ( [ r25 ] ) is formally hermitian , _ i.e. _ , given @xmath96 , it is hermitian with respect to the inner product @xmath437 . but any two solutions of ( [ r25 ] ) , @xmath438 and @xmath439 , corresponding to different values of the energy , @xmath96 and @xmath440 , are not orthogonal . this is due to the nonlinear dependence of @xmath436 on @xmath96 . orthogonality can be instated by using the following definition of inner product : @xmath441 this inner product follows directly by substitution of eq . ( [ r24 ] ) into eq . ( [ r15 ] ) . the set of first order equations ( [ r25 ] ) can also be expressed as a second order equation . for this purposeit is convenient to permute the elements of @xmath438 by means of the matrix @xmath442 , where @xmath443 \quad { \rm for } \quad p=(-)^{j\pm1},{\label{r27}}\ ] ] and @xmath444 \quad { \rm for } \quad p=(-)^{j } . { \label{r28}}\ ] ] then , in terms of the two - dimensional blocks @xmath445 , \ ] ] the equations take the form @xmath446 where @xmath447 . elimination of @xmath448 leads to the following equation for the @xmath449 vector - function @xmath450 : @xmath451\psi_1(r ) = 0.\ ] ] eigenstates @xmath450 , @xmath452 corresponding to different values of the energy are orthogonal with respect to the inner product : @xmath453 which also follows from the reduction procedure . the form of equations ( [ r32 ] ) is convenient for examining the energy spectrum of @xmath26 bound states perturbatively to @xmath24 . for this purpose we introduce the dimensionless quantities , @xmath454 where @xmath455 is the reduced mass . we now perform a perturbative expansion in @xmath226 of equation ( [ r32 ] ) . to order @xmath456 equation ( [ r32 ] ) takes the hamiltonian form ( hereafter we omit the subscript 1 " of the wave function @xmath450 ) : @xmath457 where @xmath458\ ] ] is a two - component wave function , @xmath419 is the hamiltonian divided by @xmath459 and expressed in terms of dimensionless quantities ( [ p1 ] ) , @xmath460 is the unperturbed ( i.e. , @xmath54th - order ) hamiltonian , and @xmath461 is the perturbative correction to ( [ p4 ] ) . the form of the symmetric @xmath462 matrices @xmath463 , @xmath464 , @xmath465 depends on the parity . in order to obtain the energy spectrum to @xmath24 , it is sufficient to calculate the eigenvalues of @xmath117 to @xmath466 , _ i.e. _ , @xmath467 , where @xmath468 will be calculated exactly , while for @xmath469 first order perturbation theory in @xmath456 is sufficient . hence , the dependence of @xmath470 on @xmath117 is not crucial : to the accuracy required , it can be replaced by@xmath468 . in addition , the kernel of the inner product ( [ r32a ] ) can be set to unity . in the case @xmath412 we have @xmath471 , so that @xmath472 is the ( dimensionless ) radial coulomb hamiltonian @xmath473 with the angular momentum @xmath474 , repeated twice : @xmath475 , \qquad h_j = -{1 \over 2 } \left\{\frac{d}{d\rho^2 } - \frac{j(j+1)}{\rho^2}\right\ } - \frac{1}{\rho } .\ ] ] the matrices @xmath464 and @xmath476 are @xmath477,{\label{p7}}\\ { \cal m}(\rho,\lambda)&= & \frac18\left[\mbox{\scriptsize$ \begin{array}{cc } ( 1-\delta^2)\lambda(\lambda-\frac2{\rho } ) + \frac{1-\delta^2 - c^2\lambda ( 1+\delta^2)}{\rho^2}+\frac{1+\delta^2(1 -2c^2)}{\rho^3 } & \frac{2c\delta}{\rho^3}\\ \frac{2c\delta}{\rho^3 } & ( 1-\delta^2)\lambda^2 - \frac{c^2(1+\delta^2)}{\rho^2}(\lambda+\frac1{\rho } ) \end{array}$ } \ ! \right ] { \label{p8}}\end{aligned}\ ] ] the eigenvalues of the @xmath54th - order hamiltonian ( [ p6 ] ) , namely @xmath478 are two - fold degenerate , each with the two eigenstates @xmath479 \qquad { \rm and } \qquad \psi_{(2)}^{(0 ) } = \left[\begin{array}{c } 0\\ \end{array } \right],\ ] ] where @xmath480 is a solution of the coulomb problem @xmath481 . thus , the correction @xmath469 must be calculated appropriately for the degenerate situation : @xmath482,\ ] ] where the matrix @xmath483 is defined as follows : @xmath484.{\label{p12}}\end{aligned}\ ] ] the mass spectrum @xmath485 obtained with the use of ( [ p9 ] ) , ( [ p11 ] ) and ( [ p12 ] ) coincides neither with the muonium spectrum found in @xcite nor ( if @xmath486 ) with the spectrum of parapositronium ( see @xcite and refs . therein ) . the reason , like for the @xmath23 states , lies in the use of the empty vacuum ( [ 3.4 ] ) , which leads to single - particle states of positive and negative energy , and subsequently to the breit equation with its spurious term @xmath487 ( [ 8 - 7 ] ) in the hamiltonian @xcite , as discussed in connection with the @xmath23 states . we are going to show that the elimination of the contribution of @xmath487 from the spectrum leads to the correct result . first of all we transform @xmath487 into a radial representation . for this purpose we note that @xmath488 ( @xmath489 in the nonrelativistic limit ) , i.e. , @xmath490 contains only those components of @xmath390 ( [ 6 - 1 ] ) which are coefficients of bispinor harmonics " @xmath491 . similarly , @xmath492 ( @xmath493 in the nonrelativistic limit ) , i.e. , @xmath490 contains only coefficients of the bispinor harmonics " @xmath494 . thus , for @xmath412 parity states the spurious term ( divided by @xmath495 ) takes the following radial form : @xmath496=\frac{1-\delta^2}{8\rho^2}\left[\mbox{\scriptsize$ \begin{array}{cc } 4 & 0\\ 0 & 1 \end{array}$ } \right ] , \ ] ] where @xmath497 , and the corresponding matrix elements are @xmath498 . { \label{p15}}\end{aligned}\ ] ] if we now use @xmath499 , instead of @xmath483 , in ( [ p11 ] ) and ( [ p13 ] ) , we obtain the spectrum , @xmath500 which coincides with the results of connell @xcite and hersbach @xcite for the parity @xmath501 states . thus , correcting for the spurious terms in the breit hamiltonian , we obtain the expected @xmath24 results . in the @xmath502 case , the matrix @xmath463 is not diagonal : @xmath503.\ ] ] it can be diagonalized by means of the orthogonal transformation , using the matrix : @xmath504.\ ] ] then @xmath505,\ ] ] so that the @xmath54th - order hamiltonian becomes : @xmath506.\ ] ] it possesses the doubly degenerate eigenvalues ( [ p9 ] ) with eigenstates : @xmath507 \qquad { \rm and } \qquad \psi_{(2)}^{(0 ) } = \left[\begin{array}{c } 0\\ \end{array } \right],\ ] ] the first - order correction @xmath508 ( [ p5 ] ) , with the matrices @xmath509,{\label{p22}}\ ] ] @xmath510/\rho^2 } } & & \\ & & \hspace{-1em } \lefteqn{\scriptstyle { -a^2[7j^2 + 22j+9 - \delta^2(3j^2 + 8j+3)]/\rho^3 } } & \\ & & & \hspace{-1em } \lefteqn{\scriptstyle { + ( 1-\delta^2)j(j+1)^2(j+4)/\rho^4 } } \hspace{13em } \end{array } & \frac{ab(1-\delta^2)}{\rho}\left(2\lambda+\frac{1 + 2\lambda}{\rho } - \frac{c^2}{\rho^2}\right)\\ \frac{ab(1-\delta^2)}{\rho}\left(2\lambda+\frac{1 + 2\lambda}{\rho } - \frac{c^2}{\rho^2}\right ) & \begin{array}{llll } \lefteqn{\scriptstyle { ( 1-\delta^2)\lambda(\lambda + 2b^2/\rho ) } } & & & \\ & \lefteqn{\scriptstyle { -b^2[\{\lambda(6j^2+j-3)-1\ } - \delta^2\{\lambda(2j^2 + 3j-1)-1\}]/\rho^2 } } & & \\ & & \lefteqn{\scriptstyle { -b^2[7j^2 - 8j-6 - \delta^2(3j^2 - 2j-2)]/\rho^3 } } & \\ & & & \lefteqn{\scriptstyle { + ( 1-\delta^2)j(j+1)^2(j-3)/\rho^4 } } \hspace{16em } \end{array } \end{array}$ } \right ] , { \label{p23}}\ ] ] generates diagonal matrix elements only : @xmath511 \nonumber\\ & = & \left[\mbox{\scriptsize$ \begin{array}{cc } \frac{11+\delta^2}{32n^4 } - \frac{(7+\delta^2)j^2+(17-\delta^2)j+8 - 2\delta^2}{4(2j+1)(j+1)(2j+3)n^3 } & 0\\ 0 & \frac{11+\delta^2}{32n^4 } - \frac{(7+\delta^2)j^2 - 3(1-\delta^2)j-2}{4(2j-1)j(2j+1)n^3 } \end{array}$ } \right].{\label{p24}}\end{aligned}\ ] ] the energy spectrum ( [ p13 ] ) with @xmath512 , @xmath513 contains the contribution of the spurious term ( [ 8 - 7 ] ) . again , we present this term in radial form : @xmath514 \nonumber\\ = \frac{1-\delta^2}{8\rho^2 } \left[\mbox{\scriptsize$ \begin{array}{cc } b^2 & ab\\ ab & a^2 \end{array}$ } \right ] , \ ] ] where , for the present @xmath412 parity case , @xmath515 . the corresponding matrix elements are : @xmath516 . { \label{p26}}\end{aligned}\ ] ] the substitution of @xmath517 and @xmath518 into ( [ p13 ] ) yields the spectrum : @xmath519 which coincides with the results of connell @xcite and hersbach @xcite for the parity @xmath520 states . we have studied a reformulation of qed , in which the coupled dirac - maxwell field equations are partially decoupled by expressing the mediating photon field in terms of the dirac particle field , using covariant green s functions . this allows us to reformulate the hamiltonian of the theory so that the photon propagator appears directly in a quartic , nonlocal interaction term . we then consider a truncated model , in which there are no free ( physical ) photons . for such a model , each @xmath521-particle segment of the fock space of the quantized , equal - time hamiltonian is an invariant space , that is , there is no coupling among the various @xmath521-fermion segments . this is achieved by introducing an unconventional `` empty '' vacuum state . as a consequence , there exist exact few - particle eigenstates of the truncated hamiltonian , which lead to dirac - like two- and three - fermion wave equations . we show , in particular , that the two - fermion wave equation , in the coulomb gauge , is just the breit equation . for specific @xmath93 states , the breit equation reduces to radial form , and then to dirac - like equations for @xmath23 states , and to a coupled pair of schrdinger - like equations for @xmath26 states . perturbative solution of these equations yields @xmath522-corrections to the nonrelativistic rydberg spectrum , which do not reproduce the muonium spectrum as calculated by connell @xcite and hersbach @xcite ( nor the positronium spectrum in the @xmath486 case ) . the apparent reason for this disagreement is the mixing of positive- and negative - energy states , which is characteristic of the breit equation @xcite . however , agreement is achieved if we subtract the contribution of spurious operator ( [ 8 - 7 ] ) , which appears in the breit equation ( _ c.f . ( [ 8 - 9 ] ) , ( [ 8 - 5 ] ) , ( [ p16 ] ) , ( [ p27 ] ) ) . we have not been able to obtain analytic solutions of the radial equations . these radial wave equations have , in general , a singular point at @xmath523 , where @xmath96 is the two - fermion bound state energy ( rest mass ) . the existence of such an `` interior '' singularity makes it difficult to obtain numerical solutions of the radial boundary value problem by standard methods . the only exception is the case @xmath25 , @xmath486 , for which the radial equations are regular , and which we have therefore studied numerically . ( it is noteworthy that this is the case where the contribution of the operator ( [ 8 - 7 ] ) is zero . ) our numerical results for the equal - mass @xmath25 states , show that the dependence of the energy @xmath96 on the coupling constant @xmath226 is qualitatively similar to that obtained earlier for the coulomb qed model ( for which transverse - photon interactions are ignored ) @xcite . for low @xmath226 , the numerically obtained eigenenergies are in agreement with the result derived perturbatively . thereafter , @xmath245 decreases monotonically to @xmath524 as @xmath226 approaches a critical value @xmath247 . we find that @xmath261 ( in contrast to cqed value of @xmath525 ) . 99 j. w. darewych , annales fond . l. de broglie ( paris ) * 23 * , 15 ( 1998 ) , and `` interparticle interactions and nonlocality in quantum field theory '' in _ causality and locality in modern physics _ , g. hunter et al . ( eds . ) , kluwer academic publishers , dordrecht , 1998 , 333 - 344 . j. w. darewych , can . j. phys . * 76 * , 523 ( 1998 ) . m. barham and j. darewych , j. phys . a * 31 * , 3481 ( 1998 ) . wiley , new york , 1975 ) . j. d. jackson , _ classical electrodynamics _ ( john wiley , new york , 1975 ) . a. o. barut , _ electrodynamics and classical theory of fields and particles _ ( dover , new york , 1980 ) . w. t. grandy , jr . , _ relativistic quantum mechanics of leptons and fields _ ( kluwer , dordrecht , 1991 ) . a. o. barut , `` the schrdinger and dirac equation - linear , nonlinear and integrodifferential '' in _ geometrical and algebraic aspects of nonlinear field theory_. s. de filippo , m. marinaro , g. marmo and g. vilasi , eds . elsevier ( north holland ) , 1989 , pp . darewych and l. di leo , j. phys . a : math . gen . * 29 * , 6817 ( 1996 ) . g. hardekopf and j. sucher , phys . a * 30 * , 703 ( 1984 ) ; * 31 * , 2020 ( 1985 ) . p. m. stevenson and i. roditi , phys . d * 33 * , 2305 ( 1986 ) . r. koniuk and j. w. darewych , phys . b*176 * , 195 ( 1986 ) . w. dykshoorn and r. koniuk , phys . rev . a * 41 * , 64 ( 1990 ) . darewych and m. horbatsch , j. phys . b : at . mol . opt . * 22 * , 973 ( 1989 ) ; * 23 * , 337 ( 1990 ) . j. w. darewych , m. horbatsch and r. koniuk , phys . d * 45 * , 675 ( 1992 ) . b. ding and j. darewych , j. phys . g * 26 * , 907 ( 2000 ) . t. c. scott , j. shertzer and r. a. moore , phys . a * 45 * , 4393 ( 1992 ) . e. e. salpeter , phys . rev . * 87 * , 328 ( 1952 ) . h. a. bethe and e. e. salpeter , _ quantum mechanics of one- and two - electron atoms _ ( springer , berlin , 1957 ) . g. breit , phys . rev . * 34 * , 553 ( 1929 ) . a. eddington , proc . a * 122 * , 358 ( 1929 ) . j. a. gaunt , phil . trans . * 228 * , 151 ( 1929 ) ; proc . roy . soc . a * 122 * , 153 ( 1929 ) . v. n. golubenkov and ia . a. smorodinskii , zh . * 31 * , 40 , ( 1956 ) . jetp * 4 * , 55 ( 1957 ) ] . l. d. landau and e. m. lifshitz , _ the classical theory of fields _ ( pergamon , new york , 1975 ) , 4th revised english edition . b. m. barker and r. f. oconnell , phys . 78a * , 231 ( 1980 ) . b. m. barker and r. f. oconnell , ann . ( n. y. ) * 129 * , 358 ( 1980 ) . g. schfer , phys . lett . * 100a * , 128 ( 1984 ) . b. m. barker and r. f. oconnell , can . j. phys . * 58 * , 1659 ( 1980 ) . r. p. gaida , yu . b. kluchkovsky and v. i. tretyak , in _ constraint s theory and relativistic dynamics . proc . workshop _ , firenze , 1986 , g. longhi and l. lusanna , eds . ( world scientific , singapore , 1987 ) , pp . 210 - 241 . j. malenfent , phys . d * 38 * , 3292 ( 1998 ) ; phys . a * 43 * , 1233 ( 1991 ) . j. darewych , condensed matt . phys . * 1 * , 593 ( 1998 ) . v. shpytko and j. darewych , phys . d * 64 * , 045012 , 2001 . g. e. brown and d. g. ravenhall , proc . ( london ) * a208 * , 552 ( 1951 ) . z. v. chraplyvy , phys . rev . * 91 * , 388 ( 1953 ) . w. a. barker and f. n. glover , phys . * 99 * , 317 ( 1955 ) . j. h. connell , phys . d * 43 * , 1393 ( 1991 ) . i. t. todorov , phys . d * 3 * , 2351 ( 1971 ) . h. hersbach , phys . a * 46 * , 3657 ( 1992 ) . e. h. de groot and th . w. ruijgok , nucl . phys . * 101b * , 95 ( 1975 ) . j. r. sapirstein and d. r. yennie , theory of hydrogenic bound states " , eq . ( 2.3 ) p. 568 , in _ quantum electrodynamics _ , t. kinoshita , ed . , world scientific , singapore , 1990 . p. j. mohr , tests of fundamental physics " , eq . ( 28.39 ) p. 344 , in _ atomic , molecular and optical physics handbook _ , g. w. f. drake , ed . , american institute of physics , woodbury , n. y. , 1996 .

since the initial work of kobayashi and maskawa , the standard model ( sm ) of electroweak interactions is known to be able to accommodate the experimentally observed cp violation through a unique phase , @xmath1 , in the cabibbo kobayashi maskawa mixing matrix ( ckm ) . however , the available experimental information , namely @xmath2 and @xmath3 , is not enough to establish this phase as the only source of cp violation . most of the extensions of the sm include new observable phases that may significantly modify the pattern of cp violation . supersymmetry is , without a doubt , one of the most popular extensions of the sm . indeed , in the minimal supersymmetric extension of the sm ( mssm ) , there are additional phases which can cause deviations from the predictions of the sm . after all possible rephasings of the parameters and fields , there remain at least two new physical phases in the mssm lagrangian . these phases can be chosen to be the phases of the higgsino dirac mass parameter ( @xmath4 $ ] ) and the trilinear sfermion coupling to the higgs , ( @xmath5 $ ] ) @xcite . in fact , in the so called constrained minimal supersymmetric standard model ( cmssm ) , with strict universality at the grand unification scale , these are the only new phases present . it was soon realized that for most of the cmssm parameter space , the experimental bounds on the electric dipole moments of the electron and neutron constrained @xmath6 to be at most @xmath7 . consequently these new supersymmetric phases have been taken to vanish exactly in most studies of cmssm . however , in the last few years , the possibility of having non - zero susy phases has again attracted a great deal of attention . several new mechanisms have been proposed to suppress edms below the experimental bounds while allowing susy phases @xmath8 . methods of suppressing the edms consist of cancellation of various susy contributions among themselves @xcite , non - universality of the soft breaking parameters at the unification scale @xcite and approximately degenerate heavy sfermions for the first two generations @xcite . in a recent work @xcite , we showed that , in a model with heavy sfermions of the first two generations and in the large @xmath9 regime , @xmath2 and @xmath10 could receive very sizeable contributions from these new susy phases . similar studies @xcite including a larger set of experimental constraints have reported the impossibility of such large supersymmetric contributions . and @xmath11 @xcite . ] in this work , we are going to complete our previous analysis with the inclusion of all the relevant constraints in a cmssm scenario . in doing so we adopt a different perspective . we will assume from the very beginning that both supersymmetric phases are @xmath8 , ignoring for the moment edm bounds . in these conditions , and taking into account other cp conserving constraints , we will analyze the effects on the low energy cp violation observables , especially @xmath2 and @xmath10 . it should be noticed that the model used in @xcite can be easily obtained as a limit of the cmssm by decoupling the first two generations of squarks and neglecting the intergeneration mixing in the sfermion mass matrices . hence , our results in the more general cmssm will include this model as a limiting case . in the next section we study the new sources of flavor mixing present at the electroweak scale in any supersymmetric model . in section 3 we are going to analyze neutral meson mixing , i.e. @xmath12@xmath13 and @xmath14@xmath15 mixings , with large supersymmetric phases . section 4 will be devoted to the study of the branching ratio of the decay @xmath16 . in section 5 we will show the impact of the measured @xmath16 branching ratio on the supersymmetric contributions to @xmath2 and @xmath10 . section 6 will present our conclusions and in appendices a and b we collect , respectively , the formulas for the integration of relevant rge s and the different loop functions appearing in the text . the issue of flavor changing neutral current ( fcnc ) interactions in the cmssm has been widely investigated in the literature . for the completeness of the discussion , we briefly recall those properties which will be relevant for our analysis . the cmssm is the simplest supersymmetric structure we can build from the sm particle content . this model is completely defined once we specify the soft - supersymmetry breaking terms . these are taken to be strictly universal at some ultra - high energy scale , which we take to be the grand unification scale ( @xmath17 ) . @xmath18 that is , there is a common mass for all the scalars , @xmath19 , a single gaugino mass , @xmath20 , and all the trilinear soft - breaking terms are directly proportional to the corresponding yukawa couplings in the superpotential with a proportionality constant @xmath21 . now , with the use of the renormalization group equations ( rge ) of the mssm , as explained in appendix a , we can obtain the whole supersymmetric spectrum at the electroweak scale . all the supersymmetric masses and mixings are then a function of @xmath19 , @xmath20 , @xmath22 , @xmath23 , @xmath24 and @xmath9 . we require radiative symmetry breaking to fix @xmath25 and @xmath26 @xcite with tree level higgs potential . and @xmath27 are switched off . for this reason we prefer to give the relevant rge s with non vanishing susy phases in appendix a. ] it is important to notice that , even in a model with universal soft breaking terms at some high energy scale as the cmssm , some off diagonality in the squark mass matrices appears at the electroweak scale . working on the basis where the squarks are rotated parallel to the quarks , the so called super ckm basis ( sckm ) , the squark mass matrix is not flavor diagonal at @xmath28 . this is due to the fact that at @xmath17 there exist two non - trivial flavor structures , namely the two yukawa matrices for the up and down quarks , which are not simultaneously diagonalizable . this implies that through rge evolution some flavor mixing leaks into the sfermion mass matrices . in a general supersymmetric model , the presence of new flavor structures in the soft breaking terms would generate large flavor mixing in the sfermion mass matrices . however , in the cmssm , the two yukawa matrices are the only source of flavor change . as always in the sckm basis , any off - diagonal entry in the sfermion mass matrices at @xmath28 will be necessarily proportional to a product of yukawa couplings . the rge s for the soft breaking terms are sets of linear equations , and thus , to match the correct quirality of the coupling , yukawa couplings or tri - linear soft terms must enter the rge in pairs , as we can see in eqs . ( a.1-a.3 ) in appendix a. in fact , in the up ( down ) squark mass matrix the up ( down ) yukawas will also be diagonalized and so will mainly contribute to diagonal entries while off diagonal entries will be due to the down ( up ) yukawa matrix . this means , for instance , that in this model the off - diagonality in the @xmath29 matrix will roughly be @xmath30 . with @xmath31 a proportionality factor that typically is , @xmath32 as expected from the loop factor and the running from @xmath17 to @xmath28 nevertheless , we have to keep in mind that this is simply a typical estimate and the final value of @xmath31 can suffer a sizeable variation depending on many other factors not present in this simple estimate . on the other hand , this has clear implications on the @xmath9 dependence of these off diagonal entries of the sfermion mass matrices . in the basis where the down yukawa matrix is diagonal , we can write the up and down yukawas as , @xmath33 with @xmath34 the cabibbo - kobayashi - maskawa mixing matrix and @xmath35 the diagonalized mass matrices for the quarks . we can see in this equation that , for @xmath36 , the up yukawa matrix will maintain similar values when going to large @xmath9 . hence , the off diagonal entries in the down squarks mass matrix will be roughly stable with @xmath9 . in the up squark mass matrix we have the opposite situation and the @xmath9 dependence is very strong . in this case the off diagonal entries depend on the down yukawa matrix that grows linearly with @xmath9 for large @xmath9 . this means that we can expect the flavor change in the up and down squark mass matrix to be similar when @xmath37 . while for @xmath38 the flavor change in the up mass matrix will be approximately @xmath39 times smaller ( see appendix a for details ) . these points also apply to the left right sub matrices where again flavor changing entries will be due to the opposite isospin yukawa matrix . in fact , this left right sfermion mixing only appears after electroweak symmetry breaking . the expression for these matrices in the sckm basis is , @xmath40 ) . this is always true for all the generation changing entries that are produced by the @xmath41 matrices . however , in the down mass matrix , this suppression can be partially compensated by a large value of @xmath9 in the diagonal terms proportional to @xmath42 . these are all well known facts in the different studies of fcnc processes in the framework of the cmssm @xcite and imply that flavor mixing is still dominantly given by the usual ckm mixing matrix in w - boson , charged higgs and chargino vertices . in this work , we are especially interested in cp violating observables . then we must also consider the presence of observable phases in the sfermion mass matrices . in the following we will take the ckm matrix exactly real to isolate pure effects of the new supersymmetric phases @xcite . the sfermion mass matrices contain several physical phases that give rise to cp violation phenomena . in particular , before rge evolution , these phases ( @xmath23 , @xmath24 ) are confined to the left - right part of the sfermion mass matrix while both the left left , @xmath43 , and right right , @xmath44 , matrices are real and diagonal . however this is not true anymore at @xmath28 : @xmath23 leaks into the off - diagonal elements of these hermitian matrices through rge evolution . from the explicit rge in the mssm , eq.([mqrge ] ) , it is clear that this phase only enters the @xmath45 evolution through the combinations @xmath46 or @xmath47 . at @xmath17 these matrices have a common phase , and so the combination @xmath48 is exactly real . so , to the extent that the @xmath41 matrices keep a uniform phase during rge evolution , no phase will leak into the @xmath49 matrices . however , we can easily see from eqs.(a.2 ) and ( a.3 ) that this is not the case , and different elements of the @xmath41 matrices are renormalized differently . in this equation , we can see that only the terms involving two yukawa and one @xmath41 matrix can produce a mismatch in the phases . moreover , these terms will only be important when there are no small yukawas involved . then , we can expect a mismatch only on the off diagonal elements involving the third generation . keeping this in mind , the general form of the @xmath49 matrix at @xmath28 in terms of the initial conditions is , @xmath50 where the coefficients @xmath51 are @xmath52 matrices with real numerical entries . in this expression we can see that the presence of imaginary parts will be linked to the non - symmetric part of the @xmath53 matrices . as is clear from the mass matrices in appendix a ( eqs . ( a.5-a.7 ) and ( a.12-a.14 ) ) , these non symmetric parts of @xmath54 are always more that three orders of magnitude smaller than the corresponding symmetric parts . this means that , in the sckm basis , the imaginary parts of any mass insertion are present only in one part per @xmath55 , and are always associated with @xmath56 mi , eqs . ( a.8-a.11 ) and ( a.15-a.18 ) . a very similar situation was also found by bertolini and vissani , in the cmssm with vanishing susy phases for the leakage of @xmath1 @xcite . so , we conclude that in the processes we will consider , we can take both @xmath57 and @xmath58 as real to a very good approximation . in the following we will analyze the new effects of this model on indirect cp violation in @xmath59 and @xmath60 systems . in doing so , we will use both the exact vertex mixing method and the mass insertion ( mi ) approximation @xcite . notice that the mi approximation is extremely good in the case of the cmssm where all the off - diagonal entries are sufficiently small . the size of these off diagonal entries directly gives in the mi approximation the amount of flavor changing induced by the sfermion mass matrices . a possible exception may arise in the stop and sbottom sectors that , in any case , could be diagonalized to ensure the validity of the mi approximation @xcite . as we will see in the next sections , this is frequently useful to understand the exact results obtained in the vertex mixing method . in the sm neutral meson mixing arises at one loop through the well known @xmath61box . however , in the cmssm , there are new contributions to @xmath62 processes coming from boxes mediated by supersymmetric particles . these are charged higgs boxes ( @xmath63 ) , chargino boxes ( @xmath64 ) and gluino - neutralino boxes ( @xmath65 , @xmath66 ) . the amount of the indirect cp violation in the neutral meson @xmath67 system is measured by the well known @xmath68 parameter : @xmath69 where @xmath70 is the @xmath67@xmath71 mass splitting . @xmath68 depends on the matrix elements of the @xmath62 hamiltonian , @xmath72 , which can be decomposed as @xmath73 where the relevant four fermion operators are given by @xmath74 with @xmath75 for the @xmath59 and @xmath60systems respectively and @xmath76 as color indices . in the cmssm , these are the only three operators present in the limit of vanishing @xmath77 . at this point , we are going to divide our discussion into two parts . we analyze separately the effective operator @xmath78 that preserves chirality along the fermionic line , and the operators @xmath79 and @xmath80 that change chirality along the fermionic line . as we will see below the flavor mixing in the sfermion mass matrix and the experimental constraints on both kinds of operators are very different . in eq.([df=2 ] ) , @xmath78 is the only operator present that does not involve a chirality change in the fermionic line . with respect to the associated sfermion , no chirality change in the sfermion propagator will be needed , and so , the suppression associated with left right sfermion mixing can be avoided . in general , @xmath81 can be decomposed as follows @xmath82 the usual sm @xmath61box , where all the couplings are purely left handed can only contribute to this effective operator . however , with @xmath83 , @xmath84 does not contain any complex phase and hence can not contribute to the imaginary part in @xmath85 . in any case it will always be , in the cmssm , the dominant contribution to @xmath70 . similarly , the charged higgs contribution , @xmath86 depends on the same combination of ckm elements with no other cp violating phase @xcite . so it will not contribute to our cp violating observable . gluino and neutralino contributions to @xmath87 are specifically supersymmetric . they involve the superpartners of quarks and gauge bosons . here , the source of flavor mixing is not directly the usual ckm matrix . it is the presence of off diagonal elements in the sfermion mass matrices , as discussed in the previous section . from the point of view of cp violation , we will always need a complex wilson coefficient . in the sckm basis all gluino vertices are flavor diagonal and real . this means that in the mi approximation , we need a complex mass insertion in one of the sfermion lines . as explained in the previous section , these mi are proportional to yukawa couplings and real up to 1 part in @xmath88 . the complete expressions for the gluino contributions to @xmath62 processes in the mi approximation can be found in @xcite . the bounds obtained there for the real and imaginary parts of the mass insertions required to saturate @xmath89 and @xmath2 are , @xmath90 where @xmath91 is an average squark mass . in the cmssm , as we can see in appendix a , these mass insertions are much smaller . in particular , the fact that the bound on @xmath89 , the real part of the mi , is satisfied implies that the imaginary parts are at least two orders of magnitude below the required value to saturate @xmath2 . hence , no sizeable contributions to @xmath2 from gluino boxes are possible . the situation in @xmath14@xmath15 mixing is completely analogous : assuming that the minimum phase from the mixing observable in the b factories is around @xmath92 radians , we would need an imaginary contribution not more than one order of magnitude below the real one . with the arguments given above , this is clearly out of reach for gluino boxes in the cmssm . neutralino contributions are generally smaller than gluino due to smaller couplings with the same source of flavor mixing . in fact , although neutralino vertices in the sckm basis also involve the complex neutralino mixings , any imaginary part on this operator will only be due to a complex mass insertion . this can be seen in the explicit expressions in @xcite where all neutralino mixings in this operator appear in pairs with its complex conjugate counterpart . finally , the charginos also contribute to @xmath93 . in this case , flavor mixing comes explicitly from the ckm mixing matrix , although off diagonality in the sfermion mass matrix introduces a small additional source of flavor mixing . @xmath94 \end{aligned}\ ] ] where @xmath95 , @xmath96 , and @xmath97 represent the coupling of chargino , @xmath98 , and squark @xmath99 to the left handed down quark , @xmath100 . finally @xmath101 is an intermediate up quark index associated with the factorization of the ckm mixing matrix . the expression for this coupling is then , @xmath102 where @xmath103 and @xmath104 are @xmath105 matrices such that the @xmath106 unitary matrix @xmath107 diagonalizes the up squark mass matrix , @xmath108 . @xmath109 is one of the matrices that diagonalize the chargino mass matrix through a bi unitary transformation @xmath110 , with , @xmath111 from these equations it is clear that @xmath112 will in general be complex , as both @xmath24 and @xmath23 are present in the different mixing matrices . the loop function @xmath113 is given in eq . ( b.1 ) of appendix b. the main part of @xmath114 in eq([chwc ] ) will be given by pure ckm flavor mixing , neglecting the additional flavor mixing in the squark mass matrix @xcite . in this case , @xmath115 and @xmath116 , we have , @xmath117 \end{aligned}\ ] ] but , taking into account that @xmath118 is symmetric under the exchange of any pair of arguments we have , @xmath119 and so @xmath120 , is exactly real @xcite . this is not exactly true in the cmssm , where there is additional flavor change in the sfermion mass matrices . here , some imaginary parts appear in the @xmath114 in eq([chwc ] ) . being associated to the size of intergenerational sfermion mixings , these imaginary parts will be maximal for large @xmath9 . in fig([imk40 ] ) we show in a scatter plot the size of imaginary and real parts of @xmath114 in the k system for a fixed value of @xmath121 . the region of susy parameters explored in this and all of the following scatter plots is @xmath122 and @xmath123 . with these initial conditions we impose that all squarks are heavier than @xmath124 with the exception of the stops that , as the charginos , are only required to be above @xmath125 . furthermore we impose the constraint from the @xmath16 decay . notice that , as we will see later , this is a conservative attitude , in the sense that other constraints that we do not impose could only make our conclusions stronger . under these conditions , we can see here that , in the cmssm , this wilson coefficient is always real up to a part in @xmath126 . fig([imb40 ] ) is the equivalent plot for the case of @xmath14@xmath15 mixing . here , imaginary parts are relatively larger but , in any case , out of reach for the foreseen b factories . taking this into account , from the point of view of experimental interest , we will always neglect imaginary parts in the wilson coefficient @xmath127 within the cmssm . notice that this would not apply in a general model with non - universality at the gut scale @xcite , and each particular model should be considered separately . from the point of view of flavor change and cp violation , operators @xmath79 and @xmath80 are different from @xmath78 . these two operators always involve a change in the chirality of the external quarks , and consequently also a change of the chirality of the associated squarks or gauginos . in particular , this implies the direct involvement of the supersymmetric phases . on the other hand , these operators are suppressed by the presence of down quark yukawa couplings , and so can only be relevant in the region of large @xmath9 @xcite . we can write the different contributions to @xmath128 and @xmath129 as , @xmath130 in first place , charged higgs contributes only to @xmath128 , but parallel to the discussion for @xmath131 , the absence of phases prevents it from contributing to @xmath132 . gluino and neutralino boxes contribute both to @xmath79 and @xmath80 . however flavor change will be given in this case by an off diagonal left right mass insertion . in the cmssm these mi are always proportional to the mass and are never enhanced by large @xmath9 values ( see eq.([lr ] ) ) of the right handed squark . this implies that these left right flavor transitions from gluino will always be smaller in the cmssm than the corresponding chargino contributions , where flavor change is directly given by the ckm matrix . in fact , this is already well known for the case of @xmath16 decay @xcite , which is completely equivalent from the point of view of flavor change . hence , the most important contribution , especially for light stop and chargino , will be the chargino box . before the inclusion of qcd effects , it contributes solely to the coefficient @xmath129 , @xmath133 where @xmath134 is the coupling of chargino , @xmath98 , and squark , @xmath99 , to the right handed down quark @xmath100 , with , @xmath135 and @xmath136 is given in eq . ( b.2 ) . unlike the @xmath114 wilson coefficient , @xmath137 is complex even in the absence of intergenerational mixing in the sfermion mass matrices @xcite . in fact , the presence of flavor violating entries in the up squark mass matrix hardly modifies the results obtained in their absence @xcite . so , in these conditions we have , @xmath138\\ f_s(\alpha , k,\gamma , l , i , j)= \frac{m_q^2}{2 m_w^2 \cos^2 \beta } h^{(\alpha , k)i}g^{(\alpha , k)j * } g^{(\gamma , l)i * } h^{(\gamma , l)j } y_2(z_{k } , z_{l } , s_i , s_j ) \nonumber\end{aligned}\ ] ] where we have used ckm unitarity and degeneracy of the first two generations of squarks . due to the differences between @xmath139 and @xmath140 couplings , this contribution is always complex in the presence of susy phases . the most relevant feature of eqs.([chwcr ] ) and ( [ chwcr0 ] ) is the explicit presence of the external quark yukawa coupling squared , @xmath141 . this is the reason why this contribution is usually neglected in the literature @xcite . however , as we showed in @xcite , this contribution could be relevant in the large @xmath9 regime . for instance , in @xmath14@xmath15 mixing we have @xmath142 that for @xmath143 is larger than 1 and so , it is not suppressed at all when compared with the @xmath114 wilson coefficient . this means that this contribution can be very important in the large @xmath9 regime @xcite and could have observable effects in cp violation experiments in the new b factories . however , in our previous work @xcite , we did not include the additional constraints coming from @xmath16 decay . in the next sections we will analyze the relation of @xmath132 with this decay , and the constraints imposed by its experimental measure . the decay @xmath16 has already been extensively studied in the context of the cmssm with vanishing susy phases @xcite . because the branching ratio is a cp conserving observable , the presence of new phases will not modify the main features found in @xcite concerning the relative importance of the different contributions . however , in the presence of the new susy phases , these contributions will have different phases and will be observable through the interference . as we will see next , the experimental constraints will also have a large impact on the imaginary parts of the decay amplitudes . this decay is described by the following @xmath144 effective hamiltonian @xmath145 where the relevant operators are given by @xmath146 here @xmath147 , and the wilson coefficients @xmath148 can be decomposed according with the particles in the loop , @xmath149 among these contributions , the @xmath61 penguin is exactly the same as in the sm and it does not depend on any supersymmetric parameters , it is simply a function of sm couplings and masses . this contribution is @xcite , @xmath150 with @xmath151 and @xmath152 the charge of the up quarks . similarly , in the charged higgs penguins all the variables are known with the exception of @xmath153 . again this contribution is unchanged by the inclusion of the new susy phases , @xmath154 where @xmath155 . this contribution gives a sizeable correction to the @xmath16 decay that constrains the mass of the charged higgs in two higgs doublet models or in the mssm with low @xmath9 . however , in the case of moderate large @xmath9 , chargino contributions may partially compensate this charged higgs contribution relaxing the constraints @xcite . in addition to the @xmath156 and charged higgs contributions analyzed above , there are three specifically supersymmetric contributions mediated by gluino , neutralino and chargino . in gluino or neutralino diagrams flavor change is due to the off diagonality in the sdown mass matrix . being left right flavor off diagonal transitions , they are suppressed by the mass of the @xmath157 quark . indeed , smallness of gluino and neutralino contributions has already been established in @xcite where it was shown that , in the cmssm , such contributions are roughly one order of magnitude smaller than the chargino contribution . together with the @xmath158 and charged higgs , the most important supersymmetric contribution will be , especially in the large moderate @xmath9 regime , the chargino contribution . in the @xmath61 and charged higgs contributions , the necessary chirality flip for the dipole amplitude is always proportional to @xmath159 . however , in the chargino penguin the chirality flip can be made either through a chargino mass insertion in the loop or through an external leg mass insertion proportional to @xmath159 . in fact , as pointed out in @xcite , this enhancement due to @xmath160 is partially compensated by the presence of the @xmath157 yukawa coupling . nevertheless , this compensation is only effective for low values of @xmath161 . in terms of the chargino quark squark couplings used in the previous section , these contributions are , @xmath162 with the loop functions defined in appendix b. similarly to the situation for the wilson coefficient @xmath129 , we can , to a very good approximation , neglect the presence of intergenerational mixing in the up squark mass matrix @xcite , then , @xmath163 where , once more , we use ckm unitarity and degeneracy of the first two generations of squarks . the second term in @xmath164 in eq . ( [ char ] ) , which corresponds to the chargino mass insertion in the loop , is dominant in the large @xmath9 regime . notice that both @xmath165 and @xmath166 are products of the squark and chargino mixing matrices that can be @xmath167 ( in the case of flavor diagonal stop mixings ) . then , for stop and chargino masses around the electroweak scale , this term has an extra enhancement of @xmath168 . this means that , for large @xmath9 , we can approximate these wilson coefficients as , @xmath169 as we have seen in section iii , chargino contribution to the @xmath129 wilson coefficient , eq.([chwcr0 ] ) , is the main contribution to indirect cp violation of the new supersymmetric phases for large values of @xmath9 . however , if we compare this wilson coefficient with the chargino contribution to the decay @xmath16 , eqs . ( [ chwcr0 ] ) and ( [ char ] ) , we can see that both chargino contributions are deeply related . in fact , if we make a rough approximation and assume that the two different loop functions involved are of the same order , i.e. , @xmath170 we would obtain , @xmath171 of course , this can not be considered as a good approximation . as we can see from their explicit expressions in appendix b , the loop functions are clearly different . anyway , they can be expected to give results of the same order of magnitude . so , the order of magnitude of @xmath129 is determined by the allowed values of @xmath172 , as we will explicitly show below . to reach this goal , we will follow @xcite , where they constrain in a model independent way new physics contributions to the wilson coefficients involved in the @xmath16 decay . in terms of these wilson coefficients , the branching ratio , @xmath0 is , @xmath173 + 0.161 re[\xi_{8 } ] + 0.083 re[\xi_{7}\xi_{8}^ { * } ] \nonumber\end{aligned}\ ] ] where @xmath174 . the different coefficients appearing in eq.([constrain ] ) are the sm renormalization group evolved contributions , that must be recovered in the limit @xmath175 . the numerical values are taken from @xcite . we have not taken into account the errors associated with the choice of the scale and the restrictions on the photon energy that do not modify our conclusions . now , using the experimental measure , @xmath176 , we can constrain the allowed values of the complex variables @xmath177 and @xmath178 . in fact , we can already see from eq . ( [ constrain ] ) , that in the approximation @xmath179 this is simply the equation of an ellipse in the @xmath180$]@xmath181 $ ] plane . in the case of supersymmetry with large @xmath9 , the new physics contribution to @xmath177 and @xmath178 will be mainly due to the chargino . the allowed values of @xmath182 directly constrain then the chargino contributions to @xmath183 and indirectly constrain the values of @xmath184 . in figure [ figc7 ] , we show a scatter plot of the allowed values of @xmath185 versus @xmath186 in the cmssm for a fixed value of @xmath9 with the constraints from eq.([constrain ] ) . notice , that a relatively large value of @xmath9 , for example @xmath187 , is needed to compensate the @xmath61 and charged higgs contributions and cover the whole allowed area with positive and negative values . however , the shape of the plot is clearly independent of @xmath9 , only the number of allowed points and its location in the allowed area depend on the value considered . in this figure we take @xmath188 because only a large value could give rise to observable cp violation @xcite . the values of @xmath172 and @xmath189 used here are the values obtained in the cmssm for a given set of initial conditions . although we do not use the approximation @xmath179 this does not modify the elliptic shape of the plot . figure [ figc3 ] shows the allowed values for a re scaled wilson coefficient @xmath190 corresponding to the same allowed points of the susy parameter space in figure[figc7 ] . as we anticipated previously , the allowed values for @xmath191 are close to the square of the values of @xmath172 in figure [ figc7 ] slightly scaled by different values of the loop functions . this is the proof of the importance of the @xmath16 constraint on the chargino contributions to indirect cp violation . we can immediately translate this result to a constraint on the size of the chargino contributions to @xmath132 . @xmath192\end{aligned}\ ] ] in this expression @xmath193 , @xmath194 and @xmath195 denote the mass , mass difference and decay constant of the neutral meson @xmath196 . the coefficient @xmath197 @xcite includes the rge effects from @xmath28 to the meson mass scale , @xmath42 , and @xmath198 is the b parameter associated with the matrix element of the @xmath80 operator @xcite . then , for the @xmath59 system , using the experimentally measured value of @xmath199 we obtain , @xmath200 \approx 0.4 \times 10^{-7 } im[\bar{c}_3]\end{aligned}\ ] ] given the allowed values of @xmath191 in fig.[figc3 ] , this means that in the cmssm , even with large susy phases , chargino can not produce a sizeable contribution to @xmath2 . we have seen in section iii that also gluino and neutralino give negligible contributions in the cmssm or in a model without off diagonal soft breaking terms at the gut scale . hence indirect cp violation in the kaon system will be mainly given by the usual sm box and the presence of a cp violating phase in the ckm matrix , @xmath1 is still needed . the case of @xmath14@xmath15 mixing has a particular interest due to the arrival of new data from the b factories . in fact , as explained at the end of section iii and in @xcite , in the large @xmath9 regime chargino contributions to indirect cp violation can be very important . however , for any value of @xmath9 we must satisfy the bounds from the @xmath16 decay . so , if we apply these constraints to the @xmath14@xmath15 mixing , @xmath201 \approx 0.5 \times 10^{-3 } im[\bar{c}_3]\end{aligned}\ ] ] where once again , with the allowed values of fig.[figc3 ] we get a very small contribution to cp violation in the mixing . we must take into account that the mixing induced cp phase , @xmath202 , measurable in @xmath14 cp asymmetries , is related to @xmath10 by @xmath203 . the expected sensitivities on the cp phases at the b factories are around @xmath204 radians , so this supersymmetric chargino contribution will be completely out of reach . gluino and neutralino contributions to indirect cp violation can also be discarded in the cmssm . once again we have to conclude that no new contributions to indirect cp violation from the new susy phases will be observable in @xmath14 cp asymmetries in the framework of the cmssm . recently , cdf @xcite has provided preliminary indications that @xmath205 is in agreement with the sm predictions . clearly , from the above result , eq . ( [ epsb ] ) , it appears that the cmssm contribution is too small by itself to account for this result . in this work , the effects of non - vanishing supersymmetric phases on indirect cp violation in @xmath59 and @xmath60 systems have been analyzed within the cmssm . we have found that operators involving only left handed external quarks are not sensitive to these new phases at an observable level . this is due to the absence of intergenerational mixings beyond those originated from the ckm matrix . on the contrary , operators involving both right and left handed quarks are in general complex , even in the absence of @xmath1 , and could be relevant in the large @xmath9 regime . however , we have shown that these contributions are deeply related with the @xmath0 decay . so , taking into account the constraints coming from this decay these contributions also turn out to be too small to be measured experimentally . although these conclusions are specific for indirect cp violation , they could also be implemented for chargino mediated direct cp violation in the decays . again , in these decays the same chargino quark squark couplings are involved and we can also expect a big impact of the @xmath16 constrain . in fact , the conclusions reached in this paper are far more general . the correlation between @xmath16 and susy induced indirect cp violation exists in any supersymmetric model with sufficiently small intergenerational mixings in the sfermion mass matrices . this would include specifically all the models without new flavor structures beyond the usual ckm matrix at the gut scale and simplified models as the one the authors used in @xcite . in summary , concerning the simpler supersymmetric models , like cmssm , the constraints coming from @xmath0 decay are sufficient to rule out pure supersymmetric indirect cp violation in @xmath59 and @xmath60 systems , even in the absence of any electric dipole moment constraints . this has very important consequences for the supergravity induced models where a cancellation between different supersymmetric contributions allows large supersymmetric phases while respecting edm bounds @xcite . in these models , even in the regions of parameter space where this cancellation occurs , no observable effect of the large susy phases will appear on indirect cp violation experiments . however , as pointed out by baek and ko @xcite , these phases would still be observable in cp asymmetries in the @xmath16 decay . all this means that the presence of large susy phases is not sufficient to produce observable effects at the low energy experiments . in particular , new sources of flavor change beyond the usual ckm matrix are needed . and so , any deviation from the sm expectations at indirect cp violation experiments due to supersymmetry should be taken as a sign of non universality of the soft breaking terms . in this context one recalls the recent studies on superstring compactifications with non universal gaugino masses @xcite . in this appendix we describe the new features of the integration of rge s in the cmssm with non - vanishing susy phases relevant to our analysis . the complete matrix form of the rg equations can be found in @xcite . using their notation and conventions , with the only change of @xmath206 , we will mainly concentrate on the left left scalar quark mass matrix and the tri linear soft breaking coupling evolution , @xmath207 \nonumber + \\ & \frac{1}{2 } [ \tilde{y}_d\tilde{y}_d^\dagger m_q^2 + m_q^2 \tilde{y}_d \tilde{y}_d^\dagger + 2 ( \tilde{y}_d m_d^2 \tilde{y}_d^\dagger + \bar{\mu}^2_1 \tilde{y}_d\tilde{y}_d^\dagger + \tilde{a}_d \tilde{a}_d^\dagger ) ] \end{aligned}\ ] ] @xmath208 \end{aligned}\ ] ] @xmath209 \end{aligned}\ ] ] except for the yukawa coupling matrices , the rge s of all other quantities are linear @xcite . this means , in particular , that rge s of all soft masses , though coupled , can be solved as a linear combination of the gut scale parameters @xmath210 , @xmath211 and @xmath212 at any scale below @xmath213 . however , one notices that the initial conditions on the trilinear couplings require the knowledge of the particular yukawa texture at the unification scale . to do this , we numerically integrate the yukawa rge with a given value of @xmath9 and in terms of the fermion masses and the ckm matrix . specifying the gut scale initial conditions in this way , it is straightforward to compute all soft masses at @xmath214 for arbitrary values of @xmath210 , @xmath211 and @xmath212 . thanks to the linearity of the corresponding rge s the soft masses at @xmath214 admit the following expansion , @xmath215 where the coefficients @xmath101 and @xmath51 are @xmath52 matrices with real numerical entries . one notices that the matrices @xmath216 would be completely real were it not for the non symmetric terms in the matrix @xmath217 . however , it will be seen from the specific examples that this matrix remains nearly symmetric , and thus , cp violating entries @xmath216 are extremely suppressed . moreover , one notices that @xmath218 carries , in general , large cp violating phases ; however , these terms are effective only for intragenerational @xmath219-type mixings . hence , this particular observation shows the importance of chargino contributions for cp violation in fcnc processes , as explained in sec . ii . as mentioned before , due to the non linearity of the rge s for yukawa matrices , it is not possible to give a fully analytic solution for the soft mass parameters . nevertheless , once we fix @xmath220 , we can numerically integrate the yukawa rge . therefore , below we give semi analytic solutions of rge s for @xmath221 and @xmath222 to illustrate the small and large @xmath220 regimes . fixing @xmath223 , we get for the relevant @xmath51 matrices in eq.([etas ] ) , @xmath224 where the vanishing off diagonal entries in the last matrix mean values smaller than @xmath225 in absolute magnitude . among the matrices involved in eq.([etas ] ) , @xmath226 is always the largest one for similar values of @xmath212 and @xmath227 . so , it sets the scale of the matrix element while @xmath228 is the only one that can produce an imaginary part . hence , we do not specify the other @xmath51matrices , which are not important for our discussion . once we obtain the @xmath229 matrix with the help of eq.([etas ] ) we can get the values of the @xmath230 and @xmath231 in the sckm basis , that give the size of flavor change in the squark mass matrices compared with the diagonal elements . for @xmath221 , those elements of the squark mass squared matrix causing @xmath232 transitions between first and second , as well as second and third generations , are given by @xmath233 now , we repeat the same quantities above for @xmath222 : @xmath234 @xmath235 a comparison of the corresponding quantities in @xmath221 and @xmath222 cases reveals the sensitivity of the results on @xmath220 . as explained in sec . ii , @xmath236 remains nearly unchanged while @xmath237 assumes an order of magnitude enhancement as @xmath220 varies from 2 to 40 . this change in @xmath237 affects various quantities as dictated by the differential equations ( a.1 a.3 ) . in this appendix we collect the different loop function used in the text . the functions @xmath238 and @xmath239 entering @xmath60@xmath240 and @xmath59@xmath241 mixings are given by @xmath242 and @xmath243\,.\end{aligned}\ ] ] @xmath245 , \\ f_{r}^{7}(x , y ) & = & \frac{1}{x } \left[q_{u}f_{4}(y / x)+f_{3}(y / x)\right ] , \\ f_{l}^{8}(x , y ) & = & \frac{1}{x}f_{2}(y / x ) , \\ f_{r}^{8}(x , y ) & = & \frac{1}{x } f_{4}(y / x)\,.\end{aligned}\ ] ] m. dugan , b. grinstein and l.j . hall , _ nucl . phys . _ * b 255 * , 413 ( 1985 ) ; + s. dimopoulos and s. thomas , _ nucl * b 465 * , 23 ( 1996 ) . t. ibrahim and p. nath , _ phys . rev . _ * d57 * , 478 ( 1998 ) ; + t. ibrahim and p. nath , _ phys . rev . _ * d58 * , 111301 ( 1998 ) ; + m. brhlik , g.j . good and g.l . kane , _ phys . * d59 * , 115004 ( 1999 ) hep - ph/9810457 . abel and j.m . rev . _ * d 55 * , 1623 ( 1997 ) ; + s. khalil , t. kobayashi and a. masiero , _ phys . rev . _ * d60 * , 075003 ( 1999 ) , hep - ph/9903544 . s. dimopoulos and g.f . giudice , _ phys . * b 357 * , 573 ( 1995 ) ; + a. cohen , d.b . kaplan and a.e . nelson , _ phys . * b 388 * , 599 ( 1996 ) ; + a. pomarol and d. tommasini , _ nucl * b 466 * , 3 ( 1996 ) . demir , a. masiero and o. vives , phys . * 82 * ( 1999 ) 2447 , hep - ph/9812337 . s. baek and p. ko , korea inst . sci report no . kaist - th-99 - 1 , april 1999 hep - ph/9904283 ; + s. baek and p. ko , phys . rev * 83 * ( 1999 ) 488 , hep - ph/9812229 . s. baek , j. h. jang , p. ko and j. h. park , korea inst . sci report no . kaist-99 - 04 , july 1999 , hep - ph/9907572 . n.k . falck , z. phys . * c30 * ( 1986 ) 247 . s. bertolini , f. borzumati , a. masiero , and g. ridolfi , nucl . phys . * b353 * ( 1991 ) 591 . p. cho , m. misiak and d. wyler , phys . rev . * d54 * ( 1996 ) 3329 , hep - ph/9601360 g. c. branco , f. cagarrinho , and f. kruger , phys . lett . * b459 * ( 1999 ) 224 , hep - ph/9904379 . s. bertolini and f. vissani , phys . lett . * b324 * ( 1994 ) 164 , hep - ph/9311293 . . hall , v.a . kostelecky and s. raby , nucl . phys . * b267 * ( 1986 ) 415 . a. j. buras , a. romanino , l. silvestrini , nucl . phys . * b520 * ( 1998 ) 3 . f. gabbiani , e. gabrielli , a. masiero and l. silvestrini , nucl . phys . * b477 * ( 1996 ) 321 hep - ph/9604387 . branco , g.c . cho , y. kizukuri and n. oshimo , _ phys . _ * b 337 * , 316 ( 1994 ) ; + g.c . branco , g.c . cho , y. kizukuri , and n. oshimo , nucl . * b449 * ( 1995 ) 483 . f. borzumati , z. phys . * c63 * , 291 ( 1994 ) ; + s. bertolini , f. vissani , z. phys . * c67 * , 513 ( 1995 ) ; + t. goto , y.y . keum , t. nihei , y. okada and y. shimizu , phys . lett . * b460 * , ( 1999 ) 333 , hep - ph/9812369 . a. l. kagan and m. neubert , eur . j. * c7 * , 5 ( 1999 ) . m. ciuchini _ _ , _ jhep _ * 10 * ( 1998 ) 008 , hep - ph/9808328 ; + r. contino and i. scimemi , _ eur . phys . j. _ * c10 * , 347 ( 1999 ) , hep - ph/9809437 . t. affolder _ _ , fermilab report no . fermlab - pub-99/225-e , september 1999 , hep - ex/9909003 . m. brhlik , l. everett , g. l. kane , and j. lykken,_phys . lett . _ * 83 * , 2124 ( 1999 ) , hep - ph/9905215 .

there are indications both from observations and theoretical studies that stars in many astrophysical environments may have the relative abundances of elements synthesized by nuclear @xmath4-capture reactions ( e.g. c , o , ne , mg , si , ti , etc ) that are enhanced with respect to the solar ones . in brief , king ( 1994 ) found the mean [ o / fe ] ratio for a handful of halo stars to be @xmath8 dex . similar results are found for globular cluster stars ( e.g. [ @xmath9=0.18 $ ] dex for 47 tuc , carney 1996 ) and galactic bulge k giants ( mcwilliam & rich 1994 , rich & mcwilliam 2000 ) . according to models of chemical evolution of the galaxy , bulge stars should be @xmath4-enhanced to reproduce the metallicity distribution ( matteucci & brocato 1990 , matteucci et al . indications that the abundance ratios of elements in elliptical galaxies are not solar ( e.g. @xmath10 = 0.3 - 0.7 $ ] dex ) come from worthey et al . also synthetic metal line strengths ( weiss , peletier & matteucci 1995 ) from high - metallicity models show that solar - scaled models do not fit the observed mg2 ( 5180 ) and fe ( 5270 and 5335 ) line indices . given these premises , it is of paramount importance to include the enhancement of @xmath4-elements in the population synthesis models designed to describe those galaxies . in the past this has been done but using models that suffered from many limitations . firstly , till recently there were no extended sets of opacities for @xmath4-enhanced mixtures covering the complete relevant range of stellar temperatures . such opacity calculations are nowadays available . secondly , it has often been assumed that the behaviour of @xmath4-enhanced isochrones can be easily mimicked by suitably scaling the metallicity ( salaris et al . 1993 ) . however , the validity of such a re - scaling has only been confirmed for low metallicities , and it holds only for some particular cases like all @xmath4-elements being enhanced by the same factor . weiss et al . ( 1995 ) already pointed out the differences in the evolution due to @xmath4-enhancement at solar and super - solar metallicities . recent work by salaris & weiss ( 1998 ) and vandenberg et al . ( 2000 ) also indicates that the simple re - scaling of the content becomes risky at high metallicities . moreover , we should keep in mind that there is no _ a priori _ reason , from the point of view of both nucleosynthetic and chemical evolution theories , to expect that all @xmath4-elements are enhanced by the same factor . from observations of halo and bulge stars , the degree of enhancement of different @xmath4-elements is , in general , not constant with differences amounting to @xmath11 dex . therefore in this paper we intend to re - analyze the impact of the enhancement of @xmath4-elements on the properties of metal - rich stellar populations ( whose metallicity ranges from half - solar to three times - solar ) . the analysis is made for a particular choice of abundance ratios , for which self - consistent opacities are available . we start calculating extended sets of stellar evolutionary tracks ( sect . [ sec_tracks ] ) , with total metal contents @xmath12 , 0.019 , 0.040 and 0.070 both for solar - scaled ( same abundance ratios as in the solar mix ) and @xmath4-enhanced patterns of abundances . particular attention is paid to the description of the adopted initial element abundances and sources of opacity . based on those stellar models we derive large grids of isochrones and single stellar populations for which we present magnitudes , colours and mass - to - light ratios in many pass - bands of the johnson and wfpc2 photometric systems ( sect . [ sec_isochrones ] ) . extensive tabulations of stellar models , isochrones and photometric properties are provided in electronic format for the purpose of general use . finally , a short summary is given in sect . [ sec_conclusion ] . vandenberg et al . ( 2000 ) recently presented extensive sets of evolutionary tracks for mixtures , limited to the range of low - mass stars ( @xmath13 ) . a constant enhancement of all @xmath4-elements has been assumed ( also in their opacity tables ) . in the present work , instead , the choice of initial abundances has been guided by observational results . it is worth remarking that , within the present errors , none of both choices seems to be strongly preferred . we calculate four grids of stellar models with initial masses from 0.15 to 20 @xmath14 and initial chemical compositions @xmath0 $ ] , @xmath1 $ ] , @xmath2 $ ] , @xmath3 $ ] . the metallicities and the helium - to - metal enrichment ratio @xmath15 are chosen in such a way to secure consistency with the girardi et al . ( 2000 ) models . the helium - to - metal enrichment law is @xmath16 . for the above values of @xmath17 $ ] , we compute both solar - scaled and @xmath4-enhanced sets of stellar tracks . the evolutionary phases covered by the grids extend from the zero age main sequence ( zams ) up to either the start of the thermally pulsing asymptotic giant branch ( tp - agb ) phase or carbon ignition . the input physics of the present stellar models is the same as in girardi et al . ( 2000 ) , apart from differences in ( i ) the opacities , and ( ii ) the rates of energy loss by plasma neutrinos . a short description of the updated ingredients and the way we deal with the enhancement of @xmath4-elements is presented in the sections below . for a fixed total metallicity @xmath18 , the solar - scaled abundance ratios of metals are taken from grevesse & noels ( 1993 ) . in the @xmath4-enhanced case , we adopt a mixture in which only the elements resulting from nuclear @xmath4-capture reactions ( o , ne , mg , si , s , ca , ti , ni ) are enhanced with respect to the solar abundance ratios . this is the same mixture as used by salaris et al . ( 1997 ) and salaris & weiss ( 1998 ) to study the age of the oldest metal - poor and disk metal - rich globular clusters . for the sake of clarity , table [ elements ] lists the abundances for the solar - scaled and @xmath4-enhanced mixtures . columns ( 2 ) and ( 4 ) show the abundance @xmath19 of the elements in logarithmic scale , @xmath20 , where @xmath21 is the abundance by number . columns ( 3 ) and ( 5 ) display the abundance by mass fraction @xmath22 , relative to the metallicity @xmath18 . columns ( 6 ) and ( 7 ) list the enhancement ratio ( relative either to fe or the total metal fraction ) in spectroscopic notation . the enhancement factors come from the determinations of chemical abundances in metal - poor field stars by ryan et al . ( 1991 , cf . also salaris & weiss 1998 ) . to derive the initial abundances of different isotopes we use the isotopic ratios ( by mass ) of anders & grevesse ( 1989 ) . c|cc|cccc element & & + & @xmath19 & @xmath23 & @xmath19 & @xmath23 & @xmath24 $ ] & @xmath25 $ ] + ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) & ( 5 ) & ( 6 ) & ( 7 ) + o & 8.870 & 0.482273 & 9.370 & 0.672836 & 0.50 & @xmath260.1442 + ne & 8.080 & 0.098668 & 8.370 & 0.084869 & 0.29 & @xmath270.0658 + mg & 7.580 & 0.037573 & 7.980 & 0.041639 & 0.40 & @xmath260.0441 + si & 7.550 & 0.040520 & 7.850 & 0.035669 & 0.30 & @xmath270.0558 + s & 7.210 & 0.021142 & 7.540 & 0.019942 & 0.33 & @xmath270.0258 + ca & 6.360 & 0.003734 & 6.860 & 0.005209 & 0.50 & @xmath260.1441 + ti & 5.020 & 0.000211 & 5.650 & 0.000387 & 0.63 & @xmath270.1875 + ni & 6.250 & 0.004459 & 6.270 & 0.002056 & 0.02 & @xmath270.3371 + c & 8.550 & 0.173285 & 8.550 & 0.076451 & 0.00 & @xmath270.3553 + n & 7.970 & 0.053152 & 7.970 & 0.023450 & 0.00 & @xmath270.3553 + na & 6.330 & 0.001999 & 6.330 & 0.000882 & 0.00 & @xmath270.3553 + cr & 5.670 & 0.001005 & 5.670 & 0.000443 & 0.00 & @xmath270.3557 + fe & 7.500 & 0.071794 & 7.500 & 0.031675 & 0.00 & @xmath270.3557 + the opacities we have adopted are the same as in salaris , deglinnocenti & weiss ( 1997 ) . they are available at : ` http://www-phys.llnl.gov/v_div/opal/existing.html ` . details are given below . for high temperatures ( @xmath28 ) we use the radiative opal opacities by rogers & iglesias ( 1995 ) in the case of solar - scaled mixtures , and by iglesias & rogers ( 1996 ) for the ones . at low temperatures these are combined with the molecular opacities provided by alexander & ferguson ( 1994 ) for the solar mixtures , and by alexander ( priv . communication ) for the ones . in both cases , low- and high - temperature opacity grids are merged by interpolation in the temperature range from @xmath29 to @xmath30 . for very high temperatures , @xmath31 , the weiss et al . ( 1990 ) opacities are used . these rely on the los alamos opacity database . the radiative opacities for c , o mixtures , needed in the post - main sequence phases , are from iglesias & rogers ( 1993 ) . finally , conductive opacities of electron - degenerate matter are from itoh et al . ( 1983 ) . it is very important to emphasize that the present models are based on _ complete and consistent opacities _ , i.e. the opacities ( both for solar and mixtures ) are generated for the same pattern of abundances adopted in the stellar models , both at low and high temperatures . the importance of this issue is illustrated by fig . [ compare_opacity_hr ] , which shows the effect in the hertzsprung - russell ( hr ) diagram of adopting different ( inconsistent ) opacities and/or mixtures and/or enhancement factors as input to a 1 @xmath14 main sequence model with @xmath32 $ ] . some of the illustrated cases are often found in literature . the tracks are evolved up to the solar age of 4.6 gyr . the meaning of the labels is the following : the first label refers to high temperature opacity ( @xmath33 ) and can vary from opal - s , i.e. solar - scaled mixture , to opal-@xmath4 , i.e. enhanced according to salaris & weiss ( 1998 ; w95 ) or rogers ( priv . communication ; r95 ) . this latter case is a special @xmath4-enhanced mixture characterized , if compared with salaris & weiss ( 1998 ) , by more c , less o , more ne , more na ( which is not an alpha - element ) , less al , more s , and more fe . the second label refers to the molecular opacity , with the same notation as for solar and @xmath4-enhanced mixtures . af stands for alexander & ferguson ( 1994 ) . the prescription enclosed in a box represents the one adopted by us . another point to emphasize is the difference between the present opacities and those used by girardi et al . two are the main sources of differences . first the adoption of the itoh et al . ( 1983 ) conductive opacities instead of the older ones by hubbard & lampe ( 1969 ) . interestingly , catelan et al . ( 1996 ) show that itoh et al . ( 1983 ) conductive opacities should not be used in rgb models , because they are valid only for a liquid phase , and not for the physical conditions existing in the interiors of rgb stars . however , we verified that the effect of using the itoh et al . ( 1983 ) opacities instead of the hubbard & lampe ( 1969 ) is to increase the core mass at the helium flash by 0.006 ( see also castellani et al . 2000 and cassisi et al . 1998 ) , which is a negligible effect . the second difference lies in the numerical technique used to interpolate within the grids of the opacity tables . in this paper we use the two - dimensional bi - rational cubic damped - spline algorithm ( see & weiss 1998 ; and spth 1973 ) , whereas girardi et al . ( 2000 ) adopted the smooth bi - parabolic spline interpolation ( ssi ) already introduced in our code by bressan et al . we verified that the differences amount to about a few percent . in general , evolutionary models of low mass differ very little by changing the opacity interpolation scheme . however , massive stars ( say @xmath34 ) are much more sensitive to the interpolation algorithm . this is shown in fig . [ compare_tracks_6msun ] , where we plot two 4 and 6 models with @xmath35 $ ] calculated both with the ssi ( continuous lines ) and the damped - spline ( dashed lines ) scheme . figure [ compare_opacity_inner ] compares the effects of the different interpolations on the opacity at the physical conditions holding in the inner regions of a 6 star when the central hydrogen mass fraction is @xmath36 , i.e. at the stage marked with a dot in fig . [ compare_tracks_6msun ] . at this stage the temperature at the schwarzschild border of the convective core is @xmath37 and the density is @xmath38 ( corresponding to the arrow in the fig . [ compare_opacity_inner ] ) . for the temperatures indicated along the curves , these represent the opacity as a function of the quantity @xmath39 ( see rogers & iglesias 1992 for more details ) . the continuous curve is based on the ssi interpolation , whereas the dashed curve refers to the method adopted in the present study . the small difference in opacity at the core border is typical of the values found throughout the evolutionary sequences , and is responsible for the different hr patterns . we notice that both interpolations are acceptable , because the percentage error associated to both of them is roughly 3 % , which is lower than the opacity uncertainty in the original opal tables ( the latter can be as high as 10 % at individual @xmath40 points ) . this shows that even the fine opal grid for some applications is not sufficiently dense and more grid points might be needed . this is also illustrated by the solar model by schlattl & weiss ( 1998 ) , where the inverse situation , i.e. spline producing a nicer opacity run , is found . the reason for the different luminosities in the high - mass models , is that in these stars the radiative gradient across the border of the convective core is shallow . therefore , even slightly larger opacities at the schwarzschild border may cause an increase of the convective core and lead to higher luminosities . this trend already begins during the core h - burning phase and remains during the subsequent evolution . the effects are detectable starting from stars with mass larger than @xmath41 ( see fig . [ compare_tracks_6msun ] ) and becomes dramatic over 60 . it is worth recalling that the maximum mass of our stellar models is 20 . energy losses by neutrinos are from haft et al . compared with the previous ones of munakata et al . ( 1985 ) used by girardi et al . ( 2000 ) , neutrino cooling during the rgb is more efficient . this causes an increase of 0.005 in the core mass at helium ignition . we have checked this on a 1 model with @xmath42 . differences of the same order have been found by other authors ( e.g. haft et al . 1994 ; cassisi et al . 1998 ) . the remaining physical input is the same as in girardi et al . ( 2000 ) , to whom we refer for further details . in summary , we follow the evolution of h , @xmath43he , @xmath44he , @xmath45c , @xmath46c , @xmath47n , @xmath48n , @xmath49o , @xmath50o , @xmath51o , @xmath52ne , @xmath53ne , mg , according to the nuclear reaction rates of caughlan & fowler ( 1988 ) and landr et al . for mg we refer to the total content of @xmath54 mg , @xmath55 mg and @xmath56 mg . the equation of state ( eos ) is that of a fully - ionized gas for temperatures higher that @xmath57 k. for lower temperatures we adopt the mihalas et al . ( 1990 and references therein ) eos . calibration of the solar model fixes the value of the mixing length parameter @xmath4 of the bhm - vitense ( 1958 ) theory at 1.68 . overshooting both from the convective core and the convective envelope are taken into account according to the ballistic algorithm by bressan et al . table [ overshooting ] summarizes the values of the overshooting parameters for the core , @xmath58 , and envelope , @xmath59 we have adopted as a function of the initial mass . as far as the mass - loss is concerned , this is taken into account and its effect on the stellar structure is calculated for masses higher than 6 . we used the same prescription adopted by fagotto et al . ( 1994 ) , where more details can be found . c|c|c @xmath60 & @xmath58 & @xmath59 + 0.5 - 1.0 & 0 & 0.25 + 1.1 - 1.4 & @xmath61 & 0.25 + 1.5 - 2.0 & 0.5 & 0.25 + 2.0 - 2.5 & 0.5 & @xmath62 + 2.5 - 20 & 0.5 & 0.7 + for the sake of illustration , in figs . [ hrtracksz008 ] , [ hrtracksz008_low ] and [ hrtracksz008_low_hb ] we present the complete set of evolutionary tracks for @xmath0 $ ] and @xmath4-enhanced mixtures . [ hrtracksz008 ] shows the models of intermediate - mass stars from the zero age main sequence ( zams ) up to the tp - agb phase and those of massive stars from the zams up to carbon ignition . [ hrtracksz008_low ] presents the low - mass tracks from the zams up to the rgb - tip . finally , fig . [ hrtracksz008_low_hb ] shows the corresponding he - burning phase from the zero age horizontal branch ( zahb ) up to the tp - agb phase . some relevant physical quantities are given in table [ timez008 ] , which lists , as a function of the initial mass , the lifetimes in years of the central h - burning and he - burning phases , @xmath63 and @xmath64 respectively , the core mass at the helium ignition , @xmath65 , and the core mass at the first thermal pulse on the agb phase , @xmath66 . for intermediate - mass and massive stars , @xmath65 refers to the core mass at the he - ignition . for the same chemical composition table [ dredgeup ] reports the changes in the surface chemical abundances of those elements that suffered from nuclear processing via the pp - chain and cno - cycle in the deep interiors , after the completion of the first and the second dredge - up by envelope mixing . the surface abundances of heavier elements are not altered by nuclear processing . the last two columns give the ratios of @xmath47n and @xmath49o with respect to their initial values . these ratios are indicative of the efficiency of the dredge - up episodes . table [ critical_masses ] displays the transition masses @xmath67 , @xmath68 and @xmath69 for models calculated with solar and initial composition . in brief , @xmath67 is the maximum mass below which stars become fully convective during the central h - burning phase , @xmath68 is the maximum mass for the core he - flash to occur , and finally @xmath69 is the maximum mass for the central c - burning to occur in highly electron degenerate gas ( deflagration or detonation followed by disruption of the star ) . c|cccc|cccc & & + @xmath60 & @xmath70yr & @xmath71yr & @xmath65 & @xmath72 & @xmath70yr & @xmath71yr & @xmath65 & @xmath72 + 0.6 & 5.45 @xmath73 & 1.05 @xmath74 & 0.4814 & 0.5108 & 4.48 @xmath73 & 1.08 @xmath75 & 0.4445 & 0.5398 + 0.7 & 3.14 @xmath73 & 1.04 @xmath74 & 0.4806 & 0.5159 & 2.45 @xmath73 & 1.05 @xmath75 & 0.4434 & 0.5614 + 0.8 & 1.87 @xmath73 & 1.03 @xmath74 & 0.4793 & 0.5192 & 1.42 @xmath73 & 1.02 @xmath75 & 0.4434 & 0.5756 + 0.9 & 1.18 @xmath73 & 1.02 @xmath74 & 0.4781 & 0.5258 & 8.28 @xmath76 & 1.01 @xmath75 & 0.4430 & 0.5926 + 1.0 & 7.64 @xmath77 & 1.02 @xmath74 & 0.4782 & 0.5272 & 5.11 @xmath76 & 1.01 @xmath75 & 0.4426 & 0.5769 + 1.1 & 5.26 @xmath77 & 1.01 @xmath74 & 0.4776 & 0.5317 & 3.64 @xmath76 & 9.92 @xmath57 & 0.4424 & 0.6049 + 1.2 & 3.90 @xmath77 & 1.00 @xmath74 & 0.4778 & 0.5312 & 2.89 @xmath76 & 9.87 @xmath57 & 0.4418 & 0.6079 + 1.3 & 3.19 @xmath77 & 9.83 @xmath78 & 0.4778 & 0.5291 & 2.39 @xmath76 & 9.71 @xmath57 & 0.4421 & 0.6049 + 1.4 & 2.68 @xmath77 & 9.79 @xmath78 & 0.4763 & 0.5314 & 2.01 @xmath76 & 9.89 @xmath57 & 0.4375 & 0.6064 + 1.5 & 2.28 @xmath77 & 1.07 @xmath74 & 0.4642 & 0.5273 & 1.71 @xmath76 & 1.11 @xmath75 & 0.4169 & 0.6022 + 1.6 & 1.87 @xmath77 & 1.14 @xmath74 & 0.4539 & 0.5245 & 1.40 @xmath76 & 1.23 @xmath75 & 0.3999 & 0.5947 + 1.7 & 1.57 @xmath77 & 1.26 @xmath74 & 0.4391 & 0.5194 & 1.16 @xmath76 & 1.36 @xmath75 & 0.3845 & 0.5875 + 1.8 & 1.33 @xmath77 & 1.44 @xmath74 & 0.4192 & 0.5065 & 9.76 @xmath75 & 1.46 @xmath75 & 0.3680 & 0.5913 + 1.9 & 1.15 @xmath77 & 3.10 @xmath74 & 0.3337 & 0.4893 & 8.28 @xmath75 & 1.70 @xmath75 & 0.3406 & 0.5849 + 2.0 & 9.90 @xmath74 & 2.71 @xmath74 & 0.3451 & 0.4909 & 7.11 @xmath75 & 2.15 @xmath75 & 0.3281 & 0.5730 + 2.2 & 7.65 @xmath74 & 2.10 @xmath74 & 0.3757 & 0.5104 & 5.33 @xmath75 & 1.71 @xmath75 & 0.3598 & 0.6114 + 2.5 & 5.44 @xmath74 & 1.44 @xmath74 & 0.4246 & 0.5452 & 3.66 @xmath75 & 1.21 @xmath75 & 0.4104 & 0.6117 + 3.0 & 3.39 @xmath74 & 7.36 @xmath78 & 0.5367 & 0.6483 & 2.16 @xmath75 & 6.60 @xmath57 & 0.5178 & 0.6844 + 3.5 & 2.31 @xmath74 & 3.98 @xmath78 & 0.6718 & 0.7806 & 1.40 @xmath75 & 3.77 @xmath57 & 0.6408 & 0.6844 + 4.0 & 1.67 @xmath74 & 2.48 @xmath78 & 0.8213 & 0.9375 & 9.71 @xmath57 & 2.34 @xmath57 & 0.7713 & 0.6844 + 4.5 & 1.27 @xmath74 & 1.67 @xmath78 & 0.9744 & 1.0910 & 7.12 @xmath57 & 1.56 @xmath57 & 0.9092 & 0.6844 + 5.0 & 1.00 @xmath74 & 1.21 @xmath78 & 1.1333 & 1.2497 & 5.45 @xmath57 & 1.10 @xmath57 & 1.0593 & 1.2062 + 6.0 & 6.72 @xmath78 & 6.73 @xmath79 & 1.4456 & - & 3.53 @xmath57 & 6.06 @xmath80 & 1.3904 & 1.5452 + 7.0 & 4.87 @xmath78 & 4.55 @xmath79 & 1.7673 & - & 2.51 @xmath57 & 3.89 @xmath80 & 1.7609 & - + 8.0 & 3.76 @xmath78 & 3.30 @xmath79 & 2.1265 & - & 1.91 @xmath57 & 2.74 @xmath80 & 2.1599 & - + 9.0 & 3.03 @xmath78 & 2.52 @xmath79 & 2.4985 & - & 1.53 @xmath57 & 2.06 @xmath80 & 2.5988 & - + 10.0 & 2.51 @xmath78 & 2.02 @xmath79 & 2.9138 & - & 1.28 @xmath57 & 1.62 @xmath80 & 3.0935 & - + 12.0 & 1.86 @xmath78 & 1.43 @xmath79 & 3.7804 & - & 9.59 @xmath80 & 1.14 @xmath80 & 3.9767 & - + 15.0 & 1.34 @xmath78 & 1.03 @xmath79 & 4.6706 & - & 7.10 @xmath80 & 8.26 @xmath81 & 5.3587 & - + 20.0 & 9.39 @xmath79 & 7.12 @xmath82 & 6.7187 & - & 5.17 @xmath80 & 6.01 @xmath81 & 7.7582 & - + @xmath68 is here defined as the initial mass for which the core mass at he ignition has its minimum value ( see table [ timez008 ] and fig . [ mflash ] ) . [ mflash ] shows that similar values for @xmath68 are found both in the and in the solar - scaled case . compared with girardi et al . ( 2000 ) , we find higher @xmath69 values . this is once more caused by the different interpolation schemes of the opacity tables . the one currently in use yields smaller convective cores for the same initial mass . it follows that with the present overshooting prescription the maximum mass for a star to develop an electron - degenerate core after the core he - burning phase is slightly higher than previously estimated . ccccccccccccc @xmath83 & h & @xmath43he & @xmath44he & @xmath45c & @xmath46c & @xmath47n & @xmath48n & @xmath49o & @xmath50o & @xmath51o & @xmath84 & @xmath85 + + all & 0.742 & 3.80@xmath86 & 0.250 & 1.37@xmath87 & 1.65@xmath86 & 4.24@xmath88 & 1.67@xmath89 & 3.85@xmath87 & 1.56@xmath89 & 8.68@xmath89 & 1.00 & 1.00 + + 0.60 & 0.730 & 3.95@xmath87 & 0.258 & 1.37@xmath87 & 1.72@xmath86 & 4.24@xmath88 & 1.66@xmath89 & 3.85@xmath87 & 1.56@xmath89 & 8.68@xmath89 & 1.00 & 1.00 + 0.70 & 0.727 & 2.73@xmath87 & 0.263 & 1.36@xmath87 & 2.50@xmath86 & 4.26@xmath88 & 1.55@xmath89 & 3.85@xmath87 & 1.56@xmath89 & 8.68@xmath89 & 0.993 & 1.01 + 0.80 & 0.725 & 1.97@xmath87 & 0.265 & 1.32@xmath87 & 3.71@xmath86 & 4.59@xmath88 & 1.44@xmath89 & 3.85@xmath87 & 1.56@xmath89 & 8.64@xmath89 & 0.964 & 1.08 + 0.90 & 0.724 & 1.55@xmath87 & 0.267 & 1.27@xmath87 & 4.02@xmath86 & 5.18@xmath88 & 1.36@xmath89 & 3.85@xmath87 & 1.57@xmath89 & 8.52@xmath89 & 0.925 & 1.22 + 1.00 & 0.723 & 1.20@xmath87 & 0.267 & 1.22@xmath87 & 4.13@xmath86 & 5.72@xmath88 & 1.30@xmath89 & 3.85@xmath87 & 1.60@xmath89 & 8.30@xmath89 & 0.891 & 1.35 + 1.10 & 0.723 & 1.01@xmath87 & 0.268 & 1.17@xmath87 & 4.24@xmath86 & 6.35@xmath88 & 1.22@xmath89 & 3.85@xmath87 & 1.70@xmath89 & 8.04@xmath89 & 0.851 & 1.50 + 1.20 & 0.725 & 8.67@xmath88 & 0.267 & 1.14@xmath87 & 4.38@xmath86 & 6.70@xmath88 & 1.18@xmath89 & 3.85@xmath87 & 1.83@xmath89 & 7.89@xmath89 & 0.829 & 1.58 + 1.30 & 0.726 & 8.00@xmath88 & 0.265 & 1.11@xmath87 & 4.18@xmath86 & 7.00@xmath88 & 1.15@xmath89 & 3.85@xmath87 & 2.01@xmath89 & 7.72@xmath89 & 0.811 & 1.65 + 1.40 & 0.724 & 6.89@xmath88 & 0.268 & 1.02@xmath87 & 4.59@xmath86 & 8.06@xmath88 & 1.02@xmath89 & 3.85@xmath87 & 2.56@xmath89 & 7.28@xmath89 & 0.742 & 1.90 + 1.50 & 0.723 & 6.11@xmath88 & 0.268 & 9.81@xmath88 & 4.49@xmath86 & 8.53@xmath88 & 9.72@xmath90 & 3.84@xmath87 & 1.02@xmath86 & 7.02@xmath89 & 0.716 & 2.01 + 1.60 & 0.724 & 5.35@xmath88 & 0.268 & 9.56@xmath88 & 4.47@xmath86 & 9.02@xmath88 & 9.44@xmath90 & 3.82@xmath87 & 1.02@xmath86 & 6.86@xmath89 & 0.697 & 2.13 + 1.70 & 0.722 & 4.76@xmath88 & 0.269 & 9.34@xmath88 & 4.50@xmath86 & 9.66@xmath88 & 9.15@xmath90 & 3.77@xmath87 & 1.70@xmath86 & 6.76@xmath89 & 0.681 & 2.28 + 1.80 & 0.720 & 4.21@xmath88 & 0.272 & 9.04@xmath88 & 4.39@xmath86 & 1.04@xmath87 & 8.88@xmath90 & 3.72@xmath87 & 1.93@xmath86 & 6.60@xmath89 & 0.660 & 2.46 + 1.90 & 0.718 & 3.78@xmath88 & 0.274 & 9.01@xmath88 & 4.45@xmath86 & 1.09@xmath87 & 8.76@xmath90 & 3.67@xmath87 & 1.63@xmath86 & 6.55@xmath89 & 0.658 & 2.57 + 2.00 & 0.718 & 3.38@xmath88 & 0.274 & 8.95@xmath88 & 4.41@xmath86 & 1.12@xmath87 & 8.72@xmath90 & 3.64@xmath87 & 1.63@xmath86 & 6.52@xmath89 & 0.653 & 2.65 + 2.20 & 0.714 & 2.76@xmath88 & 0.278 & 8.88@xmath88 & 4.51@xmath86 & 1.20@xmath87 & 8.57@xmath90 & 3.57@xmath87 & 1.53@xmath86 & 6.45@xmath89 & 0.648 & 2.82 + 2.50 & 0.711 & 2.13@xmath88 & 0.280 & 8.72@xmath88 & 4.53@xmath86 & 1.27@xmath87 & 8.37@xmath90 & 3.51@xmath87 & 1.16@xmath86 & 6.36@xmath89 & 0.636 & 3.00 + 3.00 & 0.713 & 1.50@xmath88 & 0.279 & 8.65@xmath88 & 4.47@xmath86 & 1.30@xmath87 & 8.27@xmath90 & 3.48@xmath87 & 9.83@xmath89 & 6.30@xmath89 & 0.631 & 3.08 + 3.50 & 0.717 & 1.13@xmath88 & 0.275 & 8.78@xmath88 & 4.66@xmath86 & 1.27@xmath87 & 8.23@xmath90 & 3.50@xmath87 & 6.74@xmath89 & 6.42@xmath89 & 0.641 & 2.99 + 4.00 & 0.718 & 9.02@xmath86 & 0.274 & 8.80@xmath88 & 4.83@xmath86 & 1.27@xmath87 & 8.10@xmath90 & 3.50@xmath87 & 5.56@xmath89 & 6.43@xmath89 & 0.642 & 2.99 + 4.50 & 0.720 & 7.69@xmath86 & 0.272 & 9.13@xmath88 & 4.77@xmath86 & 1.22@xmath87 & 8.49@xmath90 & 3.51@xmath87 & 3.89@xmath89 & 6.64@xmath89 & 0.666 & 2.89 + 5.00 & 0.721 & 6.49@xmath86 & 0.270 & 9.49@xmath88 & 4.64@xmath86 & 1.18@xmath87 & 8.90@xmath90 & 3.51@xmath87 & 3.70@xmath89 & 6.73@xmath89 & 0.693 & 2.79 + 6.00 & 0.722 & 4.75@xmath86 & 0.270 & 8.75@xmath88 & 4.96@xmath86 & 1.28@xmath87 & 7.79@xmath90 & 3.50@xmath87 & 4.04@xmath89 & 6.40@xmath89 & 0.638 & 3.01 + 7.00 & 0.723 & 3.85@xmath86 & 0.269 & 8.55@xmath88 & 5.00@xmath86 & 1.32@xmath87 & 7.51@xmath90 & 3.47@xmath87 & 3.84@xmath89 & 6.28@xmath89 & 0.624 & 3.12 + 8.00 & 0.721 & 3.26@xmath86 & 0.271 & 8.41@xmath88 & 4.98@xmath86 & 1.37@xmath87 & 7.31@xmath90 & 3.43@xmath87 & 3.61@xmath89 & 6.15@xmath89 & 0.614 & 3.24 + 9.00 & 0.722 & 3.06@xmath86 & 0.270 & 8.83@xmath88 & 4.99@xmath86 & 1.31@xmath87 & 7.83@xmath90 & 3.45@xmath87 & 2.77@xmath89 & 6.42@xmath89 & 0.644 & 3.08 + 10.00 & 0.721 & 2.78@xmath86 & 0.271 & 8.79@xmath88 & 4.92@xmath86 & 1.33@xmath87 & 7.79@xmath90 & 3.43@xmath87 & 2.83@xmath89 & 6.37@xmath89 & 0.641 & 3.13 + 12.00 & 0.713 & 2.36@xmath86 & 0.279 & 8.62@xmath88 & 5.03@xmath86 & 1.44@xmath87 & 7.52@xmath90 & 3.33@xmath87 & 2.49@xmath89 & 6.21@xmath89 & 0.629 & 3.40 + 15.00 & 0.687 & 1.90@xmath86 & 0.305 & 8.23@xmath88 & 4.94@xmath86 & 1.67@xmath87 & 7.15@xmath90 & 3.12@xmath87 & 2.24@xmath89 & 5.81@xmath89 & 0.601 & 3.94 + + 4.00 & 0.694 & 8.64@xmath86 & 0.298 & 8.46@xmath88 & 4.71@xmath86 & 1.42@xmath87 & 7.84@xmath90 & 3.37@xmath87 & 5.91@xmath89 & 6.17@xmath89 & 0.618 & 3.35 + 4.50 & 0.676 & 6.95@xmath86 & 0.316 & 8.35@xmath88 & 4.65@xmath86 & 1.51@xmath87 & 7.72@xmath90 & 3.29@xmath87 & 4.59@xmath89 & 6.07@xmath89 & 0.609 & 3.56 + 5.00 & 0.662 & 5.71@xmath86 & 0.330 & 8.48@xmath88 & 4.51@xmath86 & 1.56@xmath87 & 7.90@xmath90 & 3.22@xmath87 & 4.24@xmath89 & 6.01@xmath89 & 0.619 & 3.68 + 6.00 & 0.667 & 4.35@xmath86 & 0.325 & 8.08@xmath88 & 4.76@xmath86 & 1.60@xmath87 & 7.29@xmath90 & 3.22@xmath87 & 3.82@xmath89 & 5.86@xmath89 & 0.590 & 3.77 + 7.00 & 0.714 & 3.75@xmath86 & 0.278 & 8.35@xmath88 & 4.89@xmath86 & 1.41@xmath87 & 7.38@xmath90 & 3.40@xmath87 & 3.89@xmath89 & 6.11@xmath89 & 0.609 & 3.32 + c|c|c|c|c|c @xmath18 & @xmath91 & mixture & @xmath67 & @xmath68 & @xmath69 + 0.008 & 0.250 & solar - scaled & 0.36 & 1.90 & @xmath92 + 0.019 & 0.273 & solar - scaled & 0.37 & 2.10 & @xmath93 + 0.040 & 0.320 & solar - scaled & 0.36 & 2.20 & @xmath94 + 0.070 & 0.390 & solar - scaled & 0.33 & 1.90 & @xmath94 + 0.008 & 0.250 & & 0.37 & 1.90 & @xmath92 + 0.019 & 0.273 & & 0.37 & 2.10 & @xmath92 + 0.040 & 0.320 & & 0.36 & 2.10 & @xmath94 + 0.070 & 0.390 & & 0.33 & 2.00 & @xmath93 + the four sets of stellar models are used to calculate isochrones , integrated magnitudes and colours of single stellar populations with ages from @xmath78 to @xmath95 yr both in the johnson - cousins @xmath6 and hst - wfpc2 photometric systems . the age range is large enough to describe young clusters and associations as well as old globular clusters . the dense grid of stellar tracks allows us to construct detailed isochrones at small age steps ( @xmath96 ) . this also makes the data - base suitable for the simulation of synthetic colour - magnitude diagrams ( e.g. girardi 1999 ) . before theoretical isochrones are constructed , the tp - agb phase is included in all tracks of @xmath97 . therefore , this phase is present in the isochrones older than about @xmath75 yr . the reader is referred to girardi & bertelli ( 1998 ) and girardi et al . ( 2000 ) for all details about the synthetic algorithm used to follow the tp - agb evolution . the hr diagram of fig . [ compare_isoc ] shows a comparison between isochrones with chemical composition @xmath98 $ ] both for solar - scaled ( continuous lines ) and ( dotted lines ) stellar models . for the sake of clarity we plot only the isochrones with ages between @xmath99 and @xmath100 , in steps of @xmath101 . as already noted by salaris & weiss ( 1998 ) , for solar and super - solar total metallicity , the element abundance ratios among metals affect the evolution . from the hr diagram of fig . [ compare_isoc ] , two main features can be singled out . \(i ) @xmath4-enhanced isochrones have fainter and hotter turn - offs ( to ) . this is shown by the entries of table [ turnoff ] , which lists the turn - off luminosity and effective temperature of isochrones with @xmath102 but different mixtures of @xmath4-elements . the reason is that at given central hydrogen content the stellar model with solar - scaled composition has a mean opacity higher than the one . the point is illustrated in fig . [ compare_inner ] where we compare the opacity profile across the 1 models with @xmath103 $ ] and both ( dashed line ) and solar - scaled ( solid line ) mixtures . higher opacities induce steeper radiative temperature gradients @xmath104 and lower surface temperatures @xmath105 for the same central conditions . furthermore , steeper @xmath104s imply smaller burning regions and lower luminosities in turn . this effect overwhelms the one induced by the slightly larger convective core due to the higher @xmath104 ( when a convective core is present ) . combining all those effects together , at the same evolutionary stage , the solar - scaled model is fainter , cooler and older than the one . the same trend is recovered in the isochrones as well . the implications of the different turn - off temperatures and luminosities on the ages of globular clusters have already been extensively investigated ( see e.g. salaris & weiss 1998 ) . for isochrones younger than @xmath76 yr , the differences in the turn - off properties are much less marked . this is due both to ( 1 ) the weak metallicity dependence of the opacities in massive stars ( where electron - scattering becomes the main source of opacity ) and ( 2 ) the larger extension of central convection caused by the flatter profile of @xmath104 in stellar interiors . \(ii ) @xmath4-enhanced isochrones are hotter than the solar - scaled ones throughout all evolutionary phases . the effect increases with the metallicity . it is mainly due to the higher opacities of the solar - scaled mixtures . looking at the 10 gyr isochrone with chemical composition @xmath103 $ ] as an example , the temperature difference @xmath106 is about @xmath270.015 at the turn - off and increases to @xmath270.020 at the bottom of the rgb . these differences in effective temperatures result into colour differences @xmath107 of 0.05 mag and 0.08 mag , at the turn - off and bottom of the rgb , respectively . c|c|c|c|c & & + age & @xmath108 & @xmath109 & @xmath108 & @xmath109 + 7.00 & 4.491 & 4.424 & 4.473 & 4.428 + 7.20 & 4.109 & 4.367 & 4.084 & 4.370 + 7.40 & 3.749 & 4.309 & 3.722 & 4.312 + 7.60 & 3.418 & 4.253 & 3.380 & 4.255 + 7.80 & 3.095 & 4.197 & 3.046 & 4.198 + 8.00 & 2.762 & 4.143 & 2.728 & 4.143 + 8.20 & 2.390 & 4.087 & 2.367 & 4.088 + 8.40 & 2.090 & 4.030 & 2.028 & 4.031 + 8.60 & 1.791 & 3.976 & 1.733 & 3.978 + 8.80 & 1.492 & 3.924 & 1.436 & 3.926 + 9.00 & 1.207 & 3.874 & 1.154 & 3.877 + 9.20 & 0.853 & 3.834 & 0.839 & 3.837 + 9.40 & 0.587 & 3.808 & 0.528 & 3.812 + 9.60 & 0.383 & 3.787 & 0.377 & 3.795 + 9.80 & 0.379 & 3.776 & 0.406 & 3.785 + 10.00 & 0.240 & 3.759 & 0.232 & 3.770 + 10.20 & 0.083 & 3.740 & 0.082 & 3.753 + it is worth remarking that in weiss et al . ( 1995 ) , the isochrones are found to have hotter and brighter turn - offs , and cooler rgbs ( at least in large portions of it ) than in the solar - scaled case ( see their fig . 4 ; notice that the line - types there are wrong : solid lines refer to the case ) . the cooler rgbs are no longer found here or in vandenberg et al . ( 2000 ) and have been a consequence of the lack of low - temperature opacity tables for compositions in opacities in weiss et al . this confirms how important is to use self - consistent opacities . finally , one can ask whether an isochrone may be approximated by a solar - scaled one with a slightly lower metallicity . in order to answer this question let us examine a set of 10 gyr old isochrones . in fig . [ fig_isoc_10gyr ] we plot all the isochrones with this age , and with metallicity going from @xmath12 to @xmath110 for both the ( continuous lines ) and solar - scaled ( short dashed lines ) mixtures . one can see that both sets of isochrones have different shapes , the differences becoming larger at higher metallicities . more specifically , one can notice that the difference in effective temperatures at the turn - off point between an isochrone and a solar - scaled one , @xmath111 , increases with the metallicity , passing from @xmath112 to @xmath113 as @xmath18 increases from 0.008 to 0.07 . on the contrary , the temperature difference at the base of the rgb , @xmath114 , is almost constant and of about @xmath115 . for the same age we calculate , by interpolating among the grids of different metallicity , several solar - scaled isochrones with @xmath18 spanning from @xmath116 to @xmath117 , and selected among them the one best fitting the main sequence of the @xmath102 isochrone . this turns out to be the interpolated isochrone with @xmath118 , which is shown by the long - dashed line in fig . [ fig_isoc_10gyr ] . clearly , this solar - scaled @xmath118 isochrone does not reproduce well the rgb sequence of the @xmath102 one . more specifically , although @xmath111 is very small for this isochrone pair , @xmath114 is not neglegible ( @xmath119 ) . from this simple test we then conclude that for a fixed age and relatively high metallicities , an isochrone can not be reproduced by tuning the metallicity of a solar - scaled one . the above temperature differences for rgb models ( isochrones ) bear very much on integrated spectral line indices of old metal - rich populations , which are commonly used to infer ages and metallicities of elliptical galaxies . this will be the subject of a forthcoming paper . theoretical luminosities and effective temperatures are translated to magnitudes and colours by means of extensive tabulations of bolometric corrections ( bc ) and colours obtained from properly convolving spectral energy distributions ( seds ) as a function of @xmath105 , @xmath120 and [ fe / h ] . the procedure is amply described in bertelli et al . ( 1994 ) , bressan et al . ( 1994 ) , and tantalo et al . ( 1996 ) to whom the reader should refer for more details . suffice it to recall here the following basic steps : \(i ) the main body of the spectral library is from kurucz ( 1992 ) , however extended to the high and low temperature ranges . for stars with @xmath121 k pure black - body spectra are assigned , whereas for stars with @xmath122 k the catalogue of stellar fluxes by fluks et al . ( 1994 ) is adopted . this catalogue includes 97 observed spectra for all m - spectral subtypes in the wavelength range @xmath123 , and synthetic photospheric spectra in the range @xmath124 . the scale of in fluks et al . ( 1994 ) is similar to that of ridgway et al . ( 1980 ) for spectral types earlier than m4 but deviates from it for later spectral types . since ridgway s et al . ( 1980 ) scale does not go beyond the spectral type m6 , no comparison for more advanced spectral types is possible . the problem is further complicated by possible effects of metallicity . the ridgway scale of is based on stars with solar metallicity ( @xmath125 ) and empirical calibrations of the -scale for @xmath126 are not available . to cope with this difficulty , we have introduced the metallicity- relation of bessell et al.(1989 , 1991 ) using the ( @xmath127 ) colour as a temperature indicator . an interpolation is made between the of bessell et al . ( 1989 ) and the ( @xmath127 ) colours given by fluks et al . ( 1994 ) for the spectral types from m0 to m10 . \(ii ) at a given metallicity , surface gravity @xmath128 , and , bcs are determined by convolving the corresponding stellar sed with the response functions of the various pass - bands . \(iii ) we recall that kurucz does not provide atmospheric models for both solar - scaled and mixtures . it is not therefore possible to estimate the impact ( if any ) of atmospheres on the isochrone colours and magnitudes . [ [ johnson - cousins - photometry . ] ] johnson - cousins photometry . + + + + + + + + + + + + + + + + + + + + + + + + + + + the response functions for the various pass - bands in which magnitudes and colours are generated are from the following sources : buser & kurucz ( 1978 ) for the @xmath129 pass - bands , bessell ( 1990 ) for the @xmath130 and @xmath131 cousins pass - bands , and finally bessell & brett ( 1988 ) for the @xmath132 pass - bands . re - normalization of the colours obtained by convolving the seds with pass - band , has been made by convolving the sed of vega and imposing that the computed colours strictly match the observed ones ( kurucz 1992 ) . finally , the zero point of the bcs is fixed by imposing that the bc for the kurucz model of the sun is @xmath133 . [ [ hst - wfpc2-photometry . ] ] hst - wfpc2 photometry . + + + + + + + + + + + + + + + + + + + + + the transformation to the wfpc2 photometric system ( filters f170w , f218w , f255w , f300w , f336w , f439w , f450w , f555w , f606w , f702w , f814w , and f850lp ) requires some additional explanations . the main parameters defining this photometric system are given in table [ hstfilter ] . columns ( 1 ) to ( 3 ) report the filter name , the mean wavelength @xmath134 and the r.m.s . band width ( bw ) , respectively . c|c|c|c|c|c filter & @xmath134 ( ) & bw ( ) & @xmath135 & @xmath136 & note + ( 1 ) & ( 2 ) & ( 3 ) & ( 4 ) & ( 5 ) & ( 6 ) + f170w & 1747 & 290 & @xmath2720.6926 & @xmath270.4070 & @xmath27 + f218w & 2189 & 171 & @xmath2720.8737 & @xmath270.2263 & @xmath27 + f255w & 2587 & 170 & @xmath2721.0940 & @xmath270.0060 & @xmath27 + f300w & 2942 & 325 & @xmath2721.1336 & + 0.0336 & wide u + f336w & 3341 & 204 & @xmath2721.1876 & + 0.0876 & u - john . + f439w & 4300 & 202 & @xmath2720.4325 & @xmath270.6675 & b - john . + f450w & 4519 & 404 & @xmath2720.6339 & @xmath270.4661 & wide b + f555w & 5397 & 522 & @xmath2721.0798 & @xmath270.0202 & v - john . + f606w & 5934 & 637 & @xmath2721.4250 & + 0.3250 & wide v + f702w & 6862 & 587 & @xmath2721.8594 & + 0.7594 & wide r + f814w & 7924 & 647 & @xmath2722.3309 & + 1.2309 & i - john . + f850lp & 9070 & 434 & @xmath2722.7578 & + 1.6578 & @xmath27 + with the wfpc2 detector at least three kinds of magnitudes are commonly in use : @xmath137 , @xmath138 and @xmath139 ( see the _ synphot user s guide _ by white et al . 1998 , distributed by the stsdas group ) . in the following we limit ourselves to consider only the theoretical counterparts of @xmath137 and @xmath138 . for radiation with flat spectrum and specific flux @xmath140 ( in @xmath141 ) impinging on the telescope , the @xmath137 magnitude is defined by @xmath142 which means that @xmath143 for @xmath144 @xmath145 . let us then define the pass - band @xmath146 as the product of the filter transmission by the response function of the telescope assembly and detector in use . in this case , @xmath140 is related to the _ counts _ rate @xmath147 i.e. the number of photons per second registered by the detector through @xmath148 where @xmath149 is the effective collecting surface of the telescope , and @xmath150 and @xmath151 are the planck constant and speed of light , respectively . @xmath147 , the quantity actually measured at the telescope , can be easily converted to a specific flux by means of eq . ( [ countrate ] ) . therefore , specific fluxes @xmath140 and @xmath152 magnitudes can be defined for incoming spectra of any shape . in the case of a stellar source of radius @xmath130 , located at a distance @xmath153 , and emitting a specific flux @xmath154 , eq . ( [ stmagdef ] ) can be replaced by @xmath155 \label{stmagdefstar}\ ] ] which defines the apparent @xmath152 magnitude of a star . in our case , we derive the absolute @xmath156 magnitudes by simply locating our synthetic stars at a distance of @xmath157 pc in eq . ( [ stmagdefstar ] ) . the specific fluxes @xmath158 are taken from the library of synthetic spectra in use as a function of @xmath105 , surface gravity @xmath159 , and chemical composition . similarly , we can define the @xmath160 magnitudes @xmath161 \label{vegamagdefstar}\ ] ] by imposing that the zero points @xmath135 are such that the synthetic magnitudes of vega match the ground - based @xmath162 johnson apparent magnitudes for this star , for the filters which are closest in wavelength to each other ( see column 6 in table [ hstfilter ] ) . we adopt the values 0.02 , 0.02 , 0.03 , 0.039 , 0.035 mag for the apparent magnitudes of vega in the @xmath162 johnson system ( see holtzman et al . 1995 ) . for the uv filters , vega is assumed to have an apparent magnitude equal to zero . we remind the reader that if we assume the apparent vega magnitude to be equal to zero in _ all _ the pass - bands , we would obtain @xmath163 and eq . ( [ vegamagdefstar ] ) would become @xmath164 this latter is the definition of _ vegamag _ in the synphot package distributed with the stsdas software . in this paper we do not make use of this calibration , but prefer to follow the slightly different one defined by holtzman et al . ( 1995 ) and described above . to get the final calibration , we insert in eq . [ vegamagdefstar ] the specific flux of vega calculated by castelli & kurucz ( 1994 ) with atmosphere models , and converted it to an absolute flux on the earth assuming for vega the angular diameter of @xmath165 mas ( code et al . 1976 ) . the absolute flux we have adopted differs ( a few percent ) from the one by hayes ( 1985 ) used by holtzman et al . ( once more , the absolute magnitude @xmath166 is obtained locating the synthetic star at a 10 pc distance . column ( 4 ) of table [ hstfilter ] lists the zero points we have obtained , whereas column ( 5 ) gives the difference @xmath167 between the zero point of the @xmath137 and @xmath138 magnitudes . the conversion from one magnitude scale to another is thus possible . finally , we calculate the bolometric corrections bc which allow one to convert theoretical bolometric magnitudes both to @xmath168 and @xmath169 : @xmath170 \nonumber \\ & + & 21.1 + 2.5 \log \frac{\int p(\lambda ) \,i_{\rm k}(\lambda)\ , \lambda \,{\mbox{d}}\lambda}{\int p(\lambda)\ , \lambda \,{\mbox{d}}\lambda } \\ { \rm bc}(p)^{m_{\rm vega}}_{t_{\rm eff},g } & = & { \rm bc}(p)^{m_{\rm st}}_{t_{\rm eff},g } -21.1 - z_p \label{bc_stmag}\end{aligned}\ ] ] where @xmath171 is the stefan - boltzmann constant . for the solar luminosity and absolute bolometric magnitude we adopt @xmath172 ( bahcall et al . 1995 ) and @xmath173 . owing to the presence of contaminants inside the wfpc2 ( see baggett & gonzaga 1998 and holtzman et al . 1995 ) , @xmath146 changes slowly with time ; in particular the uv throughput degrades , changing the photometric performances . to cope with this drawback , small corrections are usually added to the definition of the instrumental magnitudes in order to bring the magnitudes back to the optimal conditions ( see holtzman et al . 1995 ) . because of this , in the calculation of the bolometric corrections we do not use the present - day pass - bands @xmath146 provided by synphot but insert the pre - launch pass - bands , as in holtzman et al . ( 1995 ) . in this section we present the integrated magnitudes , colours and mass - to - light ratios for single stellar populations ( ssp ) of the same age and chemical compositions of the isochrones above , both in the johnson - cousins and wfpc2 photometric systems . the integrated magnitudes and colours are computed assuming an initial mass function @xmath174 and adopting the normalization @xmath175 where @xmath176 and @xmath177 are the lower and upper limits of zero age main sequence stars . by doing so , the integrated magnitudes of these ideal ssps can be easily scaled to produce integrated magnitudes for stellar populations of any arbitrary total _ initial _ mass @xmath178 , by simply adding the factor @xmath179 . the theoretical mass - to - light ratios @xmath180 are calculated following the same procedure as in alongi & chiosi ( 1990 ) and chandar et al . ( 1999 ) . the fate of a single star depends on its initial mass ( binary stars are neglected here ) . in a brief and over - simplified picture of stellar evolution the following mass limits and groupings can be identified : ( i ) stars more massive than @xmath181 have a lifetime shorter than the current estimate of the hubble age of the universe , say @xmath182 gyr . stars lighter than this limit are not of interest here because once formed they live for ever . ( ii ) stars more massive than @xmath69 explode as supernovae leaving a neutron star remnant of @xmath183 . the possibility that massive stars ( @xmath184 ) may end up as black holes is not considered here . ( iii ) stars less massive than @xmath185 and more massive than @xmath186 terminate their evolution as white dwarfs of suitable masses @xmath187 that depend on the initial mass and efficiency of mass loss during the rgb and agb phases . with the current estimates of the mass loss efficiency @xmath188 , i.e. @xmath189 . the possibility , however , exists that @xmath190 . in such a case stars in the mass interval @xmath185 to @xmath69 reach the c - ignition stage and deflagrate as supernovae leaving no remnant . for the purposes of the present study , this latter case is neglected and @xmath191 is always assumed . finally , let @xmath192 be the initial mass of the most massive star still alive in a ssp of a certain age . it is worth recalling that all the above mass limits depend on the initial chemical composition and that this dependence can be properly taken into account . therefore , at any given age @xmath193 the total mass @xmath194 of a ssp is made of two contributions : @xmath195 where the first right - hand side term is the total mass of the stars still alive at the age @xmath193 ( i.e. burning a nuclear fuel ) and the second one is the total mass in stellar remnants , i.e. white dwarfs or neutron stars depending on the initial mass of the progenitor ( i.e. @xmath196 for @xmath197 , and @xmath198 for @xmath199 ) . we adopt an upper mass limit @xmath200 . the choice for the lower mass limit , @xmath176 , is explained below . it is worth recalling here that , due to eq . [ eq_imf ] , @xmath201 is the current mass of an ideal ssp with initial mass equal to @xmath202 . @xmath201 decreases as a function of time from its initial value , due to the mass lost by both stellar winds ( for all masses ) and supernova explosions ( from massive stars ) . finally , the total luminosity @xmath203 of the ssp in any pass - band @xmath204 is obtained by integrating , along the isochrone , the luminosity of individual stars whose number is given by the imf @xmath205 to proceed further , one has to specify the initial mass function and @xmath176 . here , we adopt the salpeter ( 1955 ) law with slope @xmath206 over the whole mass range @xmath176 to @xmath177 . concerning @xmath176 , the choice @xmath207 is made by imposing that the observational value @xmath208 of the lmc cluster ngc 1866 is matched ( see girardi & bica 1993 ) . the age and metallicity of ngc 1866 are @xmath74 years and @xmath12 , respectively , whereas its @xmath209 is close to the mean value of 0.20 for lmc clusters of the same age ( see battinelli & capuzzo - dolcetta 1989 , and references therein ) . we have computed extended sets of evolutionary tracks and isochrones at relatively high metallicities . a major improvement , with respect to previous calculations of tracks , is that we made use of self - consistent opacities , i.e. computed with the same chemical compositions over the complete range of temperatures . the main result is that in general all evolutionary phases ( isochrones ) have higher effective temperature with respect to solar - scaled models ( isochrones ) of same metal content @xmath18 . the temperature shift is caused by the lower opacities of the mixtures . moreover , we find that at relatively high metallicities and old ages , an isochrone can not be mimicked by simply using a solar - scaled isochrone of lower metallicity . obvious limitation of the present results is that they refer to a particular choice for the chemical compositions derived from observations of a particular sample of metal - poor field stars . future observational data may suggest a different partition of metals in stars , and hence different results for the stellar models and corresponding isochrones . testing a large number of possible combinations of ratios , though feasible , would not be of practical use owing to the present uncertainty on the abundance ratios . therefore , the results of this study ought to be taken as indicative of the _ potential _ effect of element ratios . moreover , it is likely that variations of single @xmath4-elements are less important than the whole problem of including or not the enhancement of @xmath4-elements in the initial chemical composition . finally , it is still not clear whether differences in enhancements are significant compared to the errors in the determinations . [ [ retrieval - of - the - data - sets . ] ] retrieval of the data sets . + + + + + + + + + + + + + + + + + + + + + + + + + + + complete tabulations of the relevant data for all evolutionary sequences , isochrones , integrated magnitudes and colours and mass - to - light rations together with useful summary tables can be obtained either upon request to the authors or downloaded from the address ` http://pleiadi.pd.astro.it ` . the layout of the tables of stellar models and isochrones is the same as in girardi et al . ( 2000 ) as far as the johnson - cousins photometric system is concerned . for the wfpc2 system the layout is similar , with table headers allowing to easily identify the pass - band . we like to thank a. bressan for many suggestions and useful conversations , h. schlattl for his help with the opacity interpolation code , h. aussel and m. zoccali for their useful clarifications about the hst / wfpc2 photometric system , j. holtzman for providing us with the pre - launch wfpc2 pass - bands , and the anonymous referee for useful suggestions . l.g . acknowledges support from the alexander von humboldt - stiftung during his stay at mpa . thanks mpa for the warm hospitality and support . this study has been funded by the italian ministry of university , scientific research and technology ( murst ) under contract `` formation and evolution of galaxies '' n. 9802192401 . alexander d.r . , ferguson j.w . , 1994 , apj 437 , 879 schlattl h. , weiss a. , 1998 , in `` proc . neutrino astrophysics '' , ringberg castle , tegernsee , germany , 20 - 24 oct 1997 , ed . m. altmann , w. hillebrandt , h .- t.janka and g.raffelt ( sfb astroteilchenphysik , technical university munich )

the metal - insulator transition induced by disorder in the problem of anderson localization remains poorly understood from a theoretical point of view . on the one hand , it is true that the prediction for the existence of a metal - insulator transition induced by disorder has been verified by numerical simulations . on the other hand , the critical exponents at the metal - insulating transition measured numerically still can not be extracted from the field theories which are believed to embody the metal - insulator transition.@xcite the paradigm of this unpleasant situation is the plateau transition in the lowest landau level of the integer quantum hall effect.@xcite the development of reliable methods to compute analytically critical properties at a disorder - induced metal - insulator transition remains to this date an open problem . the simplest known example of disorder - induced criticality in the context of anderson localization was solved by dyson in 1953.@xcite according to dyson , the global density of states ( dos ) in the thermodynamic limit and close to the band center is anomalous when a single quantum particle hops with a random amplitude between the nearest - neighbor sites of a chain excluding any other form of disorder , in short , the one - dimensional random hopping problem . subsequent works showed that dyson s spectral anomaly is related to a diverging localization length,@xcite an anomalous decay of the envelope of wave functions,@xcite and an anomalous probability distribution of the resistance upon approaching the band center.@xcite the coalescence of the concept of universality imported from the theory of critical phenomena with the symmetry classification imported from random matrix theory has led to a classification of diffusive regimes and disorder - induced metal - insulator transitions in terms of 10 universality classes that are uniquely characterized by the intrinsic symmetries preserved by the disorder ( assumed weak ) for any given dimensionality of space.@xcite according to this classification , dyson s singular dos is a trademark of the chiral universality classes . the chiral universality classes are realized in problems of anderson localization for which the hamiltonian anticommutes with some unitary operator for all realizations of the disorder and in which case the dos is known to be singular at the band center in zero,@xcite quasi - one,@xcite and two dimensions.@xcite even though the chiral universality classes are critical at the band center , they have been far less studied than the orthogonal , unitary , and symplectic universality classes , which we will refer to as the standard universality classes . for example , the probability distribution of the so - called mesoscopic local density of states ( ldos ) has been known since 1989 from the work of altshuler and prigodin for a single chain with weak on - site disorder,@xcite whereas it has not yet been computed for the one - dimensional random - hopping problem . the first aim of this paper is to fill this gap . if a diffusive regime exists , perturbative techniques can be used to describe the crossovers between universality classes.@xcite however , little is known quantitatively on the crossover between two different universality classes , here defined in terms of nonlinear - sigma models , say , from the diffusive to the localized regimes . this is even true in one dimension , which is by far the most studied laboratory for anderson localization as was recently illustrated by a flurry of works on the validity of one - parameter scaling in one dimension.@xcite the absence of a diffusive regime renders the concept of universality classes ambiguous in one dimension in that symmetry alone does not specify a universality class in one dimension . for example , the ratio of the localization length to the mean free path in a wire of finite width interpolates smoothly between its two limiting values for fully preserved or completely broken time - reversal symmetry , respectivley , as a function of a weak magnetic field.@xcite similarly , fine tuning of microscopic parameters is required to achieve delocalization ( i.e. , diverging localization length ) of quasiparticles at the fermi energy in a dirty superconducting wire of finite thickness with both broken spin - rotation and time - reversal symmetries whereas delocalization becomes generic in the thick wire scaling limit.@xcite one can nevertheless define a _ symmetry class _ in one dimension by demanding that single - parameter scaling holds . lack of universality in one dimension can then be understood as the fact that there is no scaling limit for which single - parameter scaling becomes a generic property of an ensemble of random microscopic hamiltonians of a given symmetry . symmetry classes in one dimension can not be construed as an enumeration of stable or unstable fixed points of some putative effective field theory . computing the crossover between symmetry classes in one dimension nevertheless remains a well - defined problem . time - reversal symmetry can not be broken for a spinless particle constrained to move in a one - dimensional and simply connected world , in which case the only possible crossover between symmetry classes takes place between the chiral and standard classes . the second aim of this paper is to compute the crossover between these two symmetry classes in one dimension for the probability distributions of the conductance and of the ldos from the ballistic to the localized regimes assuming a weak disorder at the microscopic level . in that regard , our results are complementary to the analysis made by schomerus and titov of the validity of one - parameter scaling in one dimension in which they computed exactly the first four cumulants of the logarithm of the conductance to leading order in the ratio of the length of the chain to the localization length assuming a weak disorder at the microscopic level.@xcite we thus propose an approximate ansatz for the crossover whereby the approximation consists of assuming that the phase and amplitude of the reflection coefficient separate . by comparing the cumulants of the lyapunov exponent of the transfer matrix computed with our approximate ansatz to the ones computed with the large - deviation ansatz of schomerus and titov deep in the localization regime , we deduce that our ansatz captures the first cumulant , but fails with the higher cumulants . the paper is organized as follows . we formulate the one - dimensional problem of anderson localization as a continuous schrdinger equation with a relativistic kinetic energy in sec . [ model ] . a functional renormalization group equation is derived for the probability distribution of the reflection coefficient in sec . [ functional renormalization group equations ] . exact solutions to these coupled functional renormalization group equations are given for the chiral and the standard classes , as well as an approximate solution that describes the crossovers between these two classes in sec . [ sec : solutions to the functional renormalization group equations ] . the accuracy of the approximate solution for the crossover regime is estimated in sec . [ sec : validity of the separation ansatz for the crossover regime ] . the probability distribution of the ldos is computed exactly in the chiral and standard classes whereas it is computed approximately in the crossover regime between these two classes in sec . [ sec : probability distribution of the ldos ] . we conclude in sec . [ sec : conclusion ] . ( 220,150)(-120,-130 ) ( -80 , 20)(1,0)140 ( -80,-10)(1,0)140 ( 60,-10)(0,1)30 ( -40 , 5 ) ( 0,10)-180 ( -40 , 15)(-1 , 0 ) 20 ( -40,-5)(-1 , 0 ) 20 ( -120 , 17)@xmath0 ( -95 , 5)@xmath1 ( -110 , -5)@xmath2 ( -70,-10)(0,1)30 ( -70,20)(10,0)10(1,-1)30 ( -70,10)(1,-1)20 ( -70,0)(1,-1)10 ( 30,20)(1,-1)30 ( 40,20)(1,-1)20 ( 50,20)(1,-1)10 ( -80,-30)(1,0)140 ( -80,-60)(1,0)140 ( 60,-60)(0,1)30 ( -40,-45 ) ( 0,10)-180 ( -40,-35)(-1 , 0 ) 20 ( -40,-55)(-1 , 0 ) 20 ( -120 , -33)@xmath3 ( -95,-45)@xmath1 ( -70,-60)(0,1)30 ( -70,-30)(10,0)10(1,-1)30 ( -70,-40)(1,-1)20 ( -70,-50)(1,-1)10 ( 30,-30)(1,-1)30 ( 40,-30)(1,-1)20 ( 50,-30)(1,-1)10 ( -80 , -80)(1,0)150 ( -80,-110)(1,0)150 ( -40 , -95 ) ( 0,10)-180 ( -40 , -85)(-1 , 0 ) 20 ( -40,-105)(-1 , 0 ) 20 ( -120 , -83)@xmath4 ( -95 , -95)@xmath1 ( -70,-110)(0,1)30 ( 60,-110)(0,1)30 ( -70,-80)(10,0)10(1,-1)30 ( -70,-90)(1,-1)20 ( -70,-100)(1,-1)10 ( 30,-80)(1,-1)30 ( 40,-80)(1,-1)20 ( 50,-80)(1,-1)10 ( -70,-120)(1,0)130 ( 60,-120)(-1,0)130 ( -5 , -130)@xmath5 in this paper we want to investigate the statistical properties of the ldos @xmath6 and dimensionless conductance @xmath7 for a single quantum particle , which is ( i ) restricted to a simply connected and strictly one - dimensional ( 1d ) geometry and ( ii ) subjected to a static and weak random environment ( disorder ) that interpolates smoothly between the standard and chiral symmetry classes . condition ( i ) implies that time - reversal symmetry is preserved for every realization of the disorder . in 1d and assuming a metallic ground state in the absence of the disorder , the ground state consists of two distinct fermi points . plane waves at the fermi points are called left- and right - movers , respectively . in the standard symmetry class , the channels for forward and backward scatterings induced by an on - site disorder potential in the basis of left- and right - movers are equally likely up to nonrandom oscillatory factors . in the chiral symmetry class , the disorder potential is off - diagonal in the basis of left- and right - movers . presuming weak disorder in condition ( ii ) allows us to choose a kinetic energy that is a first - order differential operator , for it only makes sense to linearize the spectrum about the two fermi points when the disorder potential is weak . we thus choose to model the dynamics of the single quantum particle and the static random environment by the hamiltonian [ eq : def our anderson localization problem ] @xmath8 where we have set the fermi velocity @xmath9 and @xmath10 to one . the boundary conditions that will be imposed in the sequel are depicted in fig . [ fig : geometries ] . they all obey condition ( i ) . we have denoted by @xmath11 the unit @xmath12 matrix and by @xmath13 the @xmath14 pauli matrices . the static random environment is represented by three independent potentials @xmath15 , @xmath16 that we choose , for mathematical convenience , to be supported on the interval @xmath17 , and to be white - noise and gaussian distributed with vanishing mean , ( disorder averaging is denoted by @xmath18 ) @xmath19 the three variances @xmath20 carry the dimensions of inverse length and need not be equal . a finite value @xmath21 can also be easily accommodated in our formalism.@xcite in this case , the clean system is insulating as @xmath22 opens up a dimerization gap at the fermi energy and the random environment mimics to a first approximation the static fluctuations of the phonons responsible for the dimerization sufficiently close to the peierls transition.@xcite the problem of anderson localization in the continuum defined by eq . ( [ eq : def our anderson localization problem ] ) is a coarse - grained version of a single particle hopping on a simply - connected chain with a uniform nearest - neighbor hopping amplitude subjected to weak and independent random fluctuations of an on - site potential and of the nearest - neighbor hopping amplitude . the generic symmetry obeyed by eq . ( [ eq : def our anderson localization problem a ] ) is time - reversal invariance , @xmath23 if the condition @xmath24 is satisfied , or , for that matter , any equivalent condition obtained by a redefinition of the pauli matrices through a position independent su(2 ) rotation around @xmath25 , eq . ( [ eq : def our anderson localization problem a ] ) has an additional chiral symmetry in that it anticommutes with @xmath26 , @xmath27 consequently , all nonvanishing eigenvalues of @xmath28 come in pairs with opposite signs . the origin of the chiral symmetry of the continuum model for any given realization of the disorder is a sublattice symmetry of an appropriate lattice regularization . a microscopic random hamiltonian realizing the chiral symmetry is , for example , a tight - binding hamiltonian on a hypercubic lattice with strictly vanishing on - site energies and random hopping matrix elements restricted to nearest - neighbor sites , in short , the random hopping problem . the ldos , dos , conductance , shot - noise power , etc . , can all be extracted from the elements of the @xmath14 scattering matrix @xmath29 at the energy @xmath30 that connects incoming ( `` i '' ) and outgoing ( `` o '' ) states @xmath31 the amplitude for incoming and outgoing waves to the left- and right - hand side of the disordered region is denoted by @xmath32 here . the matrix elements @xmath33 are the reflection coefficients whereas the matrix elements @xmath34 are the transmission coefficients . alternatively , one can work with the @xmath14 transfer matrix @xmath35 at the energy @xmath30 which follows from the scattering matrix ( [ eq : def s matrix ] ) through the basis transformation implied by @xmath36 conservation of probability dictates that the scattering matrix is unitary or , equivalently , that the transfer matrix is pseudo - unitary @xmath37 conservation of probability also implies the existence of the polar decompositions [ eq : polar decomposition s+m ] @xmath38 where @xmath39 , @xmath40 , @xmath41 , @xmath42 are independent complex numbers with @xmath43 and @xmath44 . the polar decomposition is not unique . for example , the sign of @xmath45 can be absorbed into a redefinition of @xmath39 , @xmath40 , @xmath41 , @xmath42 . the variable @xmath45 , when restricted to the halfline @xmath46 , has the geometrical interpretation of a radial coordinate on a riemannian manifold.@xcite the time - reversal symmetry of the hamiltonian implies the transformation laws @xmath47 which enforce the constraints @xmath48 and @xmath49 or , equivalently , @xmath50 . the chiral symmetry of the hamiltonian implies the transformation laws @xmath51 which enforce the constraints @xmath52 and @xmath53 or , equivalently , @xmath54 , @xmath55 , @xmath56 . then , at the band center @xmath57 with the condition ( [ eq : def chiral limit for g s ] ) , the scattering matrix becomes hermitian . under the assumptions that ( i ) the disorder is weak and ( ii ) @xmath39 , @xmath40 , @xmath41 , @xmath42 are all independently and uniformly distributed on the unit circle in the complex plane , it was essentially shown in ref . that the probability distribution @xmath58 of @xmath59 obeys the fokker - planck equation @xmath60 in the geometry of fig . [ fig : geometries]c . the mean free path @xmath61 is some function of @xmath62 , @xmath63 , @xmath64 , and @xmath65 . in a more general context of multi channel quantum wires assumption ( ii ) is known as the isotropy assumption,@xcite whereas the fokker - planck equation ( [ eq : fp std limit ] ) is a special case of the dorokhov - mello - pichard - kumar ( dmpk ) equation.@xcite a consequence of assumption ( ii ) is one - parameter scaling as encoded by the fokker - planck equation ( [ eq : fp std limit ] ) , i.e. , the probability distribution of @xmath66 and , consequently , of @xmath67 depends on the single dimensionless parameter @xmath68 . it is evident that assumption ( ii ) breaks down at the band center @xmath57 and with the chiral condition ( [ eq : def chiral limit for g s ] ) because @xmath69 and @xmath70 are then both real valued . correspondingly , the fokker - planck equation obeyed by the probability distribution @xmath58 of @xmath71 is different from eq . ( [ eq : fp std limit ] ) and given by the diffusion equation @xmath72 in the geometry of fig . [ fig : geometries]c.@xcite the mean free path @xmath61 is some function of @xmath64 . again , the fokker - planck equation ( [ eq : fp chiral limit ] ) encodes one - parameter scaling as the probability distribution of @xmath66 and , consequently , of @xmath67 depends on the single dimensionless parameter @xmath68 . there are no known multiple - parameter scaling equations for the ldos , dos , conductance , etc . , which describe the crossover between the two limiting cases described by eqs . ( [ eq : fp std limit ] ) and ( [ eq : fp chiral limit ] ) for values of @xmath5 ranging from the ballistic ( @xmath5 smaller than the mean free path ) to the localized regime ( @xmath5 larger than the mean free path ) to the best of our knowledge ( see refs . for discussions of two - parameter scaling ) . we fill this gap in the remainder of the paper . ( 240,160)(0,-140 ) ( 0,10)(a ) ( + 10,-10)(1,0)220 ( + 10,-50)(1,0)220 ( 50,-50)(0,1)40 ( 190,-50)(0,1)40 ( 120,-60)(-1,0)70 ( 120,-60 ) ( 1,0)70 ( 117,-70)@xmath5 ( 40,-50)(0,1)40 ( 20 , 0 ) ( 1,0)20 ( 70 , 0)(-1,0)20 ( 40 , 5)@xmath73 ( 50,-20)(1,-1)30 ( 50,-30)(1,-1)20 ( 50,-40)(1,-1)10 ( 50,-10)(10,0)11(1,-1)40 ( 190,-20)(-1,1)10 ( 190,-30)(-1,1)20 ( 190,-40)(-1,1)30 ( 40,-50)(0,10)4(1,1)10 ( 50,-50)(0,10)4(-1,1)10 ( 0,-80)(b ) ( 10,-100)(1,0)220 ( 10,-140)(1,0)220 ( 40,-140)(40,0)2(0,1)40 ( 160,-140)(40,0)2(0,1)40 ( 80,-140)(-10,0)4(-1,1)10 ( 40,-100)(10 , 0)4(1,-1)10 ( 40,-100)(0,-10)4(1,-1)10 ( 80,-140)(0 , 10)4(-1,1)10 ( 200,-140)(-10,0)4(-1,1)10 ( 160,-100)(10 , 0)4(1,-1)10 ( 160,-100)(0,-10)4(1,-1)10 ( 200,-140)(0 , 10)4(-1,1)10 ( 10,-113)@xmath74 ( 10,-130)@xmath75 ( 208,-113)@xmath76 ( 208,-130)@xmath77 ( 85,-113)@xmath78 ( 85,-130)@xmath79 ( 130,-113)@xmath80 ( 130,-130)@xmath81 ( 55,-122)@xmath82 ( 175,-122)@xmath83 we shall show in secs . [ sec : solutions to the functional renormalization group equations ] and [ sec : probability distribution of the ldos ] that the statistical properties of the dimensionless conductance @xmath7 and of the ldos @xmath6 follow from the knowledge of the statistical properties of the reflection coefficient @xmath84 which has been decomposed into its square modulus @xmath85 and phase @xmath86 . for a disordered wire of length @xmath87 as depicted in fig . [ fig : composition law and rg step ] , the reflection coefficient @xmath88 is related to the entries of the scattering matrix @xmath89 for the slice of length @xmath73 and to the entry @xmath90 of the scattering matrix @xmath91 for the much longer segment of length @xmath5 by the composition law @xmath92 when the width @xmath73 of the slice is much larger than the lattice spacing @xmath93 but much smaller than the mean free path @xmath61 , here defined by @xcite @xmath94 we infer the continuous langevin process [ eq : langevin eq for r and phi ] @xmath95 by using the relations @xmath96 which are valid up to first order in the disorder potentials @xmath97 , and by expanding the right - hand side of eq . ( [ eq : composition law reflection amplitude ] ) to the same order . we note in passing that the mean free path ( [ eq : def ell ] ) is simply given by @xmath98 in the born approximation . for later convenience , we have continued the energy @xmath30 to the upper part of the complex plane , @xmath99 @xmath100 . the continuous langevin process ( [ eq : langevin eq for r and phi ] ) can also be formulated as the fokker - planck equation obeyed by the joint probability distribution function @xmath101 . using the standard methods of chapter 3 in ref . say , one finds@xcite [ eq : master equation in terms of r and phi ] @xmath102 where @xmath103 and @xmath104 we have introduced the short - hand notations @xmath105 observe that the dependence on the polar angle @xmath106 arises solely from @xmath107 and @xmath108 in the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) . thus , provided initial and boundary conditions are also periodic on the interval @xmath109 , the joint probability distribution @xmath101 is periodic on the interval @xmath109 even though the phase @xmath106 is originally defined modulo @xmath110 . the periodicity @xmath111 is a consequence of the langevin process ( [ eq : langevin eq for r and phi ] ) being invariant under @xmath112 and @xmath113 . a dimerization gap @xmath114 produces the changes @xmath115 the polar decomposition ( [ eq : polar decomposition s+m ] ) suggests the change of variable @xmath116 in terms of which the langevin process ( [ eq : langevin eq for r and phi ] ) becomes [ eq : langevin eq for x and phi ] @xmath117 in turn , the joint probability distribution function @xmath118 obeys the fokker - planck equation [ eq : master equation in terms of x and phi ] @xmath119 where @xmath120 and @xmath121 the langevin process ( [ eq : langevin eq for x and phi ] ) or the fokker - planck equation ( [ eq : master equation in terms of x and phi ] ) are invariant under @xmath122 and @xmath113 . equations ( [ eq : master equation in terms of r and phi ] ) or ( [ eq : master equation in terms of x and phi ] ) are , from a conceptual point of view , the main result of this paper as they determine under what conditions one - parameter scaling holds in a 1d weakly disordered quantum wire . for example , eq . ( [ eq : master equation in terms of x and phi ] ) encodes the functional renormalization group flow of the joint probability distribution of the radial coordinate @xmath123 and phase @xmath106 of the reflection coefficient @xmath1 . if it is the property that the functional renormalization group flow of the probability distribution of @xmath66 depends solely on the dimensionless ratio @xmath68 aside from @xmath66 , which is understood as one - parameter scaling for @xmath66,@xcite then two examples of one - parameter scaling can be constructed from eq . ( [ eq : master equation in terms of x and phi ] ) as follows . first , the fokker - planck ( dmpk ) equation ( [ eq : fp std limit ] ) follows from insertion of the ansatz @xmath124 into eq . ( [ eq : master equation in terms of x and phi ] ) with @xmath125 and from integration over @xmath126 of the resulting equation . second , the fokker - planck ( dmpk ) equation ( [ eq : fp chiral limit ] ) follows from insertion of the ansatz @xmath127 \mathcal{x}(x;l ) \hphantom{aa}\end{aligned}\ ] ] with @xmath128 into eq . ( [ eq : master equation in terms of x and phi ] ) and from integration over @xmath106 of the resulting equation . one - parameter scaling is not the rule for generic initial and boundary conditions of eq . ( [ eq : master equation in terms of x and phi ] ) and for generic values of @xmath65 and @xmath20 due to the absence of a diffusive regime in 1d.@xcite this is most clearly seen by the fact that the localization length @xmath129 , as defined by the typical dependence of the conductance @xmath67 on @xmath130 , is finite , @xmath131 , for the standard symmetry class , but diverges in the chiral limit @xmath128 . hence , the localization length must depend on a second microscopic parameter aside from @xmath61 for generic values of @xmath65 and @xmath20 . equations ( [ eq : master equation in terms of r and phi ] ) and ( [ eq : master equation in terms of x and phi ] ) describe the full crossover from the chiral to the standard class with no restrictions on the values of @xmath66 and @xmath5 . for any given absorption @xmath132 , the parameter space @xmath133 with @xmath134 and @xmath135 of eqs . ( [ eq : master equation in terms of r and phi ] ) and ( [ eq : master equation in terms of x and phi ] ) is three - dimensional and is depicted in fig . [ fig : parms ] . remarkably , the parameter @xmath136 always enters eqs . ( [ eq : master equation in terms of r and phi ] ) and ( [ eq : master equation in terms of x and phi ] ) as a prefactor to @xmath137 or @xmath138 and conversely @xmath137 or @xmath138 are always multiplied by @xmath136 . as we shall see in sec . [ sec : solutions to the functional renormalization group equations ] this property turns out to be crucial in our study of the crossover between the chiral and standard symmetry classes . equation ( [ eq : master equation in terms of r and phi ] ) was derived in the standard class by abrikosov , melnikov , and kumar , and by rammal and doucot among others starting from the continuous nonrelativistic schrdinger equation.@xcite the large @xmath66 limit @xmath139 of eq . ( [ eq : master equation in terms of x and phi ] ) was derived by schomerus and titov when @xmath125 and @xmath140 starting from a langevin process for the eigenfunctions of the discrete schrdinger equation on a chain . @xcite ( 180,170)(0,10 ) ( + 60,+70 ) ( 0,1)100 ( 0,+55 ) ( 4,1)180 ( 30,+90 ) ( 3,-2)120 ( 20,60 ) ( 0,1)40 ( 100,80 ) ( 0,1)40 ( 80,20 ) ( 0,1)70 ( 120,30 ) ( 0,1)70 ( 160,40 ) ( 0,1)70 ( 60,15 ) ( 4,1)120 ( 20,100 ) ( 4,1)80 ( 80,90 ) ( 4,1)80 ( 95,10 ) ( -3,2)90 ( 85,90 ) ( 3,-2)90 ( 30,160 ) ( 3,-2)90 ( 80,80)(1,1)13 ( 80,75)(1,1)20 ( 80,70)(1,1)26 ( 80,65)(1,1)32 ( 80,60)(1,1)40 ( 80,55)(1,1)46 ( 80,50)(1,1)53 ( 80,45)(1,1)59.5 ( 80,40)(1,1)66.5 ( 80,35)(1,1)73 ( 80,30)(1,1)80 ( 80,25)(1,1)80 ( 80,20)(1,1)80 ( 160,95)(-1,-1)73.5 ( 160,90)(-1,-1)66.5 ( 160,85)(-1,-1)59.5 ( 160,80)(-1,-1)53 ( 160,75)(-1,-1)46 ( 160,70)(-1,-1)40 ( 160,65)(-1,-1)33 ( 160,60)(-1,-1)26.5 ( 160,55)(-1,-1)19.5 ( 160,50)(-1,-1)13 ( 30,95)(1,-1)15 ( 30,100)(1,-1)30 ( 30,105)(1,-1)45 ( 30,110)(1,-1)60 ( 30,115)(1,-1)75 ( 30,120)(1,-1)90 ( 30,125)(1,-1)90 ( 30,130)(1,-1)90 ( 30,135)(1,-1)90 ( 30,140)(1,-1)90 ( 30,145)(1,-1)90 ( 30,150)(1,-1)90 ( 30,155)(1,-1)90 ( 30,160)(1,-1)90 ( 120,95)(-1,1)15 ( 120,90)(-1,1)30 ( 120,85)(-1,1)45 ( 120,80)(-1,1)60 ( 120,75)(-1,1)75 ( 170,105)@xmath136 ( 48,165)@xmath141 ( 150,15)@xmath65 ( 65,11)@xmath142 ( 13,50)@xmath143 ( 58,60)@xmath144 ( 93,70)@xmath145 ( 62,113)@xmath146 ( 21,65)ch1 ( 102,85)ch2 ( 21,105)kw1 ( 102,125)kw2 ( 20,60 ) ( 100,80 ) ( 20,100 ) ( 100,120 ) exact and approximate solutions to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) are derived in this section . we will consider three regimes of parameter in fig . [ fig : parms ] . the first regime corresponds to the chiral symmetry class . the second regime corresponds to the standard symmetry class . the third regime describes a specific path connecting the two symmetry classes . the chiral symmetry class is given by the condition [ eq : def chiral regime parameters ] @xmath147 which defines the band center of the energy eigenvalue spectrum of hamiltonian ( [ eq : def our anderson localization problem a ] ) when @xmath148 i.e. , when hamiltonian ( [ eq : def our anderson localization problem a ] ) anticommutes with @xmath26 as in eq . ( [ eq : def chiral symmetry for h ] ) . the standard symmetry class occurs whenever [ eq : strong+weak def of std limit ] @xmath149 or @xmath150 or @xmath151 since any one of these three conditions decouples @xmath152 from @xmath106 in eq . ( [ eq : master equation in terms of r and phi ] ) and is consistent with a stationary probability distribution of @xmath106 which is uniformly distributed . in this paper we shall only consider the crossover between the chiral and standard symmetry classes along the line @xmath153 we must naturally choose the initial and boundary conditions to be imposed on the solutions to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) . our choice is motivated by the computation of the probability distribution of the ldos that we will carry out in sec.[sec : probability distribution of the ldos ] . two geometries depicted in fig . [ fig : bc for ldos ] will be considered in sec.[sec : probability distribution of the ldos ] . the relevant geometries for the probability distribution of the reflection coefficient @xmath1 that enter the computation of the ldos are depicted in fig . [ fig : geometries ] . in figs . [ fig : geometries]a and [ fig : geometries]b the disordered quantum wire is closed on the right - hand side and connected to a reservoir on the left - hand side . in fig . [ fig : geometries]a conservation of probability holds . in fig . [ fig : geometries]b conservation of probability does not hold due to the presence of an absorption encoded by a finite imaginary part @xmath154 of @xmath30 . in fig . [ fig : geometries]c the disordered quantum wire is open at both ends and flux conservation ( no absorption ) is assumed . we choose the ideal initial condition @xmath155 and @xmath156 when @xmath157 for the semiopen and fully open geometries of fig . [ fig : geometries ] , respectively . in the chiral symmetry class , we demand that the phase of @xmath1 is either 0 or @xmath111 with probability 1/2 initially . in the standard symmetry class , we demand that the phase of @xmath1 is uniformly distributed in the interval @xmath158 initially . in view of the periodicity of the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) we always impose initial conditions with a periodicity of @xmath111 . we shall now construct exact solutions to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) in the chiral and standard symmetry classes , respectively , for the three geometries of fig . [ fig : geometries ] . we shall also solve the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) in the crossover regime ( [ eq : crossover regime weak if a ] ) for the geometry of fig . [ fig : geometries]a . we could not construct exact solutions to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) in the crossover between the chiral and standard symmetry classes for the geometries of figs . [ fig : geometries]b and [ fig : geometries]c . we shall construct , however , an approximate solution to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) in a crossover regime that interpolates between the exact solutions for the geometries of figs . [ fig : geometries]b and [ fig : geometries]c . the accuracy of this approximate solution will be discussed in sec . [ sec : validity of the separation ansatz for the crossover regime ] . in geometry [ fig : geometries]a , an incoming plane wave from an ideal lead is perfectly reflected by the disordered quantum wire . hence , the reflection coefficient @xmath1 of the disordered quantum wire must be a pure phase , @xmath159 , for all @xmath160 . we thus insert the separation - of - variable ansatz @xmath161 into the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) that , after integration over @xmath152 , reduces to @xmath162 \frac{\partial\phi}{\partial\phi } \nonumber \\ & & + \left [ 2\zeta^{\ } _ { 0 } + 1 + \zeta \cos(2\phi ) \right ] \frac{\partial^{2}\phi}{\partial\phi^{2}}. \label{eq : fp if closed at one end and no absorption}\end{aligned}\ ] ] by assumption , we have switched off the absorption @xmath125 . our choice of initial condition shall depend on the proximity to the parameter regimes ( [ eq : def chiral regime parameters ] ) and ( [ eq : strong+weak def of std limit ] ) but shall always be periodic on the interval @xmath109 . in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) the fokker - planck equation ( [ eq : fp if closed at one end and no absorption ] ) reduces to @xmath163\frac{\partial^{2}\phi}{\partial\phi^{2 } } \nonumber \\ \label{eq : fp if closed at one end and no absorption ch}\end{aligned}\ ] ] and has the normalized , stationary , and periodic solution @xmath164 whenever the phase @xmath106 is @xmath144 or @xmath111 with probability @xmath165 initially . the same conclusion could have been anticipated by inspection of the langevin process ( [ eq : langevin eq for r and phi ] ) or from the fact that the chiral symmetry class ( [ eq : def chiral regime parameters ] ) imposes the condition that the scattering matrix is hermitian , see eq . ( [ eq : chiral symmetry s ] ) . in the standard symmetry class ( [ eq : strong def of std limit ] ) , @xmath107 and @xmath108 drop out from the fokker - planck equation ( [ eq : fp if closed at one end and no absorption ] ) : @xmath166 its normalized , stationary , and periodic solution is @xmath167 whenever the phase @xmath106 of @xmath1 is uniformly distributed initially . more generally , the solution to eq . ( [ eq : fp if closed at one end and no absorption std ] ) converges to the uniform probability distribution ( [ eq : solution std limit semi - open geometry ] ) as soon as @xmath168 for any given initial probability distribution . this is also true for the standard symmetry class ( [ eq : yet another def of std limit ] ) and ( [ eq : weak def of std limit ] ) as @xmath169 and @xmath170 are ineffective in these limits . when @xmath171 and @xmath172 the stationary probability distribution of @xmath1 need not be uniform in @xmath158 . for example , consider the case @xmath173 , @xmath174 , @xmath175 , for which the fokker - planck equation ( [ eq : fp if closed at one end and no absorption ] ) reduces to [ eq : band center anomaly ] @xmath176 \frac{\partial^{2}\phi}{\partial\phi^{2 } } \label{eq : band center anomaly a}\end{aligned}\ ] ] and has the stationary and normalized solution@xcite @xmath177 with a periodicity of @xmath111 . here @xmath178 denotes the value of the gamma function for @xmath179 . equation ( [ eq : band center anomaly b ] ) is a special case of the stationary solution @xmath180 to eq . ( [ eq : fp if closed at one end and no absorption ] ) with @xmath181 as shown by titov.@xcite in the chiral limit @xmath182 and @xmath183 , this solution becomes unnormalizable and we have to use eq . ( [ eq : sum of delta functions ] ) as a stationary solution instead . for simplicity , we consider the crossover regime ( [ eq : crossover regime weak if a]).@xcite the chiral symmetry class ( [ eq : def chiral regime parameters ] ) is reached when @xmath181 . the standard symmetry class ( [ eq : weak def of std limit ] ) is reached when @xmath184 . in the crossover regime , the fokker - planck equation ( [ eq : fp if closed at one end and no absorption ] ) reduces to @xmath185 a stationary solution to eq . ( [ eq : chiral 1 fp with no absorption and r=1 ] ) is constructed from the solution @xmath186 valid for @xmath187 and on the interval @xmath109 , by periodic extension to the interval @xmath158 . the constant @xmath188 is chosen so that the solution ( [ eq : sol cross if geo c ] ) is properly normalized . the solution interpolates between eqs . ( [ eq : sum of delta functions ] ) and ( [ eq : solution std limit semi - open geometry ] ) as @xmath189 and @xmath190 , respectively , as it should be . equation ( [ eq : sol cross if geo c ] ) is an important intermediary result as we will use it to construct an approximation to the probability distributions for the conductance and ldos in the crossover regime that interpolate between the exact chiral and standard limiting probability distributions . in geometry [ fig : geometries]b , an incoming plane wave from an ideal lead is not perfectly reflected by the disordered quantum wire due to a finite absorption @xmath191 . hence , the reflection coefficient @xmath1 of the disordered quantum wire is not a pure phase anymore , @xmath192 . in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) the phase @xmath106 and squared magnitude @xmath152 of the reflection coefficient @xmath1 must necessarily separate since @xmath193 is necessarily real valued , @xmath194 \mathcal{r}(r;t ) . \hphantom{aa } \label{eq : separation variable chiral geometry b}\end{aligned}\ ] ] after insertion of eq . ( [ eq : separation variable chiral geometry b ] ) into the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) one finds that @xmath195 obeys @xmath196\mathcal{r } \nonumber \\ & & + \big[2\omega r + 3 ( 1-r)(1 - 3r ) \big ] \frac{\partial \mathcal{r}}{\partial r } \nonumber \\ & & + 2r(1-r)^{2}\frac{\partial^{2 } \mathcal{r}}{\partial r^{2 } } \label{eq : fp chiral geometry b}\end{aligned}\ ] ] with the stationary solution @xmath197 the constant @xmath188 is chosen so that the solution ( [ eq : solution r for closed wire with absorption ( chiral ) ] ) is properly normalized ( this is always possible for @xmath198 ) . the stationary solution ( [ eq : solution r for closed wire with absorption ( chiral ) ] ) has a square - root singularity at @xmath199 , an essential singularity at @xmath200 , and is normalizable for @xmath201 $ ] . we shall see below that the probability for @xmath152 to be in the vicinity of 0 is enhanced in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) compared to the standard symmetry class ( [ eq : strong+weak def of std limit ] ) . in the standard symmetry class ( [ eq : strong+weak def of std limit ] ) the phase @xmath106 and squared magnitude @xmath152 of the reflection coefficient @xmath1 can reasonably be taken to separate in view of the initial condition of a uniformly distributed phase @xmath106 , @xmath202 after insertion of the separation - of - variable ansatz ( [ eq : separation variable std geometry b ] ) into the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) one finds that @xmath195 obeys @xmath203\mathcal{r } \nonumber \\ & & + \big [ 2\omega r + ( 1-r)(1 - 5r ) \big ] \frac{\partial \mathcal{r}}{\partial r } \nonumber\\ & & + r(1-r)^{2}\frac{\partial^{2 } \mathcal{r}}{\partial r^{2 } } \label{eq : fp standard limit closed both sides with abs}\end{aligned}\ ] ] with the stationary and normalized solution @xmath204 the stationary solution ( [ eq : solution r for closed wire with absorption ( std ) ] ) has an essential singularity when @xmath200 . the stationary solution ( [ eq : solution r for closed wire with absorption ( std ) ] ) was found by pradhan and kumar in ref . . when the initial condition on the probability distribution of @xmath106 is not that of a uniform distribution , it is presumed that the deviations of the solution to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) from the separation - of - variable ansatz ( [ eq : separation variable std geometry b ] ) with the stationary solution ( [ eq : solution r for closed wire with absorption ( std ) ] ) are short transients . we do not know of an analytical verification of this assumption . there is no compelling reason to believe that the squared amplitude @xmath152 and the phase @xmath106 of the reflection coefficient @xmath1 separate in the crossover regime . however , short of an explicit exact solution to the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) in the crossover regime , we shall nevertheless pursue a strategy relying on an approximation built on a separation - of - variable ansatz that allows us to interpolate between the chiral and standard symmetry classes . one possible alternative to the separation - of - variable ansatz that we tried is a perturbative expansion about the standard symmetry class ( [ eq : strong+weak def of std limit ] ) using a fourier expansion of the joint probability distribution for @xmath152 and @xmath106 . however , this approach has the drawback that it can not be expected to interpolate all the way to the chiral symmetry class ( [ eq : def chiral regime parameters ] ) in view of the nature of the singularities characterizing the limiting probability distributions . for this reason we will not present the perturbative approach in this paper . our starting point is the separation - of - variable ansatz [ eq : separation - of - variable ansatz in geometry b ] @xmath205 we further assume that the probability distribution function for the phase @xmath106 shows a quick relaxation to its stationary solution , i.e. , @xmath206 is assumed to be @xmath207-independent , @xmath208 insertion of eq . ( [ eq : separation - of - variable ansatz in geometry b ] ) into the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) followed by an integration over @xmath106 yields [ eq : separation ansatz fp for r geometry b ] @xmath209 \mathcal{r } \nonumber\\ & & + \big [ 2\omega r + ( 1-r)(1 - 5r ) \nonumber \\ & & \hphantom{aaaaaaa } -2\ , \zeta \alpha ( 1-r)(1 - 2r ) \big ] \frac{\partial\mathcal{r}}{\partial r } \nonumber\\ & & + ( 1-\zeta \alpha ) r(1-r)^{2 } \frac{\partial^{2}\mathcal{r}}{\partial r^{2 } } \label{eq : resulting ode for r if separation ansatz}\end{aligned}\ ] ] where the real constant @xmath210 is the ratio of two fourier expansion coefficients , @xmath211 a stationary solution to eq . ( [ eq : resulting ode for r if separation ansatz ] ) is @xmath212 } { \left . ^{-\zeta \alpha/(1-\zeta \alpha ) } \left(1-r\right)^{2/(1-\zeta \alpha)}}. \label{eq : solution for r if separation ansatz}\end{aligned}\ ] ] this solution reduces to the known results in the standard [ eq . ( [ eq : solution r for closed wire with absorption ( std ) ] ) ] and chiral [ eq . ( [ eq : solution r for closed wire with absorption ( chiral ) ] ) ] symmetry classes . in the standard symmetry class ( [ eq : strong+weak def of std limit ] ) the probability distribution of the phase is uniform @xmath213 and hence @xmath214 , whereas in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) it is a sum of the two delta functions @xmath215 , @xmath216 , @xmath217 . there remains considerable arbitrariness in the choice of @xmath218 because many different @xmath218 share the same @xmath210 . in practice we will choose the stationary solution ( [ eq : sol cross if geo c ] ) since we then recover the true solution in the semiopen geometry of fig . [ fig : geometries]a as @xmath219 . we postpone to sec . [ sec : validity of the separation ansatz for the crossover regime ] the discussion of the accuracy of the approximation implied by the separation - of - variable ansatz ( [ eq : separation ansatz fp for r geometry b ] ) with the choice ( [ eq : sol cross if geo c ] ) for the stationary probability distribution of @xmath106 . in geometry [ fig : geometries]c , an incoming plane wave from an ideal lead is partially reflected and partially transmitted by the disordered quantum wire . following abrikosov in ref . we trade the squared magnitude @xmath152 of the reflection coefficient @xmath1 for the resistance ( inverse of conductance ) @xmath220 \label{eq : def varrho}\end{aligned}\ ] ] under which @xmath221 obeys the fokker - planck equation [ eq : master equation in terms of varrho and phi ] @xmath222 where @xmath223 and @xmath224 in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) insertion of the separation - of - variable ansatz @xmath225 \mathcal{q}(\varrho;t)\end{aligned}\ ] ] into the fokker - planck equation ( [ eq : master equation in terms of varrho and phi ] ) yields [ eq : fp+initial geometry c chiral ] @xmath226 with the initial condition @xmath227 the normalized solution to eq . ( [ eq : fp+initial geometry c chiral ] ) has the integral representation [ eq : sol if chiral geo c ] @xmath228 with the auxiliary function and variable @xmath229 , \qquad t^{\prime}:=2t.\end{aligned}\ ] ] the integrations over @xmath230 and @xmath231 can be performed in closed form , @xmath232 . \nonumber\\ & & \end{aligned}\ ] ] by shifting the minimum of the resistance @xmath233 from 1 to 0 , @xmath234 , we recover the same asymptotics for the probability distribution of @xmath235 computed by stone and joanopoulos in ref . with a log - normal distribution of the nearest - neighbor hopping amplitudes of the discrete schrdinger equation when @xmath236 or @xmath237 . the change of variable @xmath238 yields the probability distribution @xmath239 \nonumber\\ & & \end{aligned}\ ] ] for the conductance @xmath7 in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) . the same result follows from solving the diffusion ( dmpk ) equation ( [ eq : fp chiral limit ] ) and performing the change of variables @xmath240.@xcite in the standard symmetry class ( [ eq : strong+weak def of std limit ] ) insertion of the separation - of - variable ansatz @xmath241 into the fokker - planck equation ( [ eq : master equation in terms of varrho and phi ] ) yields [ eq : abrikosov fp ] @xmath242 with the initial condition @xmath243 as shown by abrikosov in ref . , the normalized solution to eq . ( [ eq : abrikosov fp ] ) has the integral representation @xmath244 ^ 2/4 t } % \frac{1}{4 t } % \left [ % \ln % \frac{\varrho - u}{u(1-u ) } % \right]^{2 } % } } { [ u(1-u)(\varrho - u)]^{1/2 } } d_0(u;\varrho ) % \ln % \left [ % \frac{\varrho - u}{u(1-u ) } % \right ] \nonumber \\ & = & \sqrt{\frac{4}{\pi t^{3}}}\ , \int_{y^{\ } _ { \varrho}}^{+\infty } { d}y\ , \frac{\,y\ , e^{-(t/4)-(y^{2}/t ) } } { \big ( \cosh^{2}y-\varrho \big)^{1/2 } } \label{eq : sol if std geo c}\end{aligned}\ ] ] with @xmath245 $ ] and @xmath246.\ ] ] the change of variable @xmath238 yields the probability distribution @xmath247 with @xmath248 $ ] for the conductance @xmath7 in the standard symmetry class ( [ eq : strong+weak def of std limit ] ) . as was already the case in sec . [ subsubsec : crossover regime b ] we could not solve the fokker - planck equation ( [ eq : master equation in terms of varrho and phi ] ) exactly in the crossover regime . we thus try the separation - of - variable ansatz @xmath249 we are again assuming that the probability distribution @xmath218 is @xmath207 independent and we made the specific choice for @xmath218 given by eq . ( [ eq : sol cross if geo c ] ) . insertion of eq . ( [ eq : separation variable crossover ansatz geometry c ] ) into the fokker - planck equation ( [ eq : master equation in terms of varrho and phi ] ) followed by an integration over the phase @xmath106 yields [ eq : separation fp+initial cond crossover c ] @xmath250 with the initial condition @xmath251 the normalized solution to eq . ( [ eq : separation fp+initial cond crossover c ] ) has the integral representation [ eq : solution cross if separation var geo c ] @xmath252 \nonumber \\ & & \times \int_{0}^{1}{d}u\ , \frac{(\frac{1}{u}-1)^{\zeta\alpha/(1-\zeta\alpha ) } } { [ u(1-u)(\varrho - u)]^{1/2 } } % % u^{-\frac{1}{2}-\frac{\zeta \alpha}{1-\zeta \alpha } } % % ( 1-u)^{-\frac{1}{2}+\frac{\zeta \alpha}{1-\zeta \alpha } } % % ( \varrho - u)^{-\frac{1}{2 } } \nonumber \\ & & \times \int_{0}^{1}{d}v\ , \frac{e^{-d^{2}(u , v;\varrho)/4t^{\prime } } } { v^{1/2}(1-v)^{1/(1-\zeta \alpha ) } } d(u , v;\varrho ) \nonumber\\ & & \end{aligned}\ ] ] with the auxiliary function and variable @xmath229 , \quad t^{\prime}:=(1-\zeta \alpha)t . \hphantom{aa}\end{aligned}\ ] ] as it should be , we recover eq . ( [ eq : sol if chiral geo c ] ) when @xmath253 and eq . ( [ eq : sol if std geo c ] ) when @xmath254 . the approximate probability distribution @xmath255 for the conductance @xmath7 in the crossover regime is obtained from eq . ( [ eq : solution cross if separation var geo c ] ) through @xmath256 the dependence on @xmath207 for the mean and variance of the conductance in the crossover regime computed from eq . ( [ eq : prob dist g cross geo c ] ) are depicted in fig . [ fig : conductance ] . in the chiral symmetry class @xmath217 , @xmath257 , the mean and variance decay like @xmath258 . for @xmath259 and @xmath254 , the mean and variance decay exponentially with @xmath207 . ( a ) the mean ( a ) and variance ( b ) of the conductance computed from the probability distribution ( [ eq : prob dist g cross geo c ] ) as a function of @xmath160 for @xmath260 and @xmath217 . , width=283 ] ( b ) the mean ( a ) and variance ( b ) of the conductance computed from the probability distribution ( [ eq : prob dist g cross geo c ] ) as a function of @xmath160 for @xmath260 and @xmath217 . , we have constructed an approximate solution to the fokker - planck equations ( [ eq : master equation in terms of r and phi ] ) and ( [ eq : master equation in terms of varrho and phi ] ) in the crossover regime ( [ eq : crossover regime weak if a ] ) by assuming the separation of the squared magnitude @xmath152 from the phase @xmath106 of the reflection coefficient @xmath1 . in this section we shall investigate the accuracy of this approximation to the crossover regime ( [ eq : crossover regime weak if a ] ) for very large values of the length @xmath5 of the disordered quantum wire , i.e. , @xmath5 much larger than the localization length @xmath129 , which will be defined below . to this end , we shall compare the cumulants of the radial coordinate @xmath66 computed from the separation - of - variable ansatz with the same cumulants computed from a large - deviation ansatz introduced by schomerus and titov in refs . and . we begin from eq . ( [ eq : large x limit fp x phi all derivatives to right ] ) without absorption and in the crossover regime . we seek the asymptotic behavior when @xmath261 of the solutions to eq . ( [ eq : large x limit fp x phi all derivatives to right ] ) in the crossover regime . we try the separation - of - variable ansatz @xmath262 where @xmath218 is the stationary and normalized solution ( [ eq : sol cross if geo c ] ) . insertion of eq.([eq : separation of varaibale ansatz geo c large x crossover ] ) into eq . ( [ eq : large x limit fp x phi all derivatives to right ] ) with subsequent integration over @xmath106 yields @xmath263 equation ( [ eq : fp x if large t and crossover ] ) is nothing but the diffusion equation with a constant drift proportional to the amount of chiral symmetry breaking @xmath264 , i.e. , it is solved by @xmath265^{2 } } { \left ( 1-\zeta \alpha\right)t } \right ] } { \sqrt{\pi \left(1-\zeta \alpha\right ) t } } \label{eq : sol x if large t and crossover}\end{aligned}\ ] ] normalized to the real line @xmath266 . in the approximation ( [ eq : sol x if large t and crossover ] ) , the only nonvanishing cumulants are the first and second cumulants , [ eq : cumulants x in cross geo c ] @xmath267 provided @xmath268 ( i.e. , away from the chiral symmetry class ) , the random variable @xmath66 becomes self - averaging in the thermodynamic limit . it then makes sense to identify the localization length @xmath129 through @xmath269 which , together with eq . ( [ eq : first cumulant x cross geo c ] ) , gives @xmath270 in the standard symmetry class ( [ eq : strong+weak def of std limit ] ) @xmath254 and the localization length is twice the mean free path @xmath271 as is well known.@xcite in the chiral symmetry class ( [ eq : def chiral regime parameters ] ) @xmath253 and the localization length defined by eq . ( [ eq : def xi cross geo c ] ) is not a finite number anymore but a random variable as the right - hand side of eq . ( [ eq : first cumulant x cross geo c ] ) vanishes whereas the right - hand side of eq . ( [ eq : second cumulant x cross geo c ] ) remains finite . for the crossover regime ( [ eq : crossover regime weak if a ] ) , the constant @xmath210 implicitly depends on @xmath65 . the asymptotic behavior of @xmath210 for small @xmath272 can be evaluated as , with the probability distribution ( [ eq : sol cross if geo c ] ) , @xmath273 since a logarithmic divergence occurs both in the @xmath66 and @xmath274 integrals in the numerator as we take @xmath275 , while it occurs only in the @xmath274 integral in the denominator around @xmath276 . it is then natural to identify @xmath277 with the typical localization length@xcite @xmath278 when approaching the chiral symmetry class ( [ eq : def chiral regime parameters ] ) through the crossover regime ( [ eq : crossover regime weak if a ] ) . we have also computed the dependence of @xmath210 on @xmath65 by evaluating eq . ( [ eq : alpha vs varepsilon ] ) numerically . in fig . [ fig : alpha ] @xmath279 is plotted as a function of @xmath65 in the crossover regime ( [ eq : crossover regime weak if a ] ) . we observe that @xmath279 asymptotically behaves as @xmath280 in the vicinity ( @xmath281 ) of the chiral symmetry class ( [ eq : def chiral regime parameters ] ) . numerical evaluation of @xmath279 as a function of @xmath65 in the crossover regime ( [ eq : crossover regime weak if a ] ) . the integral @xmath210 defined in eq . ( [ eq : def alpha ] ) is evaluated numerically for the stationary distribution of the phase ( [ eq : sol cross if geo c ] ) . the line representing @xmath282 is a guide to the eyes . , width=283 ] schomerus and titov in refs . and devised a systematic method to compute all the cumulants of the conductance and the ldos to leading order in @xmath283 . in practice , they carried out explicitly the computation of the first four cumulants . to draw a connection to their work we note that whereas the parametrization @xmath284 of the magnitude of the reflection coefficient is the natural one from a geometrical point of view according to eqs . ( [ eq : polar decomposition s+m ] ) , ( [ eq : fp std limit ] ) , and ( [ eq : fp chiral limit ] ) , it is @xmath285 that behaves statistically as the logarithm of the conductance @xmath7 ( or the logarithm of the ldos @xmath6 according to the borland conjecture , see ref . ) in the regime defined by the condition @xmath286 we denote by @xmath287 the joint probability distribution of @xmath230 and @xmath106 , and introduce @xmath288 . we infer from eq . ( [ eq : large x limit fp x phi all derivatives to right ] ) that @xmath287 obeys the fokker - planck equation [ eq : large u limit fp u phi all derivatives to right ] @xmath289 where @xmath290 in the large @xmath230 limit and in the absence of absorption . the fact that @xmath291 can be written as a total derivative plays an essential role when solving , for any given purely imaginary @xmath292 and any given positive integer @xmath293 , the eigenvalue equation [ eq : large deviation ansatz ] @xmath294 f^{\ } _ k \label{eq : large deviation ansatz a}\end{aligned}\ ] ] for the eigenvalue @xmath295 and the normalized eigenfunction @xmath296 of @xmath106 with periodicity @xmath111 that results from insertion of the large - deviation ansatz @xmath297 into eq . ( [ eq : large u limit fp u phi all derivatives to right ] ) . for large @xmath298 , the integration over @xmath292 is dominated in eq . ( [ eq : large deviation ansatz b ] ) by the region of size @xmath299 close to the origin @xmath300 . assume that for a nonvanishing @xmath292 of order @xmath301 , all eigenvalues @xmath295 are real valued , of descending order , and that the largest eigenvalue @xmath302 is separated from @xmath303 by a finite gap . consequently , truncation of the summation over @xmath293 to the single term @xmath304 on the right - hand side of eq . ( [ eq : large deviation ansatz b ] ) produces an exponentially small error @xmath305 . \label{eq : large deviation ansatz c}\end{aligned}\ ] ] next , assume the expansions @xmath306 where the expansion coefficient @xmath307 encodes the @xmath308th cumulant of @xmath230 , @xmath309 insertion of the expansion ( [ eq : expansion mu s and f s ] ) into the eq . ( [ eq : large deviation ansatz a ] ) for @xmath304 can be solved iteratively for @xmath310 . for @xmath311 , eq . ( [ eq : large deviation ansatz a ] ) reduces to @xmath312 whose normalized solution is nothing but the stationary and normalized solution to eq . ( [ eq : fp if closed at one end and no absorption ] ) . for @xmath313 , eq . ( [ eq : large deviation ansatz a ] ) reduces to @xmath314 the expansion coefficient @xmath315 is obtained from integrating eq . ( [ eq : ode order n=1 ] ) over @xmath106 as a result of @xmath291 being a total derivative , [ eq : solving for 1 cumulant u ] @xmath316 where @xmath317 [ eq : computation alpha00 ] after comparison of eqs . ( [ eq : relating cumulants u and mun0 ] ) and ( [ eq : solving for 1 cumulant u ] ) on the one hand with eqs . ( [ eq : first cumulant x cross geo c ] ) and ( [ eq : def alpha ] ) on the other hand , we conclude that the first cumulant of the logarithm of the conductance is computed exactly to leading order in @xmath318 within the separation - of - variable ansatz for the crossover regime ( [ eq : crossover regime weak if a ] ) . as a corollary , the crossover ( [ eq : typical localization length in cross geo c ] ) for the typical localization length is given exactly by the separation - of - variable ansatz . carrying on the iteration to compute the expansion coefficients @xmath307 , however , we find that the separation - of - variable ansatz breaks down for @xmath319 as it does not agree anymore with the large - deviation ansatz from schomerus and titov in refs . and . ( 210,160)(-210,-120 ) ( -190 , 20)(1,0)190 ( -190,-10)(1,0)190 ( 0,-10)(0,1 ) 30 ( -190,-10)(-2,-1)10 ( -190 , 20)(-2 , 1)10 ( -135 , 5)(0 , 10 ) 180 ( -135 , 15 ) ( 1 , 0 ) 20 ( -135,-5 ) ( 1 , 0 ) 20 ( -128 , 5)@xmath320 ( -65 , 5 ) ( 0,10)-180 ( -65 , 15)(-1 , 0 ) 20 ( -65,-5)(-1 , 0 ) 20 ( -80 , 5)@xmath321 ( -100,-10)(0,1)30 ( -103,-20)@xmath274 ( -190,-30 ) ( 1,0)190 ( 0,-30)(-1,0)190 ( -105,-40)@xmath5 ( -210,35)@xmath0 ( -200 , -60)(1,0)200 ( -200,-90)(1,0)200 ( -200,-90)(0,1 ) 30 ( 0,-90)(0,1 ) 30 ( -135,-75)(0 , 10 ) 180 ( -135,-65 ) ( 1 , 0 ) 20 ( -135,-85 ) ( 1 , 0 ) 20 ( -128,-75)@xmath320 ( -65,-75 ) ( 0,10)-180 ( -65,-65)(-1 , 0 ) 20 ( -65,-85)(-1 , 0 ) 20 ( -80,-75)@xmath321 ( -100,-90)(0,1)30 ( -103,-100)@xmath274 ( -200,-110 ) ( 1,0)200 ( 0,-110)(-1,0)200 ( -105,-120)@xmath5 ( -210,-50)@xmath3 in this section , the probability distributions @xmath322 and @xmath323 of the ldos @xmath6 in the geometries of figs . [ fig : bc for ldos]a and [ fig : bc for ldos]b , respectively , are calculated based on the probability distribution function of the reflection coefficient @xmath192 obtained in sec . [ sec : solutions to the functional renormalization group equations ] . for any given realization of the disorder potentials @xmath97 , the ldos @xmath324 is defined by @xcite @xmath325 where the matrix elements @xmath326 in real space of the green function are obtained from solving @xmath327 g(y , z)= \delta(y - z)\sigma^{\ } _ { 0 } , \nonumber\\ & & e=\mathcal{e}^{\prime}+{i}\mathcal{e}^{\prime\prime } , \qquad \mathcal{e}^{\prime\prime}\geq0 , \nonumber\\ & & ( \eta:\hbox { positive infinitesimal } ) \label{eq : def green fct}\end{aligned}\ ] ] with the boundary conditions associated to the geometries in fig . [ fig : bc for ldos ] . as long as @xmath274 is far away from the ends of the disordered quantum wire , say in the middle , the statistical properties of @xmath324 should be independent of @xmath274 . however , the statistical properties of @xmath324 depend strongly on the boundary conditions at the ends of the disordered quantum wire.@xcite for completeness , we will understand under the chiral limit of the ldos ( [ eq : def meso ldos ] ) the property that the green function ( [ eq : def green fct ] ) anticommutes with @xmath26 , i.e. , @xmath328 whereas we will understand under the standard limit of the ldos ( [ eq : def meso ldos ] ) the conditions on the green function ( [ eq : def green fct ] ) that [ eq : def std limit from green fct ] @xmath329 or @xmath330 or @xmath331 schomerus _ et al . . expressed the probability distribution of the ldos in terms of the probability distributions of the reflection coefficient @xmath320 ( @xmath321 ) for outgoing plane waves reflected from the region to the left ( right ) of @xmath274 for the continuous nonrelativistic schrdinger equation in the standard symmetry class ( [ eq : def std limit from green fct ] ) . here we extend their analysis to the continuous relativistic hamiltonian ( [ eq : def our anderson localization problem a ] ) . a crucial observation in ref . is that , for weak disorder , the green function can be expanded in terms of free scattering states within a small interval of order of the mean free path @xmath61 since one can assume that there are no impurities in this interval . the solution to eq . ( [ eq : def green fct ] ) for @xmath274 and @xmath292 belonging to such an impurity - free interval can be written as a linear combination of @xmath332 , @xmath333 , @xmath334 , @xmath335 , except for the discontinuity at @xmath336 , @xmath337 where @xmath338 , @xmath339 , @xmath340 is the heaviside function that vanishes when @xmath341 and equals one when @xmath342 , and we have momentarily reinstated the fermi velocity @xmath9 and @xmath10 in the relativistic dispersion relation @xmath343 . the relative amplitudes among @xmath344 as well as the factor @xmath345 are determined by taking into account multiple scattering from the left and right boundaries of an interval free of impurities . combining eqs . ( [ eq : def green fct ] ) and ( [ eq : plane wave expansion green ] ) reproduces in the relativistic limit the mesoscopic relation @xmath346 between the ldos at @xmath274 and the reflection coefficients @xmath320 and @xmath321 from the disordered region to the left and right of @xmath274 , respectively , derived by schomerus _ et al . _ in ref . for the nonrelativistic continuous schrdinger equation . in the nonrelativistic case , the diagonal matrix element of the green function in real space gives the microscopic local density of states . the microscopic local density of states exhibits @xmath347 oscillations.@xcite the ldos ( [ eq : ldos ] ) is obtained after smearing out these oscillations . the effect of the relativistic approximation ( [ eq : def our anderson localization problem a ] ) is to remove the @xmath347 oscillations from the outset . henceforth let us measure the ldos @xmath324 in units of @xmath348 , the dos of a clean wire . the probability distribution function @xmath349 of the ldos @xmath324 defined by eq . ( [ eq : ldos ] ) is now simply given by @xmath350 where @xmath351 ( @xmath352 ) is the length of the segment of the disordered quantum wire to the left ( right ) of the point @xmath274 where the ldos is measured . the total length of the disordered quantum wire is evidently given by ( see fig . [ fig : bc for ldos ] ) @xmath353 the dependence of the probability distribution for the ldos on the boundary conditions enters through the dependence of the probability distributions @xmath354 and @xmath355 on the boundary conditions corresponding to the geometries of fig . [ fig : geometries].@xcite this dependence on boundary conditions is implied by the dependence on @xmath356 or @xmath357 of the probability distribution of the two reflection coefficients @xmath320 and @xmath321 , respectively . when a disordered quantum wire of finite length is closed , one must broaden the single - particle energy levels beyond the mean level spacing to make sense of the energy dependence of the mesoscopic ldos . this is achieved here by introducing a finite absorption @xmath358 , in which case the dependence on @xmath359 or @xmath360 becomes immaterial . the absorption @xmath358 is switched off when a disordered quantum wire of finite length @xmath5 is connected to a perfect lead . we consider first the geometry depicted in fig . [ fig : bc for ldos]a and then the geometry depicted in fig . [ fig : bc for ldos]b . ( a ) ( a ) three traces for the probability distribution of the ldos are plotted with @xmath361 ( from the bottom up at @xmath362 ) so as to interpolate between the chiral and standard symmetry classes ( [ eq : def chiral limit from green fct ] ) and ( [ eq : def std limit from green fct ] ) , respectively . each trace is obtained from numerical integration of eq . ( [ eq : distribution for the ldos ] ) for the semi - open geometry of fig . [ fig : bc for ldos]a with @xmath363 . ( b ) three traces for the probability distribution of the ldos are plotted with @xmath361 ( from the bottom up at @xmath364 ) so as to interpolate between the chiral and standard symmetry classes ( [ eq : def chiral limit from green fct ] ) and ( [ eq : def std limit from green fct ] ) , respectively . each trace is obtained from numerical integration of eq . ( [ eq : distribution for the ldos ] ) for the closed geometry of fig . [ fig : bc for ldos]b with @xmath365 . we have verified that numerical integration of eq . ( [ eq : distribution for the ldos ] ) agrees with a representation of eq . ( [ eq : distribution for the ldos ] ) in terms of elementary functions when possible . , width=283 ] ( b ) ( a ) three traces for the probability distribution of the ldos are plotted with @xmath361 ( from the bottom up at @xmath362 ) so as to interpolate between the chiral and standard symmetry classes ( [ eq : def chiral limit from green fct ] ) and ( [ eq : def std limit from green fct ] ) , respectively . each trace is obtained from numerical integration of eq . ( [ eq : distribution for the ldos ] ) for the semi - open geometry of fig . [ fig : bc for ldos]a with @xmath363 . ( b ) three traces for the probability distribution of the ldos are plotted with @xmath361 ( from the bottom up at @xmath364 ) so as to interpolate between the chiral and standard symmetry classes ( [ eq : def chiral limit from green fct ] ) and ( [ eq : def std limit from green fct ] ) , respectively . each trace is obtained from numerical integration of eq . ( [ eq : distribution for the ldos ] ) for the closed geometry of fig . [ fig : bc for ldos]b with @xmath365 . we have verified that numerical integration of eq . ( [ eq : distribution for the ldos ] ) agrees with a representation of eq . ( [ eq : distribution for the ldos ] ) in terms of elementary functions when possible . , width=283 ] to calculate the probability distribution of the ldos in a disordered quantum wire connected to a reservoir on the left and closed on the right , we need ( i ) the joint probability distribution @xmath366 for the square @xmath367 of the magnitude of the reflection coefficient @xmath320 and its phase @xmath368 on the segment of length @xmath351 to the left of the point at which the ldos is measured and ( ii ) the probability distribution @xmath369 of the phase @xmath370 of the reflection coefficient @xmath321 on the segment of length @xmath352 to the right of the point at which the ldos is measured . the calculation of @xmath369 is outlined in sec.[subsec : disordered wire closed on the right - hand side without absorption ] . the calculation of @xmath366 is outlined in sec.[sec : disordered wire opened at both ends without absorption ] . the probability distribution of the ldos in the chiral symmetry class ( [ eq : def chiral limit from green fct ] ) is ( see the appendix [ app sec : proof ] ) [ eq : log normal distribution ldos if ch in geo semi - open ] @xmath371 this is the probability distribution of a log - normal distributed random variable with @xmath372 the average ldos is @xmath373 which grows exponentially with @xmath374 . for comparison , the probability distribution of the ldos in the standard symmetry class ( [ eq : def std limit from green fct ] ) , which was obtained in refs . and , is also log - normal , [ eq : log normal distribution ldos if std in geo semi - open ] @xmath375 , \nonumber\\ & & \\ & & \left\langle \ln\nu \right\rangle^{\ } _ { l^{\ } _ { \mathrm{l}}}= -\frac{l^{\ } _ { \mathrm{l}}}{\ell } , \\ & & \left\langle \left ( \ln\nu + \frac{l^{\ } _ { \mathrm{l}}}{\ell } \right)^2 \right\rangle^{\ } _ { l^{\ } _ { \mathrm{l } } } = \frac{2l^{\ } _ { \mathrm{l}}}{\ell}.\end{aligned}\ ] ] incidentally , the average ldos is not affected by the disorder in the standard symmetry class , @xmath376 . the probability distributions for the ldos in the chiral and standard symmetry classes are depicted in fig . [ fig : distribution of the ldos]a . in the thermodynamic limit @xmath377 , the logarithm of the ldos is self - averaging in the standard symmetry class whereas this is not the case in the chiral symmetry class . this difference could have been anticipated from the identification @xmath378 , valid when @xmath379 , with the radial coordinate @xmath66 that obeys the fokker - planck equations ( [ eq : fp std limit ] ) and ( [ eq : fp chiral limit ] ) , respectively . numerical evaluation of the integrals on the right hand side of eq . ( [ eq : distribution for the ldos ] ) is depicted in fig . [ fig : distribution of the ldos]a for the crossover regime ( [ eq : crossover regime weak if a ] ) within our approximation which is encoded by the separation - of - variable ansatz ( [ eq : separation variable crossover ansatz geometry c ] ) , ( [ eq : separation fp+initial cond crossover c ] ) , ( [ eq : solution cross if separation var geo c ] ) , and ( [ eq : sol cross if geo c ] ) . to calculate the probability distribution of the ldos in a disordered quantum closed at both ends with a finite absorption @xmath380 , we need the joint probability distributions @xmath381 and @xmath382 . we can safely ignore the dependence on @xmath383 in the regime @xmath384 as is done in sec . [ subsec : disordered quantum wire closed on the right - hand side with absorption ] . the probability distribution of the ldos in the chiral symmetry class ( [ eq : def chiral limit from green fct ] ) is@xcite ( see appendix [ app sec : proof ] ) [ eq : distribution ldos if ch in geo closed ] @xmath385^{2 } \frac{1}{\nu } \exp\left [ -\frac{\omega(1+\nu^{2})}{2\nu } \right ] k^{\ } _ { 0 } \left ( \frac{\omega|1-\nu^{2}|}{2\nu } \right ) \label{eq : distribution ldos if ch in geo closed a}\end{aligned}\ ] ] where @xmath386 , @xmath310 are the modified bessel functions and has the asymptotics @xmath387 for comparison , the probability distribution of the ldos in the standard symmetry class ( [ eq : def std limit from green fct ] ) , which was obtained in refs . and , @xmath388 } { \sqrt{u-(1+\nu^2)/(2\nu ) } } \left [ u k^{\ } _ 0\left(\omega\sqrt{u^2 - 1}\right ) + \sqrt{u^2 - 1 } k^{\ } _ 1\left(\omega\sqrt{u^2 - 1}\right ) \right],\end{aligned}\ ] ] has the asymptotics @xmath389 the probability distributions for the ldos in the chiral and standard symmetry classes are depicted in fig . [ fig : distribution of the ldos]b . spectral weight from the tails @xmath390 and @xmath391 of the probability distribution in the standard symmetry class is redistributed around @xmath392 in the chiral symmetry class . consequently , in the limit @xmath219 , the mean value of the mesoscopic ldos in the chiral symmetry class becomes @xmath393 ^ 2 } \int^\infty_1{d}\nu\left(1+\frac{1}{\nu^2}\right ) \nonumber \\ & & \times \exp\left[-\frac{\omega}{2}\left(\nu+\frac{1}{\nu}\right)\right ] k_0\left(\frac{\omega}{2}\left(\nu-\frac{1}{\nu}\right)\right ) \nonumber \\ & { \buildrel \omega\to0 \over = } & \frac{1}{(\omega/2)[\ln(\omega/2)]^2}.\end{aligned}\ ] ] this result is identical to the smeared dyson singularity of the dos in the chiral symmetry class , @xmath394 ^ 2}.\end{aligned}\ ] ] numerical evaluation of the integrals on the right hand side of eq . ( [ eq : distribution for the ldos ] ) is depicted in fig . [ fig : distribution of the ldos]b for the crossover regime ( [ eq : crossover regime weak if a ] ) within our approximation which is encoded by the separation - of - variable ansatz ( [ eq : separation - of - variable ansatz in geometry b ] ) , ( [ eq : separation ansatz fp for r geometry b ] ) and ( [ eq : sol cross if geo c ] ) . the probability distribution of the mesoscopic local density of states ( ldos ) @xmath6 for a strictly one - dimensional problem of anderson localization with chiral symmetry ( the chiral symmetry class ) was computed in closed form for two simply connected geometries assuming a weak disorder . as is the case when the chiral symmetry is maximally broken ( the standard symmetry class ) , the probability distribution of @xmath6 strongly depends on boundary conditions . in a semiopen geometry we found that the probability distribution of @xmath6 is log - normal with a vanishing mean as opposed to log - normal with a finite mean in the standard symmetry class . in a closed geometry with absorption we found that the probability distribution of @xmath6 has a double - peak structure whereby the second peak turns out to be a logarithmic singularity at @xmath392 . we verified that the smeared dyson singularity is reproduced by the mean value of the ldos when the absorption is sufficiently small . furthermore , we found one exact and proposed two approximate solutions to the functional renormalization group equation obeyed by the joint probability distribution for the squared modulus and the phase of the reflection amplitude of a finite and possibly dissipative wire with semiopen and open boundary conditions , respectively , from which we could extract the approximate crossover of the ldos @xmath6 and conductance @xmath7 between the chiral and standard symmetry classes . we could show that our approximation is exact for the first cumulant but fails to describe higher cumulants of the logarithms of @xmath6 and @xmath7 in the crossover regime . we are indebted to p. w. brouwer for an initial collaboration on this project and for his generous feedback at later stages . we would also like to thank h. schomerus and m. titov for useful comments . this work was supported in part by japan society for the promotion of science ( s.r . ) . in the appendix we prove eqs . ( [ eq : prob dis ldos chiral semi - open ] ) and ( [ eq : distribution ldos if ch in geo closed a ] ) . we make use of the fact that the probability distributions for the squared modulus and phases of the reflection coefficients @xmath395 factorize in the chiral symmetry class . we have chosen here to represent @xmath396 by @xmath397 . we thus need to compute t. nagao and k. slevin , j. math . phys . * 34 * , 2075 ( 1993 ) ; j. j. m. verbaarschot and i. zahed , phys . lett . * 70 * , 3852 ( 1993 ) ; s. hikami and a. zee , nucl . b * 408 * , 415 ( 1993 ) ; a. v. andreev , b. d. simons , and n. taniguchi , nucl . b * 432 * , 487 ( 1994 ) . l. i. deych , d. zaslavsky , and a. a. lisyansky , phys . lett . * 81 * , 5390 ( 1998 ) ; l. i. deych , a. a. lisyansky , and b. l. altshuler , _ ibid . _ * 84 * , 2678 ( 2000 ) ; l. i. deych , m. v. erementchouk , a. a. lisyansky , and b. l. altshuler , _ ibid . _ * 91 * , 096601 ( 2003 ) . an alternative derivation of the fokker - planck equation ( [ eq : master equation in terms of r and phi ] ) is the following . first , a discrete langevin process with the discrete time step `` @xmath73 '' is written down by expanding @xmath409 in eq . ( [ eq : composition law reflection amplitude ] ) up to second order in @xmath97 . second , a continuous fokker - planck equation for the probability distribution @xmath410 is derived from the discrete langevin process using the methods of appendix a3.5 of ref . , say . here , @xmath411 denotes averaging over @xmath97 restricted to the thin slice of length @xmath73 .

the 7 tev and 8 tev run at the large hadron collider ( lhc ) have revealed and confirmed the existence of a higgs - like particle @xmath0 in the standard model ( sm ) with mass in the narrow range of 125 - 126 @xmath1 @xcite . this discovery is also verified by the recent tevatron final results @xcite . the observation of this new particle combines evidence in the decays @xmath2 , @xmath3 and @xmath4 . different signal strengths , defined as the product of higgs boson production cross sections from different channels and the branching ratios for different decay modes normalized to the corresponding products in sm , have been measured with good precision by both experiments at atlas and cms @xcite . these measurements will be further improved in the future 13 - 14 tev run at the lhc , and perhaps at a future international linear collider ( ilc ) should this machine ever be built . from these signal strengths measurement one can extract information on the couplings of this higgs - like particle to the gauge bosons and sm fermions . from the most recent measurements , extraction of this higgs - like particle couplings to sm particles seem to be consistent to a great extent with those of the sm higgs boson couplings @xcite . moreover , data collected both at atlas and cms indicate that this higgs - like particle has zero spin and is cp - even , _ i.e. _ @xmath5 is preferred @xcite . the discovery of the higgs - like particle at the lhc indicates for the first time that fundamental scalar exists in nature . certainly , many phenomenological models that extend the sm scalar sector with just one scalar doublet existed already in the literature . some of them are motivated by physics of the dark matter ( dm ) or neutrinos masses . among these extensions , we have models with multiple higgs doublets , with one higgs doublet and multiple singlets or triplets etc . all these extensions should have one light scalar with higgs - like couplings to sm particles in the range tolerated by signal strength measurements . indeed many studies ( see for example the references in @xcite ) have been done using these data to constrain various extensions of the scalar sector of the sm . in this paper , we concentrate on the inert higgs doublet model ( ihdm ) which is a very simple extension of the sm . it was first proposed by deshpande and ma @xcite in order to study the pattern of electroweak symmetry breaking . the ihdm is an attractive model due to its simplicity . it is basically a two higgs doublet model ( thdm ) ( see @xcite for a recent overview ) with an imposed exact @xmath6 symmetry . under the @xmath6 symmetry , all the sm particles are even representing the visible sector , while the new higgs doublet field is odd representing the inert dark sector . imposing the @xmath6 symmetry forbids the second higgs doublet developing a vacuum expectation value ( vev ) and all the inert particles in this doublet can only appear in pair in their interaction vertices . indeed , recent studies @xcite of global fits of the lhc data suggest that the couplings between the @xmath7 and @xmath8 gauge bosons with the new 125 - 126 gev higgs - like boson are very close to their sm values . the new 125 - 126 gev boson may play the entire role of electroweak symmetry breaking ( ewsb ) and leave no room for other higgs fields to develop any vevs . this favors the ihdm . as a result , ihdm exhibits very interesting phenomenology . it predicts the existence of a neutral scalar field , denoted generically by @xmath9 here , which is the lightest odd particle ( lop ) in this model and will play the role of dm candidate . the higgs mechanism provides a portal for communication between the inert dark sector and the visible sm sector . thus if kinematics allowed , the sm higgs boson may decay into a pair of dm @xmath9 and will contribute to the invisible sm higgs boson width which is now constrained by the lhc data . moreover , annihilation of @xmath9 into sm particles will provide thermal relic density and the scattering of @xmath9 onto nucleons will lead to direct detection signatures . therefore , ihdm could be considered as a simple but competitive model in the market with a weakly interacting massive particle ( wimp ) . as we will see later , ihdm could predict correct dm relic density as well as a cross section for scattering of @xmath9 onto nucleons that is consistent with existing data from direct detection . almost three decades later , ihdm was extended further by ma @xcite to include three @xmath6 odd weak singlets of right - handed neutrinos with majorana masses . in this extended model @xcite , a radiative seesaw mechanism for light neutrino masses was proposed and either @xmath9 or one of the right - handed neutrinos could be dm candidate . we will not consider this extended version of ihdm in this work but would like to return to this in the future @xcite . as mentioned earlier , there have been many attempts to introduce dm higgs models by extending the sm scalar sector with more singlets or doublets @xcite . in particular , the phenomenology of ihdm had been extensively discussed in the context of dm phenomenology @xcite and also for collider phenomenology @xcite . ihdm has been also advocated to explain the naturalness problem @xcite . in the present study , we will reconsider the ihdm model in light of the recent atlas and cms discovery of a higgs - like particle of 125 - 126 gev . we assume that the lop must fulfill the recent relic density measurement by planck @xcite . as a good dm candidate test , we also consider the constraints from dm direct and indirect detection . for the constraint from dm direct detection search , we study the impact from the most recent lux upper limit @xcite which provides a robust constraint on the parameter space . as for indirect detection , we will take into account the fermi - lat @xmath10-ray observations of the dwarf spheroidal galaxies ( dsphs ) @xcite and the galactic center ( gc ) @xcite . in addition to @xmath10-rays , we also include constraints from cosmic ray electrons / positrons from ams-02 @xcite , pamela @xcite , and fermi - lat @xcite , and cosmic ray anti - protons from pamela @xcite . these constraints will be also supplemented by the lhc constraints such as monojet and diphoton signal strength measurement as well as constraint on the higgs boson invisible decay width . some of the above aspects for ihdm have been discussed in recent studies @xcite . the compatibility of a heavy sm higgs boson with lhc results and xenon100 data @xcite were discussed in ref . @xcite . similar issues for ihdm were discussed in ref . @xcite with the inclusion of radiative corrections to the scalar masses of the model . @xcite also included renormalization group effects for the quartic scalar couplings @xmath11 in order to evaluate vacuum stability , perturbativity and unitarity constraints at a higher scale . in our analysis , we will go further by including also the following aspects : * larger parameter space for dm mass : we will scan @xmath12 from 5 gev to 4 tev . * lhc monojet constraint in the likelihood . * accurate dm indirect detection likelihood . * constraints from the first result of lux in direct detection likelihood . * future sensitivity to monojet search at lhc with 14 tev at the planned luminosity of 100 fb@xmath13 and 300 fb@xmath13 . * sensitivity to ams-02 anti - protons and xenon1 t . the layout of this paper is as follows . in section 2 , we briefly review ihdm and its parameterization . we then list the theoretical constraints such as perturbativity , perturbative unitarity and vacuum stability that must be satisfied by the scalar potential parameters . the constraints from collider searches that ihdm is subjected to are discussed in section 3 . these include : electroweak precision test constraints , @xmath7 and @xmath8 width constraints , negative search for charginos and neutralinos from lep - ii that could restrict the inert higgs bosons masses , diphoton signal strength measurement as well as monojet constraint from dm search at lhc . in section 4 , we will discuss the relic density measurement by planck as well as dm direct detection and indirect detection constraints . in section 5 , we present our methodology for likelihood analysis and explain how all the constraints are included . we present our numerical results in section 6 . future experimental constraints from lhc-14 , xenon1 t and ams-02 are discussed in section 7 . we conclude in section 8 . in this section , we briefly review the salient features of ihdm and discuss some existing theoretical constraints . the ihdm @xcite is a rather simple extension of the sm higgs sector . it contains the sm higgs doublet @xmath14 and an additional higgs doublet @xmath15 . this model has a @xmath6 symmetry under which all the sm fields including @xmath14 are even while @xmath15 is odd under @xmath6 : @xmath16 . we further assume that @xmath6 symmetry is not spontaneously broken _ i.e. _ @xmath15 field does not develop vev . these doublets can be parameterized as : h_1 = ( c g^+ + ( v + h + i g^0 ) ) , h_2 = ( c h^+ + ( s + i a ) ) where @xmath17 and @xmath18 are the charged and neutral goldstone bosons respectively , which will be absorbed by the @xmath19 and @xmath8 to acquire their masses . the scalar potential with an exact @xmath6 symmetry forbids the mass term @xmath20 which mixes @xmath14 and @xmath15 . thus it has one fewer term than in thdm , _ i.e. _ @xmath21 the electroweak gauge symmetry is broken when @xmath14 doublet gets its vev : @xmath22 while @xmath23 . this pattern of symmetry breaking ensures unbroken @xmath6 symmetry and results in one more cp - even neutral scalar @xmath24 , one cp - odd neutral scalar @xmath25 , a pair of charged scalars @xmath26 and @xmath27 in addition to the sm cp - even scalar higgs @xmath0 . note that since @xmath0 is the sm higgs boson , it is @xmath6 even , while @xmath24 , @xmath25 and @xmath28 are @xmath6 odd . moreover , the exact @xmath6 symmetry naturally imposes the flavor conservation . only sm higgs boson couples to sm fermions while the inert higgses @xmath24 , @xmath25 and @xmath28 do not . the @xmath6 symmetry also ensures the stability of the lightest scalar ( @xmath24 or @xmath25 ) that can act as a dm candidate . dm phenomenology of ihdm had been studied extensively in the literature @xcite . the above scalar potential in eq . ( [ potential ] ) has 8 real parameters : 5 @xmath11 , 2 @xmath29 and the vev @xmath30 . minimization condition for the scalar potential eliminates @xmath31 in favour of the higgs mass and the vev @xmath30 is fixed to be 246 @xmath1 by the weak gauge boson masses . we are left with 6 independent real parameters . the masses of all the four physical scalars can be written in terms of @xmath32 and @xmath33 as the following @xmath34 where _ l , a=(_3 + _ 4 _ 5 ) . four of the five quartic couplings can be written in terms of physical scalar masses and @xmath35 as the following expressions _ 1 & = & , _ 3 = ( m_h^^2 - _ 2 ^ 2 ) , + _ 4 & = & , _ 5 = . [ lambds ] we are then free to take @xmath36 and @xmath35 as 6 independent parameters , or equivalently , the following set \{m_h , m_s , m_a , m_h^ , _ 2 , _ l } [ eq : input_par_2 ] which is more convenient for our purposes to describe the full scalar sector . in our tree level parameterization , @xmath37 can be expressed as _ a = _ l - _ 5 = _ [ eq : lambda_a ] it is clear from eq . ( [ eq : lambda_a ] ) that @xmath38 for @xmath39 and @xmath40 for @xmath41 . in our systematic scan in the following numerical work , we will consider both cases where @xmath42 or @xmath43 being the lop . thus , the dm mass is defined as m_=m_s , m_a . in order to illustrate constraint on @xmath44 with @xmath24 or @xmath25 being the lop , we define the coupling @xmath45 as @xmath46 the coupling @xmath47 shows up directly in the relic density computation depends on whether @xmath48 or @xmath49 . the parameters of the scalar potential of the ihdm are severely constrained by theoretical constraints . first , to trust our perturbative calculations we have to require all quartic couplings in the scalar potential of eq . ( [ potential ] ) to obey @xmath50 . second , in order to have a scalar potential bounded from below we must also demand the following constraints @xcite : @xmath51 third , to further constrain the scalar potential parameters of the ihdm one can impose tree - level unitarity in a variety of scattering processes among the various scalars and gauge bosons . for the unitarity constraints , it is convenience to define the following twelve parameters @xmath52 @xcite : @xmath53 the perturbative unitarity constraints are then imposed on all @xmath52 satisfying @xcite the strongest constraints on @xmath54 when @xmath55 , which translate into @xmath56 in fact , from eq . ( [ mh - lambda1 ] ) with @xmath57 gev and @xmath58 gev we have @xmath59 which is well below the unitarity bound given by eq . ( [ eq : lam2cons ] ) . in this section , we discuss dm constraints from the collider search experiments . we will focus on the constraints from the electroweak precision test ( ewpt ) experiments at lep - ii , neutral and charged higgs search at lep - ii , as well as the mass and the invisible width of the higgs , diphoton signal strength and monojet search from the lhc . * ewpt is a common approach to constrain physics beyond sm by using the global electroweak fit through the oblique @xmath24 , @xmath60 and @xmath61 parameters @xcite . it is well known that in the sm the ewpt implies a close relation between the three masses @xmath62 , @xmath63 and @xmath64 . similarly , in the ihdm , the ewpt implies constraints on the mass splitting among the higgs boson masses @xcite . in this study , we will use the pdg values of @xmath24 and @xmath60 with @xmath61 fixed to be zero @xcite . we allow @xmath24 and @xmath60 parameters to be within 95% c.l . ( confidence level ) . the central value of @xmath24 and @xmath60 , assuming a sm higgs boson mass of @xmath57 gev , are given by @xcite : s = 0.05 0.09 , t = 0.08 0.07 . the correlation between @xmath24 and @xmath60 is 91% in this fit . analytic expressions for @xmath24 and @xmath60 in ihdm can be found in ref . * other lep constraints come from the precise measurements of @xmath7 and @xmath8 widths . in order not to affect these decay widths we demand that the channels @xmath65 and/or @xmath66 are kinematically not open . this leads to the following constraints : @xmath67 , @xmath68 and @xmath69 @xcite . + additional constraints on the charged higgs boson @xmath28 , cp - even @xmath24 and cp - odd @xmath25 masses can be derived . note that lep , tevatron and lhc bounds on @xmath28 and @xmath25 can not apply because the standard search channels assumes that those scalars decays into a pair of fermions which are absent in the ihdm due to @xmath6 symmetry . + in the ihdm , if @xmath24 is the lop the cp - odd @xmath25 could decay like @xmath70 , while the charged higgs boson @xmath28 could decay into @xmath71 and/or @xmath72 followed by @xmath70 . therefore the final states of the two production processes @xmath73 and @xmath74 would be multi - leptons or multi - jets , depending on the decay products of @xmath19 and @xmath8 , plus missing energies . to certain extents , the signatures for the charged higgs case would be similar to the supersymmetry searches for charginos and neutralinos at @xmath75 or at hadron colliders @xcite . and @xmath76 in the ihdm case @xcite . the cross sections for fermionic and scalar pair production are scaled by @xmath77 and @xmath78 respectively , where @xmath79 is the velocity of the final state particle in the center - of - mass frame . hence , the scalar pair will be suppressed by an extra factor of @xmath79 as compared with the fermionic case . ] taking into account these considerations , we will safely choose in our scan for the charged higgs mass @xmath80 being always greater than @xmath81 . for the neutral inert higgses @xmath24 and @xmath25 , neutralinos search at lep - ii via @xmath82 followed by @xmath83 @xcite could apply here since the process @xmath74 followed by the cascade @xmath84 would give similar signals . such analysis had been carefully done in ref . their limits on @xmath85 and @xmath86 can be summarized as @xmath87 @xmath88 . however , in the present study , we will use the exact exclusion region as given in ref . * : + in the ihdm , the sm higgs boson @xmath0 have similar couplings to sm fermions and gauge bosons . therefore , as long as @xmath0 decays into sm final states , all the measurements from atlas and cms experiments about sm higgs boson properties can be used . in particular , we will require the mass of the sm higgs boson of the ihdm should lie within the measurement @xcite : @xmath89 * the openings of one of the non - standard decays of the higgs boson such as @xmath90 or @xmath91 , 70 gev in our numerical scan ] hence @xmath92 is not open . can modify the total width of the higgs boson and can have significant impact on lhc results . since either @xmath42 or @xmath25 is the lightest @xmath6 odd particle , it will be stable and the decay @xmath93 will be invisible . both atlas and cms had performed searches for invisible decay of the higgs boson @xcite . using the higgs - strahlung sm cross section for @xmath94 with a 125 gev sm higgs boson , atlas @xcite has excluded an invisible branching ratio of the higgs boson larger than 65% with 95% c.l .. cms also studied the invisible decay of the higgs boson produced via the vector boson fusion ( vbf ) mechanism and obtained an upper limit for the invisible branching ratio of 69% with 95% c.l .. when the two production mechanisms are combined the upper limit becomes 54% with 95% c.l . this constraint on the invisible decay is rather weak compared to the one derived from various works of global fits to atlas and cms data @xcite . these global fits studies suggest that the branching ratio of the invisible decay of the higgs boson should not exceed 19% at 95% c.l . in the case where the higgs boson has sm - like couplings to all sm particles plus additional invisible decay mode which is exactly the case as in ihdm . on the other hand , if one allows for deviation in the @xmath95 ( and @xmath96 as well on general grounds but not for ihdm ) coupling from its sm value the 95% c.l . limit on the invisible higgs decay branching ratio moves up to 29% @xcite . * assuming that the production cross section of the higgs boson is dominated by the gluon gluon fusion process , the diphoton signal strength in the ihdm normalized to the sm value can be simplified as @xmath97 where in the first line we have used the narrow width approximation and in the second line we used the fact that @xmath98 is the same in both the sm and ihdm . thus the signal strength @xmath99 in ihdm is simply given by the ratio of the branching ratios , which is not necessarily one since the charged higgs boson in ihdm can provide extra contribution other than the sm particles to the triangle loop amplitude of @xmath100 . + at atlas , the overall signal strength for diphoton is about @xmath101 , which corresponds to about 2@xmath102 deviation from the sm prediction @xcite , while the other channels are consistent with sm . however , at cms , the new analysis for diphoton mode based on multivariate analysis @xcite gives a signal strength about @xmath103 , which is consistent with sm . many proposals based on physics beyond sm , including ihdm , have been suggested to explain the diphoton excess , but the actual disagreement between atlas and cms does not allow to draw any definite conclusions yet , given the current level of statistics . in the present analysis we will not try to explain the diphoton excess but rather study the impact of the other constraints on the ratio @xmath99 . * besides using the invisible width of the higgs decay , another strategy to look for dm at the lhc is to study high @xmath104 monojet balanced by a large missing transverse energy @xmath105 @xcite . such kind of signature is possible in ihdm by producing the sm higgs boson @xmath0 in association with an energetic jet followed by the invisible decay of @xmath0 . in our analysis we will consider the following parton processes : * * @xmath106 : @xmath107-channel and @xmath108-channel tree level diagrams with the higgs boson radiated from @xmath109 quark legs , * * @xmath110 : @xmath108-channel diagram through tree level gluon - quark - anti - quark vertex and one - loop @xmath96 effective vertex , * * @xmath111 : @xmath108-channel diagram through tree level three gluon vertex and one - loop @xmath96 effective vertex , * * @xmath112 : @xmath107-channel diagram through tree level gluon - quark - anti - quark vertex and one - loop @xmath96 effective vertex . + in all these processes , the final state consists of a pair of invisible dm particles plus a quark or gluon jet . for the experimental cuts , see the later discussion of the likelihood function for the monojet data in section [ section:5 ] . it is well known that annihilation of @xmath9 into sm particles and other inert higgs bosons can contribute to thermal relic density as well as indirect dm signals of high energy gamma - rays , positrons , antiprotons or neutrinos , while the scattering of @xmath9 onto nuclei will lead to direct detection signals by measuring the recoil energy of the nuclei via scintillation light , heat or ionization or some combinations of these three different signals using different technologies . * assuming a standard thermal evolution of our universe , we compute the relic density from the following channels : @xmath113 ( @xmath114 ) , @xmath115 and @xmath116 . since @xmath9 can be either @xmath24 or @xmath25 , we consider @xmath117 or @xmath118 annihilation . + note that in the case where @xmath119 , we take into account the annihilation into 3-body final state from @xmath120 or 4-body final state from @xmath121 ( @xmath122 ) . all the annihilation into sm particles channels proceed through @xmath107-channel higgs boson exchange while the annihilation into inert higgs particles such as @xmath123 , @xmath124 and @xmath118 will proceed through both @xmath107-channel and @xmath108-channel higgs boson exchange as well as the contact interactions with the quartic couplings for the @xmath125 , @xmath126 and @xmath127 vertices . the calculation is done using the public code ` micromegas ` @xcite . the outcome of our relic density calculation should be in agreement with the recent planck measurement @xcite : @xmath128 + as is well known if the mass splitting between the lop and next - lightest odd particle ( nlop ) is @xmath129 or so , the number densities of these nlops have only slight boltzmann suppression with respect to the lop number density . therefore , the contributions to the relic density from the scattering of lop - nlop and nlop - nlop have to be taken into account in order to have a more precise relic density prediction . these mechanisms are known as coannihilation @xcite . as these are implemented already in the package ` micromegas ` @xcite , we can take the @xmath130 , @xmath131 and @xmath132 coannihilation into account at ease . * at present the most stringent limit on the spin - independent component of elastic scattering cross section @xmath133 for @xmath134 comes from lux @xcite . they improved the minimum @xmath133 upper limit obtained by xenon100 @xcite about an overall factor of 3 . this result then sets the limit on the spin - independent cross section , @xmath135 pb for dm mass @xmath136 gev . in this study , we include the 90% upper limit obtained in @xcite for @xmath133 versus the dm mass in our likelihood function . however , one should bear in mind that @xmath133 may be susceptible to large theoretical uncertainties from the hadronic matrix elements . we will take into account the uncertainties in the hadronic matrix elements , as will be discussed later in section [ section:5 ] . * we consider the fermi - lat observations of @xmath10-rays from dsphs @xcite and gc @xcite . the @xmath137 dsphs as adopted in @xcite will be used in this work . four years of the fermi - lat data , recorded from 4 august 2008 to 2 august 2012 with the pass 7 photon selection , are employed in this analysis . the energy range of photons is chosen from @xmath138 mev to @xmath139 gev , and the region - of - interest ( roi ) is adopted to be a @xmath140 box centered on each dsphs . for the gc analysis , a slightly smaller roi region of @xmath141 is chosen to avoid too many sources in the analysis . in the likelihood analysis , the normalization of the diffuse background models gal_2yearp7v6_v0.fits and iso_p7v6source.txt , and the point sources located in the rois in the second lat catalog @xcite are left free to do the minimization . the fermi - lat data are binned into @xmath142 energy bins logarithmically spaced between @xmath143 and @xmath144 gev , and we calculate the likelihood map of fermi - lat dsphs and gc observations on the @xmath145flux plane following the method developed in @xcite . such a method is very efficient to derive the final likelihood of any specific @xmath10-ray spectrum , and is tested to be consistent with the standard analysis procedure using fermi scientific tool @xcite . * the cosmic ray positron fraction measured by pamela @xcite and most recently by ams-02 @xcite show clear evidence of excess compared with the secondary production as expected from the cosmic ray propagation model . the fluxes of the total @xmath75 measured by atic @xcite , fermi - lat @xcite , hess @xcite and magic @xcite also show the deviation from the extrapolation of the low energy pamela data @xcite , which further supports the existence of extra @xmath75 sources . there are many models proposed to account for the @xmath75 excesses , including the astrophysical sources such as pulsars and supernova remnants , and dm annihilation / decay ( see _ e.g. _ the review articles @xcite ) . + while the dm model would suffer from strong constraints from @xmath10-ray observations @xcite , it has been shown that the pulsars with reasonable parameters can explain the positron fraction as well as the electron plus positron flux data @xcite . therefore in this work we first fit both data set with the background plus pulsar - like models , and then add the dm contributions from ihdm to calculate their likelihoods @xcite . the framework for doing such calculation in this astrophysical setting can be found in @xcite . basically , a markov chain monte carlo ( mcmc ) based global fitting tool was used to determine the model parameters . the observational data used in the fit include the ams-02 positron fraction @xcite , pamela electron spectrum @xcite , and the total @xmath75 spectra of fermi - lat @xcite and hess @xcite . note that for the background electron spectrum we employ a three - piece broken power - law function in order to fit simultaneously the above data @xcite . + the solar modulation affects the fluxes of the particles at low energy . in this work we simply adopt the force - field approximation to account for the solar modulation effect @xcite . it was found that the modulation potential @xmath146 mv can fit both the positron fraction and electron spectra . however , for the cosmic ray protons , a smaller modulation potential @xmath147 mv is favoured by the pamela data @xcite . we leave this as an open question because the solar modulation may indeed depend on the mass - to - charge ratio of particles . the best - fitting positron fraction and electron spectra compared with the observational data are shown in fig . [ fig : bkgep ] . the model fits the data well and the reduced @xmath148 is about @xmath149 . + + when calculating the likelihood after adding the dm contributions in ihdm , we further multiply a factor of @xmath150 to the fluxes of the background components ( @xmath151 for the @xmath152 background , @xmath153 background and pulsar @xmath154 respectively ) , in order to take the uncertainties of the modelling into account @xcite . the parameters @xmath155 and @xmath156 are left free and treated as nuisance parameters in our analysis . they are allowed to vary in the range of @xmath157 and @xmath158 to calculate the maximum likelihood of a specific dm model point . * the precise measurement of the antiproton - to - proton ratio and antiproton flux by pamela show relatively good agreement with the cosmic ray background model expectation @xcite , which leaves limited space for the dm models @xcite . we calculate the expected antiproton flux in the same propagation model used to explain the @xmath75 data , as shown in fig . [ fig : bkgap ] . the solar modulation potential is adopted to be @xmath139 mv as suggested by pamela @xcite . same as the @xmath155 and @xmath156 in @xmath75 case , an adjustment factor @xmath159 with nuisance parameters @xmath160 and @xmath161 varied in the same respective range is employed to account for the uncertainties of the background estimation . in table [ tab : exp_constraints ] , we summarize all the experimental constraints mentioned in this and previous section . to avoid words cluttering in later presentation , we denote the first block of relic density and collider constraints together with the theoretical constraints as * rc * ( relic density and collider ) , the second block of lux constraint as * dd * ( direct detection ) and the third block of constraints as * i d * ( indirect detection ) . additionally , we reject those points during our parameter scans which violate any one of the theoretical constraints on ihdm mentioned in section [ section:2b ] . we note that the data in the * rc * block does not involve large theoretical uncertainties compared with the other two blocks , so we will take special care of this block by including it only at the scan level . . the experimental constraints that we include in our likelihood functions to constrain the ihdm model . we denote the first block of relic density and collider constraints together with the theoretical constraints as * rc * ( relic and collider ) , lux constraint as * dd * ( direct detection ) and the last block of constraints as * i d * ( indirect detection ) . [ cols="<,<,<,<,<,<",options="header " , ] in table [ tab : bestpts ] , we summarize the best - fit points for the three different blocks from our scan . the second column is only with * rc * constraints in the likelihood while the third and fourth columns are with * rc+id * and * rc+dd+id * in the likelihood , respectively . we would like to stress that there is no preference of @xmath42 or @xmath43 due to the symmetry between @xmath24 and @xmath25 in the model . we find that the maximum likelihood of @xmath162 and @xmath163 are roughly the same . therefore , the fact that the best - fit points are located at @xmath42 or @xmath43 region is just due to the fact that we collected the maximum likelihood before hitting the sampling stop criteria . in other words , we can not tell the dark matter in ihdm must be a scalar or pseudoscalar from this analysis . however , we can see that the best - fit points of three sets of constraint are all located at the lower @xmath12 region . the reasons for this are mainly due to the ewpt and indirect detection constraints . first , the nearly degeneracy between @xmath80 and @xmath12 required by relic density constraints at the @xmath164 region implies @xmath24 and @xmath60 are always negligible . on the other hand , @xmath24 and @xmath60 can be enhanced in the @xmath165 region due to larger mass splitting between @xmath28 and @xmath9 . hence , we can obtain the better likelihood in small @xmath12 region . second , the lower @xmath12 region may amplify the indirect detection signal too , because the signal fluxes of indirect detection experiments are inversely proportional to @xmath166 . hence , lower @xmath12 region can have stronger signals to fit current * i d * constraints than astrophysical background only . on the other hand , the fluxes from dm annihilation in the @xmath164 region can be suppressed and may be lower than the astrophysical background . as seen from figs . [ fig : bkgep ] and [ fig : bkgap ] , there are still some rooms for dm in indirect detection experiments . perhaps not a discovery with large statistical significance , but a weak signal usually fits the likelihood better than using the background only hypothesis . certainly , our understandings of the astrophysical backgrounds could be too naive . the @xmath167 profile likelihood of the * rc+id+dd * constraints projected on the ( @xmath12 , @xmath168 ) plane for the monojet result at lhc-14 . the black crosses will be excluded by the future 100 @xmath169 data . the orange square will still be allowed by 100 @xmath169 data but disfavoured by 300 @xmath169 data . the red dots will be allowed by 300 @xmath169 data . they are all in @xmath167 significance . , width=384 ] in this section , we will consider the sensitivities of three future experiments : ( 1 ) lhc-14 monojet with luminosities 100 fb@xmath13 and 300 fb@xmath13 , ( 2 ) xenon1 t , and ( 3 ) ams-02 one year antiproton data , given the parameter space obtained in previous section that satisfies the * rc+dd+id * constraints . at the 14 tev run of lhc ( lhc-14 ) , we will discuss the impact of the monojet search with luminosities of 100 fb@xmath13 and 300 fb@xmath13 . we assume a null measurement of dm with background events same as observed , _ i.e. _ @xmath170 where @xmath109 and @xmath171 are the background and observed events at 100 fb@xmath13 or 300 fb@xmath13 obtained by scaling the current cms data at luminosity of @xmath172 . using the standard sensitivity formula , @xmath173 , we compute the significance of monojet for lhc-14 . in fig . [ fig : mx_laml_lhc14 ] , we present the potential power of lhc monojet search with 100 fb@xmath13 and 300 fb@xmath13 on the ( @xmath12,@xmath168 ) plane . in linear scale , we zoom into the region @xmath174 where the invisible higgs decay is open . all the points shown satisfy the * rc+dd+id * constraints in @xmath167 . with 100 fb@xmath13 of data , only the orange boxes will be allowed while the few black crosses located near the boundary where @xmath175 will be disfavoured in the @xmath167 subset . however , with 300 fb@xmath13 of data , the range of @xmath168 will be extended to the region of red dots in fig . [ fig : mx_laml_lhc14 ] where @xmath176 . in fig . [ fig : mx_sigv_xenon1 t ] , we show the disfavoured region by future xenon1 t sensitivities subjected to the * rc+id+dd * constraints in @xmath167 significance . the left panel is for @xmath177 and the right panel is for ( @xmath12 , @xmath178 ) . the red dots are favoured but gray dots / crosses are disfavoured by xenon1 t limit . although the region of @xmath164 can not be entirely ruled out by xenon1 t from our global analysis based on tree level calculation , a recent paper @xcite pointed out that electroweak corrections can significantly alter the theoretical prediction of @xmath133 , especially for large @xmath12 region . as shown in their computation , @xmath133 is not expected to be lower than @xmath179 pb even when one loop corrections are included @xcite . we thus expect next generation of ton - sized detectors for dm direct detection can probe most of the parameter space of ihdm . the left and right panels show the sensitivity of xenon1 t on the ( @xmath12 , @xmath168 ) and ( @xmath12 , @xmath178 ) planes respectively subjected to the * rc+id+dd * constraints in @xmath167 significance . the red ( gray ) area will be allowed ( excluded ) by xenon1 t sensitivity . , title="fig:",width=288 ] the left and right panels show the sensitivity of xenon1 t on the ( @xmath12 , @xmath168 ) and ( @xmath12 , @xmath178 ) planes respectively subjected to the * rc+id+dd * constraints in @xmath167 significance . the red ( gray ) area will be allowed ( excluded ) by xenon1 t sensitivity . , title="fig:",width=288 ] we generate the simulated one year ams-02 data of the antiprotons following ref . @xcite . the expected antiproton flux @xmath180 is adopted to be the one described in section [ section:2 ] ( see fig . [ fig : bkgap ] ) . the number of antiproton events in a given energy bin is approximately @xmath181 where @xmath182 is the simulated geometry factor given in @xcite , @xmath183 is the width of the energy bin and @xmath184 is the exposure time . we generate the data from 1 to 300 gev , with 50 bins logarithmically evenly distributed according to the binning of the positron fraction measurement by ams-02 @xcite . for the `` observed '' number of events we apply a poisson fluctuation on @xmath185 , with statistical error @xmath186 . the systematic error is simply adopted to be @xmath187 @xcite , which is added quadratically to the statistical error . the simulated antiproton flux for one year observation of ams-02 is shown in the left panel of fig . [ fig : ap_ams02 ] . using the simulated antiproton data , we calculate the @xmath148 of each dm model point with the same method described in sections [ section:4 ] and [ section:5 ] . in the right panel of fig . [ fig : ap_ams02 ] , the exclusion power of the ams-02 one year antiproton data is shown . the red dots are allowed but gray crosses are disfavoured by future sensitivity . comparing with the exclusion power of xenon1 t sensitivity ( right panel of fig . [ fig : mx_sigv_xenon1 t ] ) , we can see most of parameter space excluded by ams-02 antiproton data are also excluded by xenon1 t . however , for the @xmath188 at the higgs resonance region ( near the top of the second vertical branch from the left ) , one can find some small fractions of red dots that can be excluded by ams-02 antiproton data but not yet ruled out by xenon1 t . this is because the dm annihilation channels are dominant by @xmath189 or @xmath190 final state at these points . however the @xmath191 coupling is irrelevant to direct detection and the @xmath192 coupling can contribute to direct detection only by integrating this heavy @xmath109 quark to obtain the @xmath193 coupling . hence its contributions to @xmath133 is also small as compared with light quarks . , @xmath12 ) plane ( _ right _ ) . the red dots ( gray crosses ) will be allowed ( excluded ) by ams-02 anti - proton one year sensitivity . , title="fig:",width=288 ] , @xmath12 ) plane ( _ right _ ) . the red dots ( gray crosses ) will be allowed ( excluded ) by ams-02 anti - proton one year sensitivity . , title="fig:",width=288 ] we finally show in fig . [ fig : mx_laml_allfuture ] the total impact from the combined sensitivities from the above three future experiments on the ( @xmath12,@xmath168 ) plane . for the lhc-14 monojet , we will assume 300 fb@xmath13 of data in making these two plots . left and right panels correspond to @xmath12 less than 100 gev and greater than 500 gev respectively . with all three future experiments sensitivities , from the left panel of a zoomed - in view of low @xmath12 region , we see that the lower limit of @xmath12 is lifted slightly from @xmath194 to @xmath195 . in this low mass region where the invisible decay @xmath196 is open , while we do not found any upper limit on @xmath12 , @xmath168 is found to lie between @xmath197 and @xmath198 . on the other hand , if the invisible mode is closed , the upper and lower limit of @xmath168 is varied with respect to @xmath12 ( right panel ) . comparing with current experimental data , these three future experiments sensitivities are robust but neither the lower @xmath12 region nor the larger @xmath12 region can be entirely ruled out . the scatter plots to present the future combined sensitivity from xenon1 t , lhc-14 monojet ( 300 fb@xmath13 ) , and ams-02 antiproton flux measurements on ( @xmath12,@xmath199 ) plane subjected to the * rc+id+dd * constraints in @xmath167 significance . the left and right panels are for @xmath200 gev and @xmath201 gev respectively . , title="fig:",width=288 ] the scatter plots to present the future combined sensitivity from xenon1 t , lhc-14 monojet ( 300 fb@xmath13 ) , and ams-02 antiproton flux measurements on ( @xmath12,@xmath199 ) plane subjected to the * rc+id+dd * constraints in @xmath167 significance . the left and right panels are for @xmath200 gev and @xmath201 gev respectively . , title="fig:",width=288 ] + despite ihdm was proposed more than three decades ago , it is still one of the most simplest models and yet viable for scalar dark matter . we have performed a global fit analysis on this model . this analysis has been performed in light of the recent atlas and cms discovery of a 125 - 126 gev higgs - like particle , taking into account the recent relic density measurement by planck , dm direct detection from lux and indirect detection from pamela , fermi - lat and ams-02 . we have shown that the constraint from dm direct detection search , such as the latest lux upper limit of year 2013 , provides a robust constraint on the parameter space . in particular , if the invisible decay of the sm higgs boson is open , the upper limit of the higgs invisible width from lhc together with the lux constraint could put some interesting limits on the sm higgs boson couplings to the dm in ihdm . indeed , an invisible decay of the sm higgs boson with a branching ratio larger than 30% and a scalar dark matter mass within the range of 20 - 60 gev are excluded by current data . we emphasize that there is no preference of @xmath42 or @xmath25 in our study . however , we found that @xmath165 region is slightly favoured by ewpt and * i d * constraints than @xmath202 region . in addition , in the 95% c.l . of * rc+dd+id * constraints , @xmath12 has the lower limit around @xmath194 . the exclusion power of the ams-02 one year antiproton data on the model parameters can be inferred from figs . [ fig : ap_ams02 ] and [ fig : mx_laml_allfuture ] . the results show that ams-02 does have the potential to constrain certain parameter space of ihdm , especially in the low mass region . however , compared with the direct detection limit expected from xenon1 t , the constraint from ams-02 antiproton data is weaker as comparing the two figs . [ fig : mx_sigv_xenon1 t ] and [ fig : ap_ams02 ] . nevertheless , future data from lhc-14 monojet , xenon1 t direct detection and ams-02 antiproton flux can further reduce the ihdm parameter space constrained by the existing * rc+dd+id * data , as is shown inevitably in the red regions as compared with the gray regions in the two plots at the left and right panels of fig . [ fig : mx_laml_allfuture ] for the low and high dm mass regions respectively . we also note that the likelihoods obtained in this work were obtained using the tree level relation for @xmath199 and @xmath37 ( eq . ( [ eq : lambda_a ] ) ) and the tree level formula for the coupling @xmath47 ( eq . ( [ eq : lamchi ] ) ) . higher order corrections will necessarily modify these relations and hence the relic density prediction will be affected . the profile likelihoods for the ihdm will be modified as well . although loop corrections have been shown to affect significantly the dm scattering cross - section on nucleons in the ihdm @xcite and thus modify the impact of direct detection searches on the viable parameter space of the model , it is beyond the scope of the present analysis to take them into account . one should bear in mind that such corrections will also probably affect the positions of the best - fit points . in summary , we note that the overall shapes of the 95% c.l . contours presented in this work are mainly determined by the planck relic density measurement . lhc monojet is only relevant when the invisible decay of the higgs is open , in which case we obtain the limit for the coupling @xmath203 which is not very stringent . on the other hand , our best - fit points are located at @xmath204 70 gev and so do not allow the opening for invisible decay of the sm higgs into dm . the current * i d * and * dd * data are only sensitive to higgs resonance region . except for relic density constraint , currently no other experimental data sets are sensitive to the ihdm parameter space at @xmath202 . however , one expects future xenon1 t can probe this region . moreover , future instruments such as dampe , gamma-400 and cta will test this higher @xmath12 region as well @xcite . as mentioned previously , experimental search for the inert higgs ( neutral and charged ) behaves like susy search for charginos and neutralinos . therefore , one of the most popular signatures for inert higgs searches would be also trilepton and/or dilepton plus missing @xmath205 @xcite . a dedicated analysis for the ihdm has been performed in @xcite where it has been demonstrated that the experimental reach for the inert higgses is only about 300 gev which is somewhat smaller than the lhc reach for the charginos and neutralinos . the main reason is that the cross section for the scalar pair production @xmath206 is smaller than the gaugino pair production @xmath207 . as we have seen in our analysis , inert higgses with masses @xmath208 tev are consistent with all experimental and theoretical constraints . one concludes that the ihdm is here to stay for another decade . the ihdm can be further extended by including inert right - handed neutrinos with majorana masses @xcite . masses for the sm light neutrinos can be generated through radiative processes with only inert particles running inside the loop . this extension of ihdm would exhibit intricate interplay between dark matter and neutrino physics . detailed global analysis of this extended model is also quite interesting and will be presented in a future publication . aa would like to thank the national science council ( nsc ) of taiwan for financial support during his stay at the institute of physics ( iop ) , academia sinica ( as ) . st would like to thank the hospitality and support from the high energy theory group at the iop , as . this work is supported by national natural science foundation of china under grant number 11105155 ( qy ) and by the nsc of taiwan under project number 102 - 2811-m-001 - 032 ( aa ) and grant number 101 - 2112-m-001 - 005-my3 ( tcy ) . 99 g. aad _ et al . _ [ atlas collaboration ] , phys . b * 716 * , 1 ( 2012 ) [ arxiv:1207.7214 [ hep - ex ] ] . s. chatrchyan _ et al . _ [ cms collaboration ] , phys . b * 716 * , 30 ( 2012 ) [ arxiv:1207.7235 [ hep - ex ] ] . t. aaltonen _ et al . _ [ cdf and d0 collaborations ] , phys . lett . * 109 * , 071804 ( 2012 ) [ arxiv:1207.6436 [ hep - ex ] ] ; phys . d * 88 * , 052014 ( 2013 ) [ arxiv:1303.6346 [ hep - ex ] ] . g. aad _ et al . _ [ atlas collaboration ] , phys . b * 726 * , 88 ( 2013 ) [ arxiv:1307.1427 [ hep - ex ] ] ; atlas - conf-2013 - 079 , 19-july-2013 . d. carmi , a. falkowski , e. kuflik and t. volansky , jhep * 1207 * , 136 ( 2012 ) [ arxiv:1202.3144 [ hep - ph ] ] ; a. azatov , r. contino and j. galloway , jhep * 1204 * , 127 ( 2012 ) [ erratum - ibid . * 1304 * , 140 ( 2013 ) ] [ arxiv:1202.3415 [ hep - ph ] ] ; m. klute , r. lafaye , t. plehn , m. rauch and d. zerwas , phys . lett . * 109 * , 101801 ( 2012 ) [ arxiv:1205.2699 [ hep - ph ] ] ; i. low , j. lykken and g. shaughnessy , phys . d * 86 * , 093012 ( 2012 ) [ arxiv:1207.1093 [ hep - ph ] ] ; a. azatov , s. chang , n. craig and j. galloway , phys . rev . d * 86 * , 075033 ( 2012 ) [ arxiv:1206.1058 [ hep - ph ] ] ; j. r. espinosa , c. grojean , m. muhlleitner and m. trott , jhep * 1205 * , 097 ( 2012 ) [ arxiv:1202.3697 [ hep - ph ] ] ; j. ellis and t. you , jhep * 1209 * , 123 ( 2012 ) [ arxiv:1207.1693 [ hep - ph ] ] ; p. p. giardino , k. kannike , m. raidal and a. strumia , phys . b * 718 * , 469 ( 2012 ) [ arxiv:1207.1347 [ hep - ph ] ] . g. aad _ et al . _ [ atlas collaboration ] , arxiv:1307.1432 [ hep - ex ] ; g. aad _ et al . _ [ atlas collaboration ] : atlas - conf-2013 - 040 . s. chatrchyan _ et al . _ [ cms collaboration ] , phys . lett . * 110 * , 081803 ( 2013 ) [ arxiv:1212.6639 [ hep - ex ] ] ; cms - pas - hig-13 - 002 . n. g. deshpande and e. ma , phys . d * 18 * , 2574 ( 1978 ) . for a review of the general two higgs doublet models , see g. c. branco , p. m. ferreira , l. lavoura , m. n. rebelo , m. sher and j. p. silva , phys . rept . * 516 * , 1 ( 2012 ) [ arxiv:1106.0034 [ hep - ph ] ] . k. cheung , j. s. lee and p. -y . tseng , jhep * 1305 * , 134 ( 2013 ) [ arxiv:1302.3794 [ hep - ph ] ] . g. belanger , b. dumont , u. ellwanger , j. f. gunion and s. kraml , phys . b * 723 * , 340 ( 2013 ) [ arxiv:1302.5694 [ hep - ph ] ] . g. belanger , b. dumont , u. ellwanger , j. f. gunion and s. kraml , phys . rev . d * 88 * , 075008 ( 2013 ) [ arxiv:1306.2941 [ hep - ph ] ] . j. r. espinosa , m. muhlleitner , c. grojean and m. trott , jhep * 1209 * , 126 ( 2012 ) [ arxiv:1205.6790 [ hep - ph ] ] ; o. lebedev , h. m. lee and y. mambrini , phys . lett . b * 707 * , 570 ( 2012 ) [ arxiv:1111.4482 [ hep - ph ] ] ; c. englert , m. spannowsky and c. wymant , phys . lett . b * 718 * , 538 ( 2012 ) [ arxiv:1209.0494 [ hep - ph ] ] . e. ma , phys . d * 73 * , 077301 ( 2006 ) [ hep - ph/0601225 ] . a. arhrib , y .- l . sming tsai , q. yuan and t. c. yuan , work in progress . v. silveira and a. zee , phys . b * 161 * , 136 ( 1985 ) . k. cheung , y. -l . s. tsai , p. -y . tseng , t. -c . yuan and a. zee , jcap * 1210 * , 042 ( 2012 ) [ arxiv:1207.4930 [ hep - ph ] ] . j. mcdonald , phys . d * 50 * , 3637 ( 1994 ) ; c. p. burgess , m. pospelov and t. ter veldhuis , nucl . b * 619 * , 709 ( 2001 ) ; m. cirelli , n. fornengo and a. strumia , nucl . b * 753 * , 178 ( 2006 ) . m. c. bento , o. bertolami , r. rosenfeld and l. teodoro , phys . d * 62 * , 041302 ( 2000 ) ; v. barger , p. langacker , m. mccaskey , m. j. ramsey - musolf and g. shaughnessy , phys . d * 77 * , 035005 ( 2008 ) ; m. aoki , s. kanemura and o. seto , phys . * 102 * , 051805 ( 2009 ) [ arxiv:0807.0361 [ hep - ph ] ] ; phys . d * 80 * , 033007 ( 2009 ) [ arxiv:0904.3829 [ hep - ph ] ] . k. cheung and t. c. yuan , phys . b * 685 * , 182 ( 2010 ) ; x. g. he , t. li , x. q. li , j. tandean and h. c. tsai , phys . b * 688 * , 332 ( 2010 ) ; m. farina , d. pappadopulo and a. strumia , phys . b * 688 * , 329 ( 2010 ) ; m. kadastik , k. kannike , a. racioppi and m. raidal , phys . b * 685 * , 182 ( 2010 ) arxiv:0912.3797 [ hep - ph ] ; m. aoki , s. kanemura and o. seto , phys . b * 685 * , 313 ( 2010 ) ; m. asano and r. kitano , phys . d * 81 * , 054506 ( 2010 ) [ arxiv:1001.0486 [ hep - ph ] ] ; a. bandyopadhyay , s. chakraborty , a. ghosal and d. majumdar , jhep * 1011 * , 065 ( 2010 ) [ arxiv:1003.0809 [ hep - ph ] ] ; s. andreas , c. arina , t. hambye , f. -s . ling and m. h. g. tytgat , phys . d * 82 * , 043522 ( 2010 ) [ arxiv:1003.2595 [ hep - ph ] ] . m. gustafsson , s. rydbeck , l. lopez - honorez and e. lundstrom , phys . d * 86 * , 075019 ( 2012 ) [ arxiv:1206.6316 [ hep - ph ] ] . a. goudelis , b. herrmann and o. stl , jhep * 1309 * , 106 ( 2013 ) [ arxiv:1303.3010 [ hep - ph ] ] . l. lopez honorez and c. e. yaguna , jcap * 1101 * ( 2011 ) 002 [ arxiv:1011.1411 [ hep - ph ] ] . l. lopez honorez , e. nezri , j. f. oliver and m. h. g. tytgat , jcap * 0702 * , 028 ( 2007 ) [ hep - ph/0612275 ] ; t. hambye and m. h. g. tytgat , phys . b * 659 * , 651 ( 2008 ) [ arxiv:0707.0633 [ hep - ph ] ] ; p. agrawal , e. m. dolle and c. a. krenke , phys . d * 79 * , 015015 ( 2009 ) [ arxiv:0811.1798 [ hep - ph ] ] ; t. hambye , f. -s . ling , l. lopez honorez and j. rocher , jhep * 0907 * , 090 ( 2009 ) [ erratum - ibid . * 1005 * , 066 ( 2010 ) ] [ arxiv:0903.4010 [ hep - ph ] ] ; e. nezri , m. h. g. tytgat and g. vertongen , jcap * 0904 * , 014 ( 2009 ) [ arxiv:0901.2556 [ hep - ph ] ] ; s. andreas , m. h. g. tytgat and q. swillens , jcap * 0904 * , 004 ( 2009 ) [ arxiv:0901.1750 [ hep - ph ] ] ; c. arina , f. -s . ling and m. h. g. tytgat , jcap * 0910 * , 018 ( 2009 ) [ arxiv:0907.0430 [ hep - ph ] ] ; l. lopez honorez and c. e. yaguna , jhep * 1009 * , 046 ( 2010 ) [ arxiv:1003.3125 [ hep - ph ] ] ; a. melfo , m. nemevsek , f. nesti , g. senjanovic and y. zhang , phys . d * 84 * , 034009 ( 2011 ) [ arxiv:1105.4611 [ hep - ph ] ] ; m. krawczyk , d. sokolowska , p. swaczyna and b. swiezewska , jhep * 1309 * , 055 ( 2013 ) [ arxiv:1305.6266 [ hep - ph ] ] ; m. gustafsson , s. rydbeck , l. lopez - honorez and e. lundstrom , phys . d * 86 * , 075019 ( 2012 ) [ arxiv:1206.6316 [ hep - ph ] ] ; e. m. dolle and s. su , phys . d * 80 * , 055012 ( 2009 ) [ arxiv:0906.1609 [ hep - ph ] ] ; s. kanemura , y. okada , h. taniguchi and k. tsumura , phys . b * 704 * , 303 ( 2011 ) [ arxiv:1108.3297 [ hep - ph ] ] . a. arhrib , r. benbrik and n. gaur , phys . d * 85 * , 095021 ( 2012 ) [ arxiv:1201.2644 [ hep - ph ] ] . b. swiezewska and m. krawczyk , phys . d * 88 * , 035019 ( 2013 ) [ arxiv:1212.4100 [ hep - ph ] ] ; arxiv:1305.7356 [ hep - ph ] . q. -h . cao , e. ma and g. rajasekaran , phys . d * 76 * , 095011 ( 2007 ) [ arxiv:0708.2939 [ hep - ph ] ] . e. dolle , x. miao , s. su and b. thomas , phys . d * 81 * , 035003 ( 2010 ) [ arxiv:0909.3094 [ hep - ph ] ] ; x. miao , s. su and b. thomas , phys . d * 82 * , 035009 ( 2010 ) [ arxiv:1005.0090 [ hep - ph ] ] . e. lundstrom , m. gustafsson and j. edsjo , phys . d * 79 * , 035013 ( 2009 ) [ arxiv:0810.3924 [ hep - ph ] ] . r. barbieri , l. j. hall and v. s. rychkov , phys . d * 74 * , 015007 ( 2006 ) [ hep - ph/0603188 ] . p. a. r. ade _ et al . _ [ planck collaboration ] , arxiv:1303.5076 [ astro-ph.co ] . d. s. akerib _ et al . _ [ lux collaboration ] , arxiv:1310.8214 [ astro-ph.co ] . m. ackermann _ et al . _ [ fermi - lat collaboration ] , phys . rev . * 107 * , 241302 ( 2011 ) [ arxiv:1108.3546 [ astro-ph.he ] ] . x. huang , q. yuan , p. -f . yin , x. -j . bi and x. chen , jcap * 1211 * , 048 ( 2012 ) [ erratum - ibid . * 1305 * , e02 ( 2013 ) ] [ arxiv:1208.0267 [ astro-ph.he ] ] . m. aguilar _ et al . _ [ ams collaboration ] , phys . lett . * 110 * , 141102 ( 2013 ) . o. adriani _ et al . _ [ pamela collaboration ] , nature * 458 * , 607 ( 2009 ) [ arxiv:0810.4995 [ astro - ph ] ] . m. ackermann _ et al . _ [ fermi lat collaboration ] , phys . lett . * 108 * , 011103 ( 2012 ) [ arxiv:1109.0521 [ astro-ph.he ] ] . m. ackermann _ et al . _ [ fermi lat collaboration ] , phys . rev . d * 82 * , 092004 ( 2010 ) [ arxiv:1008.3999 [ astro-ph.he ] ] . o. adriani _ et al . _ [ pamela collaboration ] , phys . lett . * 105 * , 121101 ( 2010 ) [ arxiv:1007.0821 [ astro-ph.he ] ] . e. aprile _ et al . _ [ xenon100 collaboration ] , phys . rev . lett . * 109 * , 181301 ( 2012 ) [ arxiv:1207.5988 [ astro-ph.co ] ] . m. gustafsson , e. lundstrom , l. bergstrom and j. edsjo , phys . lett . * 99 * , 041301 ( 2007 ) [ astro - ph/0703512 [ astro - ph ] ] . m. e. peskin and t. takeuchi , phys . d * 46 * ( 1992 ) 381 . m. baak , m. goebel , j. haller , a. hoecker , d. kennedy , r. kogler , k. moenig and m. schott _ et al . _ , eur . j. c * 72 * , 2205 ( 2012 ) [ arxiv:1209.2716 [ hep - ph ] ] . k. nakamura _ et al . _ [ particle data group collaboration ] , j. phys . g g * 37 * , 075021 ( 2010 ) . a. pierce and j. thaler , jhep * 0708 * , 026 ( 2007 ) [ hep - ph/0703056 [ hep - ph ] ] . m. acciarri _ et al . _ [ l3 collaboration ] , phys . b * 472 * , 420 ( 2000 ) [ hep - ex/9910007 ] ; r. barate _ [ aleph collaboration ] , phys . b * 499 * , 67 ( 2001 ) [ hep - ex/0011047 ] ; j. abdallah _ et al . _ [ delphi collaboration ] , eur . j. c * 31 * , 421 ( 2003 ) [ hep - ex/0311019 ] ; g. abbiendi _ et al . _ [ opal collaboration ] , eur . j. c * 35 * , 1 ( 2004 ) [ hep - ex/0401026 ] . atlas collaboration , atlas - conf-2013 - 011 . cms collaboration , cms - pas - hig-13 - 018 . cms collaboration , cms - pas - hig-13 - 013 . atlas collaboration , atlas - conf-2013 - 012 . c. palmer [ cms collaboration ] , arxiv:1305.3654 [ hep - ex ] ; s. chatrchyan _ et al . _ [ cms collaboration ] , cms - pas - hig-13 - 001 . cms collaboration , cms - pas - exo-12 - 048 . atlas collaboration , atlas - conf-2012 - 147 . g. belanger , f. boudjema , a. pukhov and a. semenov , comput . commun . * 180 * , 747 ( 2009 ) [ arxiv:0803.2360 [ hep - ph ] ] ; g. belanger , f. boudjema , a. pukhov and a. semenov , nuovo cim . c * 033n2 * , 111 ( 2010 ) [ arxiv:1005.4133 [ hep - ph ] ] . d. larson , j. dunkley , g. hinshaw , e. komatsu , m. r. nolta , c. l. bennett , b. gold and m. halpern _ et al . _ , astrophys . j. suppl . * 192 * , 16 ( 2011 ) [ arxiv:1001.4635 [ astro-ph.co ] ] . k. griest and d. seckel , phys . d * 43 * , 3191 ( 1991 ) . p. binetruy , g. girardi and p. salati , nucl . b * 237 * , 285 ( 1984 ) . a. geringer - sameth and s. m. koushiappas , phys . lett . * 107 * , 241303 ( 2011 ) [ arxiv:1108.2914 [ astro-ph.co ] ] . y. -l . s. tsai , q. yuan and x. huang , jcap * 1303 * , 018 ( 2013 ) [ arxiv:1212.3990 [ astro-ph.he ] ] . fermi - lat collaboration , astrophys . j. suppl . * 199 * , 31 ( 2012 ) [ arxiv:1108.1435 [ astro-ph.he ] ] . j. chang , j. h. adams , h. s. ahn , g. l. bashindzhagyan , m. christl , o. ganel , t. g. guzik and j. isbert _ et al . _ , nature * 456 * , 362 ( 2008 ) . f. aharonian _ et al . _ [ h.e.s.s . collaboration ] , phys . * 101 * , 261104 ( 2008 ) [ arxiv:0811.3894 [ astro - ph ] ] ; f. aharonian _ et al . _ [ h.e.s.s . collaboration ] , astron . astrophys . * 508 * , 561 ( 2009 ) [ arxiv:0905.0105 [ astro-ph.he ] ] . d. borla tridon _ [ magic collaboration ] , arxiv:1110.4008 [ astro-ph.he ] . o. adriani _ et al . _ [ pamela collaboration ] , phys . lett . * 106 * , 201101 ( 2011 ) [ arxiv:1103.2880 [ astro-ph.he ] ] . x. -g . he , mod . a * 24 * , 2139 ( 2009 ) [ arxiv:0908.2908 [ hep - ph ] ] ; y. -z . fan , b. zhang and j. chang , int . j. mod . d * 19 * , 2011 ( 2010 ) [ arxiv:1008.4646 [ astro-ph.he ] ] ; p. d. serpico , astropart . phys . * 39 - 40 * , 2 ( 2012 ) [ arxiv:1108.4827 [ astro-ph.he ] ] ; m. cirelli , pramana * 79 * , 1021 ( 2012 ) [ arxiv:1202.1454 [ hep - ph ] ] . g. bertone , m. cirelli , a. strumia and m. taoso , jcap * 0903 * , 009 ( 2009 ) [ arxiv:0811.3744 [ astro - ph ] ] ; l. bergstrom , g. bertone , t. bringmann , j. edsjo and m. taoso , phys . d * 79 * , 081303 ( 2009 ) [ arxiv:0812.3895 [ astro - ph ] ] ; m. cirelli , p. panci and p. d. serpico , nucl . b * 840 * , 284 ( 2010 ) [ arxiv:0912.0663 [ astro-ph.co ] ] ; m. papucci and a. strumia , jcap * 1003 * , 014 ( 2010 ) [ arxiv:0912.0742 [ hep - ph ] ] . d. grasso _ et al . _ [ fermi - lat collaboration ] , astropart . phys . * 32 * , 140 ( 2009 ) [ arxiv:0905.0636 [ astro-ph.he ] ] ; t. linden and s. profumo , astrophys . j. * 772 * , 18 ( 2013 ) [ arxiv:1304.1791 [ astro-ph.he ] ] ; p. -f . yin , z. -h . yu , q. yuan and x. -j . bi , phys . d * 88 * , 023001 ( 2013 ) [ arxiv:1304.4128 [ astro-ph.he ] ] . l. bergstrom , t. bringmann , i. cholis , d. hooper and c. weniger , phys . lett . * 111 * , 171101 ( 2013 ) [ arxiv:1306.3983 [ astro-ph.he ] ] . q. yuan , x. -j . bi , g. -m . chen , y. -q . guo , s. -j . lin and x. zhang , arxiv:1304.1482 [ astro-ph.he ] ; q. yuan and x. -j . bi , phys . lett . b * 727 * , 1 ( 2013 ) [ arxiv:1304.2687 [ astro-ph.he ] ] . l. j. gleeson and w. i. axford , astrophys . j. * 154 * , 1011 ( 1968 ) . o. adriani _ et al . _ [ pamela collaboration ] , science * 332 * , 69 ( 2011 ) [ arxiv:1103.4055 [ astro-ph.he ] ] . a. de simone , a. riotto and w. xue , jcap * 1305 * , 003 ( 2013 ) [ arxiv:1304.1336 [ hep - ph ] ] ; h. -b . jin , y. -l . wu and y. -f . zhou , jcap * 1311 * , 026 ( 2013 ) [ arxiv:1304.1997 [ hep - ph ] ] . o. adriani , g. c. barbarino , g. a. bazilevskaya , r. bellotti , m. boezio , e. a. bogomolov , l. bonechi and m. bongi _ et al . _ , phys . lett . * 102 * , 051101 ( 2009 ) [ arxiv:0810.4994 [ astro - ph ] ] . f. donato , d. maurin , p. brun , t. delahaye and p. salati , phys . lett . * 102 * , 071301 ( 2009 ) [ arxiv:0810.5292 [ astro - ph ] ] ; i. cholis , jcap * 1109 * , 007 ( 2011 ) [ arxiv:1007.1160 [ astro-ph.he ] ] . m. aguilar _ et al . _ [ ams collaboration ] , phys . rept . * 366 * , 331 ( 2002 ) [ erratum - ibid . * 380 * , 97 ( 2003 ) ] . y. asaoka , y. shikaze , k. abe , k. anraku , m. fujikawa , h. fuke , m. imori and s. haino _ et al . _ , phys . lett . * 88 * , 051101 ( 2002 ) [ astro - ph/0109007 ] . et al . _ , proc . cosmic ray conf . 3 , 13 ( 2005 ) . r. r. de austri , r. trotta and l. roszkowski , jhep * 0605 * , 002 ( 2006 ) [ hep - ph/0602028 ] . l. roszkowski , e. m. sessolo and y. -l . s. tsai , phys . d * 86 * , 095005 ( 2012 ) [ arxiv:1202.1503 [ hep - ph ] ] . j. alwall , m. herquet , f. maltoni , o. mattelaer and t. stelzer , jhep * 1106 * , 128 ( 2011 ) [ arxiv:1106.0522 [ hep - ph ] ] . d. hooper and l. goodenough , phys . b * 697 * , 412 ( 2011 ) [ arxiv:1010.2752 [ hep - ph ] ] ; a. boyarsky , d. malyshev and o. ruchayskiy , phys . b * 705 * , 165 ( 2011 ) [ arxiv:1012.5839 [ hep - ph ] ] ; d. hooper and t. linden , phys . d * 84 * , 123005 ( 2011 ) [ arxiv:1110.0006 [ astro-ph.he ] ] ; k. n. abazajian and m. kaplinghat , phys . d * 86 * , 083511 ( 2012 ) [ arxiv:1207.6047 [ astro-ph.he ] ] ; c. gordon and o. macas , arxiv:1306.5725 [ astro-ph.he ] . j. stahov , h. clement and g. j. wagner , phys . b * 726 * , 685 ( 2013 ) [ arxiv:1211.1148 [ nucl - th ] ] . f. feroz , m. p. hobson and m. bridges , mon . not . soc . * 398 * , 1601 ( 2009 ) [ arxiv:0809.3437 [ astro - ph ] ] . w. a. rolke , a. m. lopez and j. conrad , nucl . instrum . a * 551 * , 493 ( 2005 ) [ physics/0403059 ] . s. kanemura , s. matsumoto , t. nabeshima and n. okada , phys . d * 82 * , 055026 ( 2010 ) [ arxiv:1005.5651 [ hep - ph ] ] . a. djouadi , a. falkowski , y. mambrini and j. quevillon , eur . j. c * 73 * , 2455 ( 2013 ) [ arxiv:1205.3169 [ hep - ph ] ] . m. klasen , c. e. yaguna and j. d. ruiz - alvarez , phys . d * 87 * ( 2013 ) 075025 [ arxiv:1302.1657 [ hep - ph ] ] . m. pato , d. hooper and m. simet , jcap * 1006 * , 022 ( 2010 ) [ arxiv:1002.3341 [ astro-ph.he ] ] . a. malinlin , http://www.atic.umd.edu/pub/ams/pub07/malinin.lomonosovconf07.pdf

rhythmic elements are basic building blocks of nature at human scales . they play particularly important roles in the functioning of biological systems , such as cardiac muscle cells , pacemaker neurons , and animals or plants following circadian or circannual rhythms . recently , synchronization among non - interacting rhythmic elements induced by common fluctuating external force has attracted much attention in diverse fields . it has been demonstrated for a wide class of rhythmic elements ranging from lasers and electronic circuits to biological systems @xcite . for example , reliable synchronous response of neurons due to common or shared input signals has been actively discussed in neuroscience @xcite . in ecology , it is known that , due to common climate fluctuations , populations of plants exhibit large - scale synchronized flowering and production of seed crops also fluctuate synchronously from year to year @xcite , and such phenomena are generally termed the `` moran effect '' @xcite . we often characterize a rhythmic element as a limit - cycle oscillator . populations of coupled limit - cycle oscillators show a variety of interesting collective behavior , including mutual synchronization , wave propagation , and chaos @xcite . it is also well known that uncoupled limit - cycle oscillators can mutually synchronize when they are driven by common impulses @xcite . the simplest classical situation is , of course , synchronization due to phase locking of an oscillator to common periodic impulses @xcite . another interesting situation is noise - induced synchronization as mentioned above , caused e.g. by common random poisson impulses @xcite . the oscillators can also synchronize when the driving impulses have intermediate statistics between the periodic and the poisson impulses , as shown by yamanobe _ et al . _ @xcite for a model of pacemaker neurons driven by gamma - distributed impulses . this prompts the question , what is the difference between the synchronization due to phase locking and noise - induced synchronization ? though drive - response configuration of the impulse source and the oscillators is the same for both cases , it is expected that there should be some difference in their synchronization dynamics , reflecting different characteristics of the driving impulses . in this paper , we address this issue by considering uncoupled limit - cycle oscillators driven by gamma impulses , which can be smoothly interpolated between regular periodic and random poisson impulses . we examine the transition from phase locking to noise - induced synchronization as the statistics of the gamma impulses is varied , and quantify their difference using phase distributions , lyapunov exponents , and information - theoretic measures . effect of common gamma impulses on limit - cycle oscillators was previously treated in the paper by yamanobe _ et al . _ @xcite , but its main focus was not on the transition between the two types of synchronization but rather on physiologically realistic characterization of pacemaker neurons . we here conduct a systematic , quantitative analysis of the transition between the two different synchronization dynamics . we will reveal that the two types of synchronization dynamics are clearly different in many aspects , e.g. their stability , fluctuations , and statistical dependence on the driving impulses . this paper is organized as follows . in sec . ii , we introduce a model of uncoupled oscillators subject to common gamma impulses and demonstrate its synchronization dynamics for different types of driving impulses . in sec . iii , we analyze phase distributions of the oscillators using frobenius - perron - type equations . in sec . iv , we focus on frequency detuning of the oscillators due to impulsive driving . in sec . v , we examine lyapunov exponents and their fluctuations in the synchronized states . in sec . vi , we characterize mutual relationships among the impulse source and the driven oscillators by utilizing information - theoretic measures . section vii summarizes the results . in this section , we introduce phase oscillators driven by common gamma impulses and demonstrate their synchronization dynamics . based on the standard phase reduction theory of limit cycles @xcite , we can describe a limit - cycle oscillator using only its phase variable , @xmath0 , defined along its stable limit cycle provided that the intervals between the impulses are sufficiently longer than the amplitude relaxation time of the oscillator state to the limit - cycle orbit . the key quantity for this description is the phase response curve ( prc ) @xmath1 of the oscillator @xcite , which encapsulates its dynamical properties . @xmath1 is a periodic function on @xmath2 that gives asymptotic response of the oscillator phase to an impulsive perturbation given at phase @xmath3 . it has been measured experimentally , e.g. , for repetitively firing neurons @xcite . when the limit - cycle oscillator is kicked by an impulse at phase @xmath3 , the phase is mapped to a new value @xmath4 . this function @xmath5 is sometimes called the `` phase transition curve '' in the literature . the dynamics of the oscillator between two successive impulses consists of a jump caused by the first impulse and subsequent free rotation until the arrival of the second impulse @xcite . the oscillator is also subject to small temporal fluctuations of various origins . we assume that each oscillator receives common driving impulses at times @xmath6 , and denote the phase of the oscillator just before the @xmath7th impulse by @xmath8 . the effect of the @xmath7th impulse is to map the oscillator phase @xmath9 to @xmath10 . to incorporate the effect of small fluctuations , we also apply a weak additive independent noise @xmath11 before the mapping . the phase @xmath12 just after the @xmath7th impulse is thus given by @xmath13 . after receiving the @xmath7th impulse , the oscillator rotates with a constant frequency until @xmath14 at which the @xmath15th impulse arrives . therefore , the phase @xmath16 just before the @xmath15th impulse is given by @xmath17 where @xmath18 is the inter - impulse interval , @xmath19 is the frequency of the oscillator , and we have taken modulo @xmath20 of the phase to confine it in @xmath2 . since @xmath21 and @xmath11 are random variables , this equation gives a random map , which we will call a `` random phase map '' hereafter . when we consider multiple oscillators , @xmath22 is common to all oscillators , whereas @xmath23 is different from oscillator to oscillator . note that @xmath11 represents small temporal fluctuations of the dynamics of oscillators , but not static heterogeneities in their natural frequencies . mean frequency of the driven oscillator exhibits qualitatively different dependence on the driving impulse between phase locking and noise - induced synchronization , as we show in sec . iv . as the simplest and typical example , we assume that the prc is given by a sinusoidal function @xmath24 in the following numerical analysis , so that the phase map is given by @xmath25 the parameter @xmath26 controls the magnitude of the nonlinearity , which may also be regarded as the intensity of the impulse . such a prc with both positive and negative lobes are called type - ii @xcite , which generally arises near the hopf bifurcation of limit - cycle oscillators @xcite . another typical example is the type - i prc with a single positive lobe , e.g. @xmath27 , which generally arises near the snic ( saddle - node on invariant circle ) bifurcation of oscillators @xcite . however , difference between these two prcs can roughly be eliminated by simply shifting the frequency @xmath19 and the origin of the phase @xmath3 in the present setup , and qualitatively similar results are expected for both prcs . actually , the two types of prcs yielded almost the same results in our numerical analysis . thus , the type - ii prc , eq . ( [ pm ] ) , gives us general insights and we only present the results for this case hereafter . for ( a ) @xmath28 , ( b ) @xmath29 , and ( c ) @xmath30 . the mean value is fixed at @xmath31 . ] we consider driving impulses that have `` intermediate '' statistics between regular periodic impulses and random poisson impulses . as such impulses , we adopt the gamma impulses @xcite , whose inter - impulse interval @xmath32 obeys the gamma distribution @xcite , @xmath33 where @xmath34 and @xmath35 are real positive parameters . the mean interval is given by @xmath36 . the gamma distribution has the following properties . firstly , when @xmath37 , @xmath38 gives an exponential distribution , @xmath39 which means that the impulses obey the standard poisson random process . secondly , by taking the limit @xmath40 and @xmath41 with @xmath42 fixed , the gamma distribution approaches the delta function , @xmath43 in this limit , the inter - impulse interval is always @xmath44 , so that the impulses become periodic . thus , the parameter @xmath34 , which we call a shape parameter , determines the property of the gamma impulses as shown in fig . [ fig.1 ] . by varying the value of @xmath34 between @xmath20 and @xmath45 with @xmath46 fixed , the gamma impulses can exhibit intermediate properties between random poisson and regular periodic impulses while keeping the same mean inter - impulse interval . we examine uncoupled oscillators driven by gamma impulses and see how synchronization dynamics of the oscillators depends on the shape parameter @xmath34 in the following . the gamma distribution has several nice mathematical properties and naturally arises under a few simple assumptions when the mean and the irregularity of impulse sequences , which respectively correspond to the mean interval @xmath47 and the shape parameter @xmath34 @xcite , are given . we thus use the gamma impulses in the present paper , though alternative approaches for varying the regularity of the driving signal should also be possible , e.g. , by using chaotic dynamical systems @xcite . we here demonstrate synchronization of uncoupled oscillators induced by common impulses for the periodic , poisson , and intermediate cases by direct numerical simulations . we fix the magnitude of nonlinearity ( or the intensity of impulses ) of eq . ( [ pm ] ) at @xmath48 throughout our numerical simulations , which is sufficiently small and therefore the map is not chaotic even for strictly periodic impulses . this is because we are focusing on the synchronization dynamics of limit cycles ( note that the sinusoidal type ii phase map with periodic impulses is nothing but the well - known circle map @xcite ) . strong poisson impulses may also yield positive lyapunov exponents and lead to desynchronization of oscillators as shown in ref . @xcite , but we do not consider such a situation in the present paper . figure [ fig.2 ] shows typical synchronization dynamics of uncoupled oscillators induced by three types of common impulses , where zero - crossing points of @xmath49 oscillators are shown in raster plots . the mean interval of the impulses is set at @xmath50 . the period of the oscillator is also taken as @xmath51 and is thus equals to the mean interval . \(i ) phase locking to periodic impulses [ fig . [ fig.2].(a ) ] . if the impulses are periodic and their intervals are nearly equal to ( or more generally rational multiples of ) the intrinsic rotation period of the oscillators , namely , if they are resonant , each oscillator becomes phase locked to the impulses . as a consequence , uncoupled oscillators driven by common periodic impulses synchronize with each other . \(ii ) noise - induced synchronization by poisson impulses [ fig . [ fig.2].(b ) ] . the oscillators also synchronize when they are driven by common poisson impulses of appropriate intensity . it has been theoretically and experimentally shown that uncoupled limit - cycle oscillators subject to weak common poisson impulses generally synchronize with each other , irrespective of their details @xcite . \(iii ) synchronization induced by gamma impulses [ fig . [ fig.2].(c ) ] . uncoupled oscillators subject to the gamma impulses with intermediate statistics between the periodic and poisson impulses can also synchronize mutually @xcite . thus , the uncoupled oscillators synchronized with each other by the effect of the common impulses for all values of the shape parameter . it is however not easy to catch the quantitative differences in their synchronization dynamics just from these figures . in the following sections , we will characterize the differences for varying types of common impulses using phase distributions , frequency detuning , lyapunov exponents , and information - theoretic measures . ) . ( b ) noise - induced synchronization ( @xmath37 ) . ( c ) synchronization induced by gamma impulses ( @xmath52 ) . ] in this section , we introduce frobenius - perron - type equations @xcite that describe evolution of phase distributions . using them , we examine how the relations among the oscillator phases and the inter - impulse intervals depend on the shape parameter , which reveal differences between phase locking and noise - induced synchronization . let us consider a single - oscillator probability density function ( pdf ) @xmath53 of the phase @xmath54 at time step @xmath7 , just before the @xmath7th impulse arrives . to obtain a frobenius - perron equation describing the dynamics of @xmath55 , it is convenient to consider a pdf @xmath56 of the `` unwrapped '' phase @xmath57 defined in @xmath58 , which obeys the following random phase map without the modulo @xmath20 , @xmath59 the pdf @xmath60 of @xmath61 is given by the following frobenius - perron equation @xcite : @xmath62 where @xmath63 is the pdf of the inter - impulse intervals , namely , the gamma distribution given in eq . ( [ eq : gamma ] ) , and @xmath64 is the probability density function of the independent additive noise , which we assume to be zero - mean gaussian of variance @xmath65 . the probability density function of the wrapped phase , @xmath66 , can be calculated as @xcite @xmath67 thus , from eq . ( [ eq:03 ] ) , the frobenius - perron equation describing @xmath55 is obtained as @xmath68 which describes the single - oscillator pdf @xmath53 of the phase @xmath54 just before the @xmath7th impulse . stationary solutions of eq . ( [ eq : fp1 ] ) gives the pdf of the oscillator phase driven by gamma impulses and additive noise in statistically steady states . in ref . @xcite , we have treated a similar frobenius - perron equation perturbatively to obtain the stationary phase pdf of the oscillator under weak poisson impulses and calculated the lyapunov exponents . in ref . @xcite , doi _ et al . _ analyzed the spectral properties of a similar frobenius - perron equation ( driven by periodic impulses ) to characterize noisy phase locking of a limit - cycle oscillator . to investigate phase relationships between a pair of oscillators ( denoted here as @xmath69 and @xmath70 ) subject to common impulses , we introduce a joint pdf @xmath71 of their phases @xmath72 and @xmath73 just before the @xmath7th impulse . the dynamics of the pair is given by @xmath74 where @xmath75 and @xmath76 are independent additive noise terms ( @xmath22 is common to both oscillators ) . similarly to the single - oscillator case , we can derive a frobenius - perron equation describing the two - oscillator phase pdf @xmath71 as @xmath77 the stationary solution of eq . ( [ eq : fp2 ] ) gives a pdf of the pair of oscillator phases driven by common impulses and independent additive noise in the statistically steady state . in ref . @xcite , we have analyzed a similar two - oscillator frobenius - perron equation by averaging out the fast dynamics of the mean phase to obtain an approximate frobenius - perron equation describing only the phase difference of two oscillators driven by common poisson impulses . since we are interested in pair phase distributions , not only in the distribution of phase differences , we directly solve eq . ( [ eq : fp2 ] ) numerically for gamma impulses in the following . of the phase @xmath3 just before the arrival of each impulse . blue curves are obtained by numerically solving the frobenius - perron equation and red squares are the results of direct numerical simulations of the random phase map . ( a ) noise - induced synchronization ( @xmath37 ) . ( b ) synchronization induced by gamma impulses ( @xmath52 ) . ( c ) phase locking ( @xmath78 ) . ] of the phase pair @xmath79 and @xmath80 just before the impulses . ( a ) noise - induces synchronization ( @xmath37 ) . ( b ) synchronization induced by intermediate impulses ( @xmath52 ) . ( c ) phase locking ( @xmath78 ) . ] of the inter - impulse interval @xmath32 and the phase @xmath3 immediately after this interval . ( a ) noise - induced synchronization ( @xmath37 ) . ( b ) synchronization induced by intermediate impulses ( @xmath52 ) . ( c ) phase locking ( @xmath78 ) . ] we here present stationary phase pdfs obtained by numerically solving the frobenius - perron equations . we set the frequency of each oscillator at @xmath81 . the period of the oscillator is @xmath82 and is equal to the mean period of impulses @xmath83 . thus , when the shape parameter @xmath34 is sufficiently large and the impulses are nearly periodic , phase locking should occur . in the numerical analysis , each variable is discretized using @xmath84-@xmath85 grid points . the calculation cost can be drastically reduced by devising numerical algorithms that use fourier representation in calculating convolutions of the frobenius - perron equations ( see appendix for details ) . figures [ fig.3](a)-(c ) show the stationary pdf @xmath86 of a single oscillator phase @xmath3 just before each impulse for nearly periodic ( @xmath78 ) , intermediate ( @xmath52 ) , and poisson ( @xmath37 ) cases . numerical solutions of the frobenius - perron equation ( [ eq : fp1 ] ) and the results of direct numerical simulations of the random phase map ( [ eqrpm ] ) agree nicely . the intensity of the weak independent gaussian noise is @xmath87 . in all cases , the oscillators synchronize with each other ( to the independent noise level ) . however , depending on the shape parameter @xmath34 , the phase pdf @xmath86 significantly differs . for nearly periodic impulses , @xmath86 has a sharp peak , indicating that the oscillator phase is actually entrained to the driving impulse with some fixed phase shift . in contrast , @xmath86 is roughly uniformly distributed for poisson impulses , which means that the relationship between the oscillator phase and the impulse timing is not fixed and hence not entrained . in the intermediate case , @xmath86 has a broad but still clear one - humped shape , implying that the phase is not completely entrained to the impulses but still possesses a certain level of coherence with respect to the driving impulses . figures [ fig.4](a)-(c ) show the stationary joint phase pdf @xmath88 of a pair of oscillators for nearly periodic ( @xmath78 ) , intermediate ( @xmath52 ) , and poisson ( @xmath37 ) impulses obtained by numerically solving the frobenius - perron equation ( [ eq : fp2 ] ) . in all cases , the oscillators synchronize with each other , so that their phases are distributed along the diagonal line , @xmath89 . however , their distribution strongly depends on the shape parameter @xmath34 . for nearly periodic impulses ( @xmath78 ) , the oscillator is almost phase - locked to the impulses and thus the pdf has a sharp peak near the center , indicating that the oscillators are not only mutually synchronized but both of them are entrained to the impulses . as the parameter @xmath34 decreases , the distribution becomes broader , and for poisson impulses ( @xmath37 ) , the phases @xmath90 and @xmath91 are broadly distributed along the diagonal line , indicating that they are synchronized but not entrained by the impulses anymore . to analyze how the driving impulses affect the oscillator phase , we also calculate the joint pdf of the inter - impulse interval and the oscillator phase just after this interval in the statistically steady states . the stationary joint pdfs @xmath92 of the impulse interval @xmath32 and the phase @xmath3 observed immediately after this interval are calculated from the frobenius - perron equation . figures [ fig.5](a)-(c ) show stationary joint pdfs @xmath93 for nearly periodic ( @xmath78 ) , intermediate ( @xmath52 ) , and poisson ( @xmath37 ) cases . the distribution has a sharp peak in the nearly phase - locked case ( @xmath78 ) , indicating that the inter - impulse interval and the oscillator phase just after this interval have almost a one - to - one correspondence , namely , the oscillator phase is almost entrained by the impulses . as we decrease the shape parameter @xmath34 , the distribution gradually elongates , and in the poisson limit ( @xmath37 ) , such clear correspondence is lost ( but they still retain a certain degree of correlation ) . thus , all phase distributions clearly capture the essential difference between the phase locking and the noise - induced synchronization . the relation among the oscillator phases and the inter - impulse intervals significantly differs depending on the shape parameter @xmath34 , even though the oscillators themselves are always mutually synchronized . as @xmath34 is increased , the oscillators tend to be more strictly phase locked to the driving impulses , whereas for smaller @xmath34 , their dependence becomes weaker . this observation will be quantified by information - theoretic measures in sec . in this section , we consider how the dynamics of the oscillator is modulated by the driving impulses . specifically , we examine the frequency detuning @xcite , i.e. , the difference between the mean frequency of the driven oscillator and that of the driving impulses , for varying values of the shape parameter . this analysis will reveal another clear difference between the two types of synchronization scenarios . from the phase pdf @xmath86 and the prc @xmath1 , we can estimate the mean frequency of the impulse - driven oscillator as follows . using the random phase map for unwrapped phase @xmath57 , eq . ( [ eq : rpm2 ] ) , the oscillator phase just before the @xmath94th impulse is given by @xmath95 therefore , long - time mean frequency @xmath96 of the driven oscillator can be calculated from this equation as @xmath97 where we have replaced the long - time average by a statistical average and used the periodicity of the prc @xmath1 and the normalization condition @xmath98 . thus , the mean frequency detuning between the oscillator and the impulses is given by @xmath99 where @xmath100 is the mean frequency of the driving impulses . we fix the mean frequency of the driving impulses at @xmath101 and vary the shape parameter @xmath34 and the natural frequency @xmath19 of the oscillator in the numerical simulations . figure [ fig.6](a ) plots the frequency detuning @xmath102 on the @xmath103 plane . for large @xmath34 and nearly resonant natural frequency , @xmath104 , the conditions for phase locking are satisfied and therefore the mean frequency @xmath96 of the oscillator is locked to that of the driving impulses @xmath105 , yielding a plateau in the figure . in contrast , for small @xmath34 , the mean frequency @xmath96 never locks to @xmath105 . figure [ fig.6](b ) plots @xmath102 as a function of @xmath19 at @xmath78 , i.e. , for approximately periodic impulses . results of direct numerical simulations agree well with the theoretical estimate , eq . ( [ eq : freqdet ] ) , with the pdf @xmath86 obtained numerically from the frobenius - perron equation . a plateau satisfying @xmath106 can clearly be seen near @xmath107 . in the intermediate case ( @xmath52 ) shown in fig . [ fig.6](c ) , clear phase - locking region no longer exist , however , the frequency of the oscillator is modulated by the driving impulses . in the poisson case ( @xmath37 ) , the mean frequency of the driven oscillator is not affected by the driving impulse as shown in fig . [ fig.6](d ) . this observation provides us with further evidence on the difference between the two synchronization scenarios . actually , as pointed out by yoshimura _ _ in ref . @xcite , absence of frequency locking between the oscillators is characteristic to common - noise - induced synchronization , in sharp contrast to ordinary phase locking due to mutual coupling ( though they considered weak gaussian noise instead of random impulses in their work ) . on the shape parameter @xmath34 and on the natural frequency @xmath19 of the driven oscillator , measured by direct numerical simulations of the random phase map . ( b)-(d ) dependence of the frequency detuning on the natural frequency @xmath19 of oscillators for several values of the shape parameter @xmath34 . blue solid lines are the results of the integral calculation , eq.([eq : freqdet ] ) , and red squares are obtained by direct numerical simulations of the random phase map . ( b ) phase locking ( @xmath78 ) . ( c ) synchronization induced by gamma impulses ( @xmath52 ) . ( d ) noise - induced synchronization ( @xmath37 ) . ] in this section , to characterize the synchronization dynamics , we focus on lyapunov exponents and their fluctuations in the synchronized states . as we will see , differences between synchronization scenarios are well characterized by the fluctuations of the lyapunov exponent . to quantify statistical linear stability of the synchronized states , we calculate the lyapunov exponent and its variance . let us consider a pair of oscillators and denote their phases at time @xmath7 as @xmath54 and @xmath108 , respectively . linearized evolution of a small phase difference @xmath109 is given by @xmath110 where @xmath111 is the instantaneous linear growth rate of the phase difference . the deviation @xmath112 at large time step @xmath94 is thus given by @xmath113,\end{aligned}\ ] ] where we have introduced the lyapunov exponent @xmath114 , defined by a long - time average of the linear growth rates as @xmath115 in the last expression , we have replaced the long - time average by a statistical average over the stationary phase pdf @xmath86 and the pdf @xmath116 of the independent noise . the synchronized state is stable if @xmath117 is negative . similarly , variance of the linear growth rates , @xmath118 , can be calculated as @xmath119 as we will see , even if the lyapunov exponent @xmath114 takes the same value , its variance @xmath120 can significantly differ depending on the shape parameter @xmath34 . on the shape parameter @xmath34 and on the frequency @xmath19 of oscillators . ( a ) mean lyapunov exponent @xmath117 . ( b ) variance of lyapunov exponent @xmath120 . ( c ) ratio of the mean to the variance , @xmath121 . ] on the parameter @xmath34 . the curves are calculated from the frobenius - perron equation and the symbols are the results of direct numerical simulations of the random phase map . ( a ) mean @xmath117 . ( b ) variance @xmath120 . ] figure [ fig.7](a ) plots the lyapunov exponent @xmath114 on the @xmath103 plane , showing its dependence on the shape parameter @xmath34 and on the natural frequency of the oscillator @xmath19 . the mean period of the impulses is fixed at @xmath122 . for sufficiently large @xmath34 , there exists bell - shaped regions near the resonant frequencies ( i.e. , @xmath81 and @xmath123 ) , where the lyapunov exponent takes large negative values . these regions correspond to nearly phase - locked states and may be seen as a kind of arnold tongues . the lyapunov exponent also takes small negative values outside this region , which corresponds to the noise - induced synchronized states . note that phase locking occurs only when the oscillator frequency is approximately resonant to the impulses , whereas the noise - induced synchronization occurs almost everywhere . dependence of the lyapunov exponent @xmath124 on the shape parameter @xmath34 for several values of @xmath19 is plotted in fig . [ fig.8](a ) . when the natural frequency is resonant to the mean inter - impulse interval ( @xmath125 ) , @xmath117 smoothly decreases as we move from the noise - induced synchronization ( @xmath37 ) to the phase - locking ( @xmath78 ) , indicating that the phase locking is more stable than the noise - induced synchronization . however , for oscillator frequencies near the edges of the phase - locking region ( @xmath126 and @xmath127 ) , @xmath117 differs only little between @xmath37 and @xmath78 , so that the noise - induced synchronization can be nearly stable as the phase locking . thus , though the lyapunov exponent @xmath117 reflects the synchronization dynamics , difference between the two types of synchronization can not be fully characterized just by looking at @xmath117 . figure [ fig.7](b ) shows dependence of the variance of the lyapunov exponent exponent @xmath120 on the shape parameter @xmath34 and on the oscillator frequency @xmath19 , and figure [ fig.8](b ) shows the dependence on the shape parameter @xmath34 for @xmath128 , and @xmath127 . in contrast to the lyapunov exponent itself , the variance @xmath120 always decreases considerably as @xmath34 is increased from the noise - induced synchronization ( @xmath37 ) to the phase - locking ( @xmath78 ) , because the phase pdf becomes much broader for smaller @xmath34 as we have seen in fig . [ fig.3 ] . thus , the difference between the phase locking and the noise - induced synchronization can be captured more clearly by the fluctuations in the lyapunov exponent , rather than by its long - time average . the difference can also be clearly visualized by plotting the ratio of the lyapunov exponent and its variance @xmath121 as in fig . [ fig.7](c ) and fig . [ fig.8](c ) . actually , this ratio characterizes the intermittent dynamics of the phase differences between the oscillators as we explain below . fluctuations of the lyapunov exponent manifests itself in the dynamics of the phase difference . in fig . [ fig.9 ] , typical time sequences of the phase difference @xmath129 between the two oscillators in the synchronized states are shown . three different shape parameters ( @xmath130 ) yielding nearly the same lyapunov exponent @xmath117 are chosen ( with @xmath126 fixed ) . in noise - induced synchronization ( fig . [ fig.9](a ) ) with small @xmath34 , the phase difference often exhibits big bursts , because the growth rate fluctuates strongly ( large @xmath120 ) . as the parameter @xmath34 increases , variation of the phase difference becomes smaller ( fig . [ fig.9](b ) ) and , in the phase - locked state , the phase difference stays small and rarely exhibits large bursts ( fig . [ fig.9](c ) ) , because fluctuations of @xmath131 are very weak ( small @xmath120 ) . the peculiar dynamics shown in fig . [ fig.9](a ) is a direct consequence of the fluctuations of the linear growth rates and intimately related to the well - known on - off ( or modulational ) intermittency @xcite . figure [ fig.10 ] shows the stationary pdfs of phase differences between the oscillators for differing values of the shape parameter @xmath34 in normal scales ( a ) and in log - log scales ( b ) . reflecting the degree of the fluctuations of @xmath131 , the width of the pdfs varies widely . the tail part of the distribution exhibits power - law decay , which is broader for smaller @xmath34 , indicating that the phase differences exhibit big bursts ( phase differences ) much more frequently . as shown in ref . @xcite , the exponent of the power - law tail is approximately given by the ratio @xmath132 that we plotted fig . [ fig.7](c ) . thus , degree of intermittency in the synchronization dynamics also characterizes the difference in noise - induced synchronization and phase locking clearly . at three different values of the shape parameter @xmath34 . ( a ) noise - induced synchronization ( @xmath37 ) . ( b ) synchronization induced by intermediate impulses ( @xmath52 ) . ( c ) phase locking ( @xmath78 ) . ] of the phase difference @xmath133 of two oscillator calculated by direct numerical simulations of the random phase map . black solid line , blue dashed line and red dotted line plot the results for @xmath134 and @xmath135 , respectively . ( a)@xmath136 . ( b)@xmath136 using logarithmic scaling on the vertical axis . ] we have observed that the joint pdfs of the oscillator phases and the impulse intervals exhibit distinct characteristics depending on the shape parameter . in this section , to quantify correlations between the oscillator phases and the impulse intervals for different types of synchronization dynamics , we introduce information measures . in particular , we focus on the mutual dependence or `` causality '' of a pair of phase oscillators driven by a common impulse source . this leads us to physically distinct interpretations of the two synchronization scenarios . the mutual information @xmath137 @xcite is a nonnegative measure that quantifies mutual dependence of two random variables @xmath138 and @xmath139 @xcite . it is defined as @xmath140 where @xmath141 is a joint probability distribution of @xmath138 and @xmath139 , and @xmath142 and @xmath143 are probability distributions of @xmath138 and @xmath139 , respectively . we often normalize @xmath144 by its upper limit as @xmath145 where @xmath146 and @xmath147 are entropies of @xmath138 and @xmath139 , respectively . the normalized mutual information satisfies @xmath148 . assuming , for example , that @xmath138 represents the inter - impulse interval @xmath32 and @xmath139 the oscillator phase @xmath3 just after this interval , @xmath137 or @xmath149 quantifies how much the oscillator phase tells us about the impulses . the interaction information @xmath150 @xcite is a generalization of the mutual information @xmath137 to three random variables @xmath138 , @xmath139 , and @xmath151 , and quantifies mutual dependence among them . it is defined as @xmath152 where @xmath153 is a joint probability distribution function of @xmath138 , @xmath139 , and @xmath151 . @xmath150 is symmetric with respect to permutations of @xmath138 , @xmath139 , and @xmath151 . the first term is simply the mutual information @xmath137 between @xmath138 and @xmath139 , and the second term on the right hand side gives conditional mutual information @xmath154 between @xmath138 and @xmath139 given @xmath151 . thus , @xmath150 can be expressed as @xmath155 which provides us with some intuitive meaning of @xmath150 . namely , it measures the effect of knowing @xmath151 in guessing the mutual dependence of @xmath138 and @xmath139 . unlike @xmath137 , @xmath150 can take both positive and negative values . the sign of @xmath150 tells us about how the three random variables @xmath138 , @xmath139 , and @xmath151 depend on each other , as illustrated in fig . [ fig.11 ] . \(i ) if @xmath156 , at least one of the three random variables is independent of the others . for example , knowing @xmath151 has no effect in guessing the mutual information between @xmath138 and @xmath139 , @xmath157 . all variables are mutually dependent if @xmath158 . \(ii ) if @xmath159 , one of the three variables tends to determine the other two variables . for example , knowing @xmath151 lowers the information on @xmath138 gained by observing @xmath139 ( or vice versa ) , i.e. , @xmath160 . in particular , the maximum value of @xmath150 , given by @xmath161 , is attained when one of the three variables dominates the other two , for example , as @xmath162 and @xmath163 where @xmath164 and @xmath165 are some maps . \(iii ) if @xmath166 , knowing e.g. @xmath151 helps in gaining information on @xmath138 by observing @xmath139 , so that @xmath167 . in particular , @xmath150 takes its maximum value when two of the random variables dominate the remaining one , for example , when @xmath168 with some map @xmath169 and moreover @xmath170 ( @xmath139 and @xmath151 are independent ) . the maximum value is equal to the negative conditional entropy @xmath171 of the dominated variable @xmath151 given the dominating variable @xmath138 . we normalize the interaction information @xmath150 by its maximum as @xmath172 so that @xmath173 is satisfied . in addition to the above normalization , @xmath150 may also be normalized by the mutual information @xmath137 as @xmath174 the quantity @xmath175 satisfies @xmath176 ( because @xmath177 in the present case as shown below ) , quantifies how much information @xmath151 conveys about the relationship between @xmath138 and @xmath139 . we use these normalized @xmath178 and @xmath175 to quantify mutual dependence among @xmath32 , @xmath179 , and @xmath180 . . each arrow indicates the dependence between the variables . ] the mutual information and the interaction information are calculated from the pdfs obtained by numerically solving the frobenius - perron equations ( [ eq : fp1 ] ) and ( [ eq : fp2 ] ) . we discretize the inter - impulse intervals and the two oscillator phases using @xmath84 grid points and use the resulting coarse - grained discrete probability distributions in the calculation . note that we need the joint probability distribution @xmath181 to calculate the interaction information , which corresponds to @xmath182 grid points . such a calculation is only feasible by the frobenius - perron approach . figure [ fig.12](a ) shows the mutual information @xmath183 of the inter - impulse interval @xmath32 and the phase @xmath3 just after this interval , as well as its upper limit @xmath184 ( entropy of the inter - impulse intervals ) as functions of the shape parameter @xmath34 for the resonant situation , @xmath185 . @xmath184 takes its maximum value in the poisson limit ( @xmath37 ) , monotonously decreases as the shape parameter @xmath34 is increased , and almost vanishes for nearly periodic impulses ( @xmath186 ) . the raw mutual information @xmath183 takes a small value for poisson impulses ( @xmath37 ) and gradually increases with the shape parameter @xmath34 . @xmath183 takes its maximum value at @xmath187 , then decreases again , and almost vanishes when the impulses become nearly periodic ( @xmath186 ) . figure [ fig.12](b ) shows the normalized mutual information @xmath188 . it increases smoothly from a small value ( nearly @xmath189 ) to @xmath20 as the shape parameter @xmath34 is increased from @xmath37 ( poisson ) to @xmath186 ( nearly periodic ) . this indicates that the oscillator phase has only little dependence on the inter - impulse interval just before it is measured in the poisson case , whereas the oscillator phase possesses almost complete information about the interval in the periodic case . these results are consistent with the shape of the single - oscillator phase pdf shown in fig . [ fig.3 ] . it is interesting to note that the raw mutual information @xmath183 has a peak at the intermediate shape parameter @xmath34 . if we regard our model as an information channel , the output oscillator phase conveys information about the input impulse interval most efficiently at this value . this can be interpreted as follows . for periodic impulses , even though @xmath3 is locked to @xmath32 , the interval @xmath32 has no information because @xmath190 and therefore @xmath191 . for poisson impulses , though @xmath184 is relatively large , @xmath3 does not faithfully represent @xmath32 . as a compromise , @xmath183 is maximized at the intermediate value of @xmath34 . ( black solid line ) and @xmath184 ( red dashed line ) . ( b ) normalized mutual information @xmath192 . ] more interesting insight can be attained by examining the interaction information @xmath193 among the inter - impulse interval and the phases of the two driven oscillators . figure [ fig.13](a ) shows the interaction information @xmath194 , its upper limit @xmath184 , and mutual informations @xmath195 and @xmath196 as functions of the shape parameter @xmath34 for @xmath81 ( resonant ) . the intermediate peak of the raw interaction information arises due to the same reason as in the previous case of the raw mutual information . interaction informations normalized in three ways , @xmath197 , @xmath198 , and @xmath199 , are plotted against @xmath34 in fig . [ fig.13](b ) . the normalized interaction information @xmath200 decreases as the shape parameter is decreased from @xmath201 ( nearly periodic ) to @xmath37 ( poisson ) . for nearly periodic impulses , @xmath202 is nearly @xmath20 , implying that @xmath32 , @xmath79 , and @xmath80 are significantly dependent on each other in the phase - locking regime . as @xmath34 is decreased , @xmath202 gradually decreases and , for poisson impulses , @xmath202 almost vanishes , indicating that at least one of the @xmath79 , @xmath80 , and @xmath32 is almost independent of the others . the normalized interaction information @xmath203 also decreases as the shape parameter is decreased from @xmath201 ( nearly periodic ) to @xmath37 ( poisson ) . for nearly periodic impulses ( @xmath201 ) , @xmath204 is nearly @xmath20 , which implies that the inter - impulse interval @xmath32 dominates the oscillator phases @xmath79 and @xmath80 . as @xmath34 is decreased , @xmath204 decreases gradually , and for poisson impulses , it becomes very close to @xmath189 , namely , @xmath32 is almost independent of @xmath79 and @xmath80 and thus @xmath32 tells very little about @xmath79 and @xmath80 . the last normalized interaction information @xmath205 also decreases as the shape parameter is decreased from @xmath201 ( nearly periodic ) to @xmath37 ( poisson ) . for nearly periodic impulses ( @xmath201 ) , @xmath206 is nearly @xmath20 , which indicates that the phase @xmath80 has almost complete information about @xmath32 and @xmath79 . it means that all @xmath32 , @xmath79 and @xmath80 depend on each other strongly . unlike @xmath204 , @xmath206 takes relatively large values around @xmath207 even in the poisson limit , which implies that we can get nearly @xmath208 of the information about @xmath32 and @xmath79 just by looking at @xmath80 . this is because the oscillators are synchronized even if they are driven by poisson impulses , and we can learn much about @xmath79 by knowing @xmath80 . though the interaction information itself is symmetric with respect to its three arguments , we can obtain further insight by taking into account the drive - response configuration of our model , namely , the fact that the inter - impulse interval @xmath32 has a distinct meaning from the other two phase variables @xmath79 and @xmath80 . therefore , for the phase - locked situation with @xmath209 , we may consider that the phases @xmath79 and @xmath80 are predominantly controlled by the driving impulse @xmath32 , as implied from the condition giving the maximum value of @xmath194 . on the other hand , in the poisson case with @xmath202 close to @xmath189 , we may conclude that @xmath32 is nearly independent of @xmath79 and @xmath80 . this indicates that the phases have very little information about the driving impulse in the noise - induced synchronized state . the above results lead us to two distinct interpretations of the two synchronization scenarios , as schematically summarized in fig . [ fig.14 ] ; in the phase locking , the two oscillators are simply oscillators dominated by the impulses , whereas they are almost free from the impulses but still behave synchronously in the noise - induced synchronization . ( black solid line ) and normalization factors @xmath184 ( red dashed curve ) , @xmath195 ( blue dot - dashed curve ) and @xmath196 ( green dotted curve ) plotted against the shape parameter @xmath34 . ( b ) normalized interaction information . @xmath210 ( red dashed curve ) , @xmath211 ( blue dot - dashed curve ) , and @xmath212 ( green dotted curve ) ] in the present study , we analyzed uncoupled oscillators driven by gamma impulses that smoothly interpolate between periodic and poisson inter - impulse intervals to quantify the difference in synchronization dynamics between these two limiting cases . by examining the dependence of various quantities on the shape parameter of the gamma impulses , we have revealed a clear difference between the two types of synchronization . using phase distributions , frequency detuning , statistics of lyapunov exponents , and information - theoretic measures , we have quantitatively confirmed our original intuition , namely , that the oscillators are principally dominated by the impulses and mutually synchronized as its consequence for the periodic driving , whereas the oscillators are not entrained by the impulses ( at least not strongly influenced by the inter - impulse interval just before the phase is measured ) but they still attain coherence in the case of poisson driving . generalization of our results for uncoupled oscillators to mutually interacting oscillators , in particular with delay , under common or correlated gamma impulses is an interesting future subject , where not only synchronization but qualitatively different behaviors @xcite can be expected . more detailed quantification of the input - output and the output - output relations of the impulse - driven oscillators from the viewpoint of information transfer would also be interesting , where effect of the functional shape of the phase response curve and its optimization would be important issues . as we found in the measurement of mutual and interaction information , the transition between the phase locking and the noise - induced synchronization may not be simply monotonic from such a viewpoint , implying the importance of intermediate randomness of the input impulses . we plan to push forward in these directions in our future studies . we thank h. suetani and h. hata ( kagoshima ) , and y. tsubo and j. teramae ( riken ) for illuminating comments . this work is supported by mext ( grant no . 22684020 ) and by the gcoe program `` the next generation of physics , spun from universality and emergence '' from mext , japan . 99 r. roy and k.s . thornburg , jr . lett . * 72 * , 2009 ( 1994 ) ; a. uchida , r. mcallister , and r. roy , phys . lett . * 93 * , 244102 ( 2004 ) ; a. uchida , k. yoshimura , p. davis , s. yoshimori , and r. roy , phys . e * 78 * , 036203 ( 2008 ) . k. yoshida , k. sato , a. sugamaga , j. sound and vibration * 290 * , 34 ( 2006 ) . k. arai and h. nakao , phys . e * 77 * , 036218 ( 2008 ) . k. nagai and h. nakao phys . e * 79 * , 036205 ( 2009 ) . z. f. mainen and t. j. sejnowski , science * 268 * , 1503 ( 1995 ) . m. d. binder and r. k. powers , journal of neurophysiology * 86 * , 2266 ( 2001 ) . r. f. galn , n. f. trocme , g. b. ermentrout , and n. n. urban , j. neurosci * 26(14 ) * , 3646 ( 2006 ) . p. danzl , r. hansen , g. bonnet , and j. moehlis , journal of computational neuroscience , * 25 * , 141 ( 2008 ) . g. b. ermentrout , r. f. galn , and n. n. urban , trends neurosci . * 31 * , 428 ( 2008 ) . t. royama , _ analytical population dynamics _ ( chapman and hall , london , uk , 1992 ) . e. ranta , v. kaitala and e. helle , oikos * 78 * , 136 ( 1997 ) . w. d. koenig and j. m. h. knops , nature * 396 * , 225 ( 1998 ) . p. a. p. moran , aust . * 1 * , 291 ( 1953 ) . a. t. winfree , _ the geometry of biological time _ ( springer - verlag , new york , 2001 ) . y. kuramoto , _ chemical oscillation , waves , and turbulence _ ( springer - verlag , tokyo , 1984 ) ( republished by dover , new york , 2003 ) . a. pikovsky , m. rosenblum , and j. kurths , _ synchronization _ ( cambridge university press , england , 2001 ) . l. glass and m. c. mackey , _ from clocks to chaos : the rhythms of life _ ( princeton university press , usa , 1988 ) . t. yamanobe and k. pakdaman , biol . cybern . * 86 * , 155 ( 2002 ) . h. nakao , k. arai , k. nagai , y. tsubo , and y. kuramoto , phys . e * 72 * , 026220 ( 2005 ) . r. f. galn , g. b. ermentrout and n. n. urban , phys . lett . * 94 * , 158101 ( 2005 ) . t. tateno and h. p. c. robinson , biophysical journal * 92 * , 683 ( 2007 ) . k. arai and h. nakao , phys . e * 78 * , 066220 ( 2008 ) . d. hansel , g. mato , and c. meunier , neural computation * 7 * , 307 ( 1995 ) . a. abouzeid and g. b. ermentrout , phys . e * 80 * , 011911 ( 2009 ) . s. marella and g. b. ermentrout , phys . e * 77 * , 041918 ( 2008 ) . e. brown , j. moehlis , and p. holmes , neural computation * 16 * , 673 ( 2004 ) . e. izhikevich , _ dynamical systems in neuroscience . the geometry of excitability and bursting . _ the mit press , 2007 . g. b. ermentrout , neural computation * 8 * , 979 ( 1996 ) . t. shimokawa , s. koyama and s. shinomoto , j. comp . doi : 10.1007/s10827 - 009 - 0194-y ( 2010 ) . g. j. escalera santos and p. parmananda , phys . e , * 65 * , 067203 ( 2002 ) . s. doi , j. inoue and s. kumagai , j. stat . phys . * 90 * , 1107 ( 1998 ) . a. lasota and m. c. mackey , _ probabilistic properties of deterministic systems _ ( cambridge university press , cambridge , 1985 ) . e. ott , _ chaos in dynamical systems _ ( cambridge university press , cambridge , 2002 ) . g. b. ermentrout and d. saunders , j. comput . neurosci . * 20 * , 179 ( 2006 ) . k. yoshimura , p. davis , and a. uchida , in proceedings of the 2007 international symposium on nonlinear theory and its applications , vancouver , canada ( ieice , vancouver , 2007 ) , p. 104 - 107 . h. fujisaka and t. yamada , prog . . phys . * 69 * , 32 ( 1983 ) ; h. fujisaka , prog . . phys . * 70 * , 1264 ( 1983 ) ; h. fujisaka and t. yamada , prog . of theor . phys . * 74 * , 918 ( 1985 ) . a. s. pikovsky , phys . a * 165 * , 33 ( 1992 ) . y. kuramoto and h. nakao , phys . lett . * 78 * , 4039 ( 1997 ) . t. a. cover and j. a. thomas , _ elements of information theory _ ( wiley , 1991 ) . w. j. mcgill , psychometrika , * 19 * , 97 ( 1954 ) . t. tsujishita , advances in applied mathematics * 16 * , 269 ( 1995 ) . k. k. lin , e. shea - brown , and l .- s . young , journal of nonlinear science * 19 * , 1432 ( 2009 ) . here we introduce fourier representation of the frobenius - perron equations ( [ eq : fp1 ] ) and ( [ eq : fp2 ] ) . based on this representation , we can use the fast fourier transform ( fft ) algorithm to numerically calculate convolution integrals included in the equations and drastically reduce calculation costs . firstly , to numerically convolute @xmath213 and @xmath214 , we transform the dirac delta function in the frobenius - perron equation , in such a way that it becomes an explicit function of the variable of integration @xmath215 rather than the original phase variable @xmath3 . @xmath216 where we have defined an inverse function of the phase map , @xmath217 since @xmath218 is a phase map , @xmath219 holds for @xmath220 , where @xmath221 is an integer . using this , we obtain @xmath222}{d\theta } r\left ( \mathcal{f}(\theta-\omega\tau)+d - k-\theta ' \right)\\ = & \int_0 ^ 1 d\theta ' \int_0^\infty d\tau w(\tau)p_{n}(\theta ' ) \frac{d\mathcal{f}(\theta-\omega\tau)}{d\theta } \sum_{k=-\infty}^{\infty } r\left ( \mathcal{f}(\theta-\omega\tau)-k-\theta ' \right)\\ = & \int_0 ^ 1d\theta ' \int_0^\infty d\tau w(\tau)p_{n}(\theta ' ) \frac{d\mathcal{f}(\theta-\omega\tau)}{d\theta } s\left ( \mathcal{f}(\theta-\omega\tau)-\theta ' \right),\end{aligned}\ ] ] where @xmath223 now , using the discrete fourier transformation , we can convolute @xmath213 and @xmath214 as @xmath224 where @xmath225 and @xmath226 are @xmath227th fourier coefficients of @xmath213 and @xmath214 , respectively : @xmath228 and @xmath229 next , we need to convolute the functions @xmath230 and @xmath231 . from the definition of the gamma distribution , eq . ( [ eq : gamma ] ) , @xmath232 holds . using this , we obtain @xmath233 where @xmath234 and @xmath235 are @xmath236th fourier coefficients of @xmath237 and @xmath238 , respectively : @xmath239 therefore , we can write the evolution of fourier coefficient of @xmath55 as @xmath240 which can easily be calculated numerically . similarly , we can write the evolution of @xmath241 , which is the @xmath242th fourier coefficient of the joint probability density function @xmath243,as @xmath244 which is also easy to calculate numerically .

this paper addresses the problem of interacting electrons in a band of a two - dimensional crystal . this problem uncovers all the macroscopic properties of the materials of which the maximum interest is centered at the present on the high-@xmath1 superconductors , in particular , on the nature of their normal state . to fix ideas and some notation let us review the highlights of the problem . start with a free theory of , say , electrons in a metal , i. e. , a free fermi gas whose ground state is known and whose low energy elementary excitations are electrons and holes obeying a free dispersion relation . then switch on an interaction among the electrons and ask the question of what the new ground state is and what is the nature of the low energy excitations of the full interacting theory . this program has been successfully addressed by landau in three spatial dimensions @xmath2 where it leads to the famous theory of landau s fermi liquid @xcite and the concept of quasiparticles . things are different in @xmath3 where an exact solution can be found and the behavior of the system is always of the luttinger type@xcite . the case @xmath4 is physically very interesting both in practice ( most of the high-@xmath1 materials are organized in two dimensional layers and experimental data are available ) and under a theoretical point of view . despite the great effort devoted to it , no exact solution is known at the present and the situation remains controversial superconductivity . ] . the ultimate origin of the trouble lies on the very complex nature of the many - body vacuum and the corresponding difficulties of doing perturbations around the fermi surface . the theory of the fermi liquid and the perturbations of it that lead to the bcs and charge density wave instabilities have recently been rephrased as effective field theories @xcite to which renormalization group techniques can be applied @xcite . the main idea of this approach is that , although we are dealing with problems that naturally have an ultraviolet cutoff @xmath5 ( the width of the band ) , we are only interested in the physics at a scale much lower than it , namely , a very narrow strip above and below the fermi surface . by integrating out in the path integral all the momenta above a certain cutoff @xmath6 , we end up with an effective field theory whose couplings are cutoff dependent . the renormalization group transformations are rescalings of @xmath7 . in this way one can visualize the fermi liquid as a fixed point of this rg transformations ( it is gapless ) that flows towards a bcs state when a weak attraction is introduced between the quasiparticles . when comparing the effective field theories that arise in this context with standard quantum field theory , we encounter two main differences . the first one is the nonrelativistic nature of the many body physics where there is a parameter , the fermi velocity , whose typical value is much smaller than the speed of light . this prevents the use of the general theorems concerning renormalizability . the more severe difference lies in the above mentioned nature of the many body vacuum . when treating the interacting system as a perturbation of the free " system constituted by the fixed point , perturbations are to be described around a fermi surface in @xmath8 , a fermi line in @xmath9 , and fermi points in @xmath3 . we are then faced with couplings that are functions of the shape of the fermi surface which is also modified by the interaction . only in the @xmath3 case we deal , as in quantum field theory , with a true vacuum around which one can expand and get a bunch of coupling _ constants_. in ref . @xcite , we developed a model for the electronic structure of free electrons in a two - dimensional honeycomb lattice . this lattice constitutes the basic building block of periodic systems made of identical atoms with planar , threefold coordination . that is the case of graphite , built of carbon atoms . the rapidly growing family of fullerene compounds @xcite can be thought of as closed , curved surfaces derived from the planar honeycomb lattice . these compounds show a variety of unexpected features , the most striking of them being the superconductivity of doped crystals of c@xmath10 molecules @xcite . moreover , this superconductivity can be considered a form of high - t@xmath11 superconductivity , as the number of electrons available to form cooper pairs is unusually low . the main feature that makes the honeycomb lattice the object of our attention in the present context is that its fermi level at half - filling consists of exactly two fermi _ points_. this peculiarity opens the possibility of performing a complete analysis of the problems mentioned above in analogy with the @xmath3 case . in particular it makes of this lattice an ideal laboratory for investigating departures from fermi liquid behavior in @xmath9 which have shown very ellusive in more conventional approaches . the major interest of this investigation relies on the fact , made clear by now from the experiments , that the normal state of high-@xmath1 materials is not of fermi liquid type . besides , it can be directly applied to the physics of the fullerenes that , as mentioned above , are two - dimensional graphitic structures at half - filling @xcite . the paper is organized as follows . in section 2 we review results on the free electrons on the honeycomb lattice that lead to the low energy effective action used as a starting point of our analysis . we emphasize the point that , unlike what usually occurs in many body physics , our effective action constitutes a genuine quantum field theory with a single , well - defined , vacuum state . next the free theory is completed with the addition of the coulomb interaction . in the field theory , this is mediated by an electromagnetic field whose propagator reproduces , in the nonrelativistic limit , the usual four - fermion interaction considered in the many - body approach . we end up with a standard local quantum field theory . a discussion is done on the differences to be expected by the use of the relativistic versus the instantaneous electromagnetic field propagator illustrated by the computation of the density of states in both cases . the results announce an infrared instability which is clarified in section 4 with the help of the renormalization group . section 3 is devoted to the study of the renormalizability of the model . it is shown that , despite its nonrelativistic nature and due to its gauge invariance , it is renormalizable . the fermi velocity and the electron wave function get renormalized while the electric charge does not . sections 4 and 5 contain the analysis of the rg flow of the model and its physical implications . the first section is devoted to the nonrelativistic limit of the theory . we see that the perturbative expansion has an effective coupling given by the ratio of the square of the electric charge and the fermi velocity . the rg shows that this coupling decreases to zero in the infrared although there is no fixed point in the nonrelativistic regime . we then investigate the possibility of having a natural length scale in the theory and make some comments on the system away from half - filling . finally we find an anomalous dimension for the electron wave function at two - loop order that points towards a non - fermi liquid behavior of the system . the previous analysis is completed in section 5 where we study the relativistic regime which is the relevant regime in the infrared . we find a fixed point of the rg flow obtained when the fermi velocity equals the speed of light . next we compare our model with standard rg analysis of the fermi liquid theory and find significant differences both from the technical and from the physical point of view . the most significant one is the anomalous dimension found for the electron green function . we discuss the non renormalization of the electric charge and establish the non - fermi liquid nature of our fixed point . we set our main conclusions and a summary of the more relevant points of the article in section 6 . in this section we first review the one - particle electronic states in the two - dimensional honeycomb lattice , which serve afterwards as a starting point for the many - body problem . the present one - particle description has been worked out with complete detail in ref . @xcite , regarding the conduction band of a two - dimensional graphite layer . the honeycomb lattice has two different atoms per primitive cell , so that two degrees of freedom arise in a variational computation of the energy eigenstates , corresponding to the respective orbitals for the two atoms . in the tight - binding approximation the problem reduces to the diagonalization of the one - particle hamiltonian @xmath12 where the sum is over pairs of nearest neighbors atoms @xmath13 on the lattice and @xmath14 , @xmath15 are canonically anticommuting operators @xmath16 the states which are simultaneously eigenvectors of the hamiltonian and of the lattice generators have to be of the form @xmath17 with coefficients @xmath18 and @xmath19 for black and blank points , respectively , as depicted in figure 1 . obviously , under the action of ( [ 6 ] ) black points are mapped into blank points and viceversa . we have , indeed , @xmath20 where @xmath21 is a triad of vectors connecting a @xmath22 atom with its nearest neighbors , and @xmath23 the triad made of their respective opposites ( see figure 1 ) . therefore , the state ( [ 8 ] ) is an eigenvector of @xmath24 provided that the coefficients @xmath18 and @xmath19 are solutions of the eigenvalue problem @xmath25 the parameter @xmath26 measures the hopping between nearest neighbors and provides the natural energy scale in the one - particle description . the diagonalization of ( [ 9 ] ) gives the band of levels @xmath27 the parameter @xmath28 being the lattice spacing . we will be interested in the consideration of this band at half - filling , which is the pertinent instance for the carbon - based materials and , in particular , graphite layers and undoped fullerenes . the dispersion relation ( [ 12 ] ) has the amazing property that the fermi level at @xmath29 is only reached by six isolated points in momentum space ( see figure 2 ) . these are the six corners of the first brillouin zone and , due to the periodicity of the dual lattice , they correspond to only two independent states . the existence of a finite number of fermi points is quite unusual in more than one spatial dimension . it has important consequences since it makes possible to encode the low - lying excitations about the fermi level into a simple field theory . when considering an arbitrarily large sample , it is appropriate to amplify the region about any of the two independent fermi points by taking the limit @xmath30 . in either case we obtain for the operator @xmath31 in equation ( [ 9 ] ) @xmath32 where is the displacement about the fermi point and @xmath33 stand for the two pauli matrices . the conclusion is that , in the long - wavelength limit @xmath30 , the one - particle ( energy ) eigenstates of the electrons in the honeycomb lattice are given by the field theory of two massless dirac spinors ( one for each independent fermi point ) in two spatial dimensions . this is all about the free theory . we now turn to introduce the electronic interaction by applying the method of second quantization . we remark that , proceeding in this way with the above set of one - particle states , we end up with a formalism which falls more into the framework of quantum field theory than into that of quantum statistical theory . the difference is subtle but significant . in quantum statistical theory the ground state of the system is , in general , a condensate with an extended fermi surface ( in more than one spatial dimension ) and the elementary excitations ( quasiparticles ) are only stable right at the fermi level . the second quantized theory built from the dispersion relation ( [ deq ] ) has , however , a definite vacuum in the sense of quantum field theory . its excitations are similar to the particle - hole excitations of the dirac sea . we may consider separately , for instance , each of the two dirac spinors and work out the many - body electron propagator of the free theory @xmath34 we denote by @xmath35 the electron operator @xmath36 in the interaction representation while @xmath37 . in terms of the respective modes @xmath38 for unoccupied and occupied levels we have @xmath39 where @xmath40 is the fermi velocity and the operators @xmath41 satisfy anticommutation relations @xmath42 taking the statistical average with the system at half - filling , we get for the fourier transform of the propagator @xmath43 @xmath44 is a standard set of @xmath45-matrices satisfying @xmath46 and related to the above quoted pauli matrices by @xmath47 . the above expression coincides precisely with the feynman propagator for a massless spinor in 2 + 1 dimensions , which confirms our assertion that the system at half - filling is equivalent to the dirac sea in quantum field theory . we must notice however the presence of the fermi velocity @xmath48 in front of the spatial part of the scalar product , which is a feature of the many - body theory . in the nonrelativistic approximation , the introduction of the electronic interaction leads , in first instance , to the second quantized hamiltonian @xmath49 we attach to this hamiltonian the name of `` coulomb '' since it describes the instantaneous interaction between electric charges . the model given in terms of the coulomb interaction is , however , a highly nonlocal field theory , what makes very awkward its investigation from a formal point of view . both for the sake of studying the properties of the quantum theory as well as for computational purposes , a complete description of the interaction with the electromagnetic field @xmath50 is more desirable . one of the points that we support in what follows is that , actually , the model of interacting electrons is sensitive to retardation effects of the electromagnetic propagation . the reason for such unconventional behaviour is the massless character of the spinor @xmath51 . this property is not accidental in the description of the two - dimensional layer , as long as the expansion ( [ deq ] ) gives only one marginal operator dictating the long - distance behaviour of the free theory . we propose , therefore , a quantum field theory description based on the second quantized hamiltonian . ] @xmath52 the interaction of the electromagnetic field @xmath53 and the electrons in the layer is described in the standard fashion , by coupling to the conserved current @xmath54 this poses some technical problems since the electromagnetic field propagates in three - dimensional space while we want the dynamics of the electrons to be confined to the two - dimensional layer . although this may not be achieved in general , it turns out to be possible by specializing to the feynman gauge@xcite , which enforces the constraint @xmath55 and has also the property of placing the @xmath56 field in the same direction as the electronic current . in this gauge we have the propagator @xmath57 and the coupling to the 2 + 1 dimensional current is perfectly consistent . in the computation of electronic properties we have just to take care of placing the points @xmath58 on the plane of the two - dimensional layer . with this description we certainly get more than we would need in a nonrelativistic theory , but we can recover at any time the nonrelativistic limit of all quantities by expanding in powers of @xmath59 . it is important to stress that we have to expect , in general , different answers in the computation of a given observable taking ( [ hcoul ] ) or ( [ hq ] ) as starting point . the dimensionality of the system is low enough so that the massless condition of the spinor reflects in the appearance of infrared divergences in perturbation theory . thus , although we should not expect any discrepancy in the computation of local quantities ( like cutoff dependent quantities ) in the nonrelativistic model ( [ hcoul ] ) and in the model given by ( [ hq ] ) in the limit @xmath60 , the response of nonlocal quantities to the interaction turns out to be quite different in the infrared regime . from a technical point of view , perturbation theory looses predictive power because of the mentioned infrared instabilities , and one has to resort to more sophisticated methods in order to obtain information from the quantum theory . let us illustrate the above fact with the computation of the quantum corrections to the density of states @xmath61 near the fermi level ( @xmath62 ) . this observable is given by @xmath63 \label{density}\ ] ] in the theory with the instantaneous coulomb interaction , the density of states is not modified to first order in perturbation theory . the instantaneous interaction can be obtained from ( [ em ] ) , for instance , by considering formally its expression for @xmath64 . in this limit we have , computing always at points @xmath65 in the two - dimensional layer @xmath66 then , it becomes obvious that this kind of interaction can not modify the @xmath67 dependence of the propagator @xmath68 , to the first perturbative order ( see also the discussion at the end of section 4 ) . however , as mentioned above , the propagator ( [ propa ] ) can not be safely used in the computation of nonlocal quantities . this remark is appropriate to the case of the density of states , whose determination in the quantum field theory given by ( [ hq ] ) goes as follows . the inverse propagator may be decomposed , as usual , into a free part and the electron self - energy @xmath69 @xmath70 @xmath69 has a perturbative expansion in terms of one - particle - irreducible diagrams@xcite , which are in general ultraviolet divergent . since our model can be treated as a genuine quantum field theory we expect these divergences to be local and susceptible of being absorbed into @xmath71 . for the moment we are only interested in finite corrections and will leave the issue of renormalization to the next section . to first order in perturbation theory , the relevant diagram is shown in figure 3 and gives ( in the limit @xmath60 ) the finite contribution to @xmath72 @xmath73 when inserted in equation ( [ density ] ) the above expression gives the first quantum corrections to the density of levels @xmath74 it is already appreciated that the perturbative expansion of @xmath61 is not well - defined , since the last integral in ( [ 21 ] ) is clearly divergent . even if we could manage to regularize in some way this infrared divergence , we still would be left with a more serious problem . at the point near which we want to measure the density of levels , i.e. near @xmath62 , the first perturbative correction dominates over the zeroth order term , and one can presume that higher order terms can produce increasingly large powers of @xmath75 . perturbation theory is clearly not reliable in the infrared regime , but the kind of infrared instability described points at new physics near the fermi level . henceforth we will pay attention to the local quantum field theory ( [ hq ] ) , which offers the possibility of analyzing the infrared behaviour by application of renormalization group methods . we study first in this section the renormalization properties of the model . though the many - body theory of the 2d layer has a natural ultraviolet cutoff , we insist in absorbing ultraviolet divergent contributions into bare parameters of the theory since this is a way of extracting relevant information about the scaling properties in the infrared@xcite . as we have mentioned before , the renormalizability of a quantum many - body theory is , in general , a nontrivial issue , since the couplings in the effective theory may depend on the point chosen on the fermi surface . in our case things become much simpler as our low - energy effective theory is a genuine quantum field theory . regarding the usual considerations for renormalizability in field theory , the only condition that we are missing is relativistic invariance since the fermi velocity of the electrons @xmath76 does not match the speed of light @xmath77 that appears in the dispersion relation of photons . for this reason it is still necessary to check that the ultraviolet divergences of the theory can be absorbed into a redefinition of the scale of the fields and of the parameters in the hamiltonian . we present here the analysis at the one - loop level , and conclude with a compelling argument showing that the model is renormalizable to all orders in perturbation theory . the conditions which make the theory renormalizable can be stated more clearly in the lagrangian formalism . corresponding to the hamiltonian ( [ hq ] ) , we have the action @xmath78 in the quantum theory the coefficients of the terms in the action have to be adjusted so as to cancel the ultraviolet divergences . we should start therefore with the bare action @xmath79 the renormalization coefficient @xmath80 may be determined from the electron self - energy @xmath69 , as well as the renormalization coefficient for the fermi velocity @xmath81 . the coefficient @xmath82 is determined from the renormalization of the time component of the interaction vertex . it turns out therefore that the spatial components have necessarily to be renormalized by the product @xmath83 . this is a nontrivial check of the renormalizability of the theory , which arises in the absence of relativistic invariance . provided that it is fulfilled , the renormalized theory may be made finite in terms of renormalized quantities @xmath84 @xmath85 in the previous section we already quoted the one - loop renormalized contribution to the electron self - energy @xmath86 . we deal here with the ultraviolet divergent part of this object , regularizing it by working in analytical continuation to dimension @xmath87@xcite . from the diagram in figure 3 we have @xmath88 in order to ensure the finiteness of the electron propagator we take the renormalization coefficients @xmath89 these coefficients have a more complicated structure than those of a relativistic theory , as long as @xmath48 does not match the speed of light which , in our units , means that @xmath90 . @xmath81 has , for instance , an infinite power series expansion in @xmath48 @xmath91 we will later use this expression in the consideration of the nonrelativistic approximation @xmath92 . now we come to the renormalization of the interaction term in the action ( [ action ] ) . the first quantum corrections to the vertex are given by the diagram in figure 4 , and we denote them symbolically by @xmath93 . the divergent contribution to the time component of the vertex is actually different than that to the spatial components . we find that @xmath94 we may think of @xmath93 as a correction to the bare vertex in ( [ action ] ) . we have then @xmath95 by taking the quotient between ( [ dos ] ) and ( [ uno ] ) , we get an independent determination of the renormalization coefficient @xmath81 @xmath96 it is not obvious that this expression coincides with the alternative form ( [ alt ] ) , though it has to for the theory to be renormalizable . in fact , after using the recursion relations which hold for hypergeometric functions@xcite one can show that the two functions in ( [ one ] ) and ( [ alt ] ) are identical . we then have the nice situation of dealing with a nonrelativistic quantum field theory in which the fermi velocity of the electrons is renormalized exactly in the same way in the different terms of the action . this precise coincidence is not casual . one is led to suspect that it has to be consequence of some symmetry operating in the theory , which turns out to be nothing but gauge invariance . actually , this should automatically lead in the quantum theory to the same renormalization of the kinetic part and the interaction term in the action , @xmath97 . it is not difficult to check that this is satisfied in our model , to the perturbative order in which we are working . from expressions ( [ zpsi ] ) and ( [ uno ] ) we have @xmath98 which become identical after use again of the relations among hypergeometric functions . the gauge symmetry @xmath99 is manifest at the classical level . provided that it holds for the bare action ( [ action ] ) , it ensures that @xmath97 . it is easily seen , by taking a gauge parameter @xmath100 independent of @xmath26 , that it also enforces the same renormalization of the fermi velocity @xmath48 in the different terms of the action . as long as one is able to preserve the gauge invariance in the quantum theory , this turns out to guarantee the renormalizability to all orders of the perturbative expansion . the rest of the renormalization coefficients in ( [ ren ] ) can be obtained from the following remark . the fact that @xmath101 has to equal @xmath82 to all orders in perturbation theory implies , as in standard quantum electrodynamics , that @xmath102 . the renormalization coefficient of the electromagnetic field @xmath103 is determined from the polarization tensor , whose first perturbative contribution shown diagrammatically in figure 5 is a finite quantity @xmath104 we conclude , therefore , that @xmath105 remarking the absence of renormalization of the electric charge to first order in perturbation theory . in this section we seek the connection of our model to the real physics of electrons in a two - dimensional layer . a single graphite sheet or any fullerene aggregate of sufficiently large size is the appropriate physical instance to look at . the first question concerns the magnitude of the parameters pertinent to the real layer . a standard value for the hopping parameter in graphite is @xmath106 . for the nearest neighbor separation in the lattice we can take @xmath107 . from these values and restoring for a moment the dependence on @xmath77 and @xmath108 , the fermi velocity in our model turns out to be @xmath109 . this is certainly much smaller than the speed of light , for which to extract relevant quantities to the electron physics in the layer we will have to take the limit @xmath60 and keep the leading order in previous formulas . for a clearer description of our approach to renormalization group we will quote the renormalization coefficients in this section as functions of the ultraviolet momentum cutoff @xmath110 rather than in terms of the dimensional regularization poles @xmath111 . at the one - loop level this just amounts to replace @xmath111 by @xmath112 in all formulas . the renormalization coefficient of the fermi velocity @xmath113 becomes , for instance , in the nonrelativistic limit @xmath114 it is not difficult to see that at each level of the perturbative expansion the leading order in the nonrelativistic approximation enters with a power @xmath115 . we define , therefore , a consistent perturbative expansion in the nonrelativistic limit by taking @xmath92 ( or , equivalently , @xmath64 ) and @xmath116 this latter quantity becomes then the effective strength of the electronic interaction . in this limit we obtain from ( [ zkin ] ) @xmath117 so that we do not find wavefunction renormalization at the one - loop level in the nonrelativistic limit . we recall also the former result ( [ charge ] ) @xmath118 we have a nonrelativistic quantum theory with no renormalization of the electric charge , but in which the effective strength of the interaction @xmath119 is renormalized . we rely on the original interpretation of renormalization adopted by wilson@xcite to determine the scale dependence of the effective interaction . since we have shown that the model is renormalizable , we may impose the independence of the renormalized theory on the momentum cutoff @xmath110 . by means of renormalization group transformations we are able therefore to relate bare theories at different scales ( different cutoff ) which correspond to the same renormalized theory@xcite . the essential information is encoded in the renormalization group equation , which for the electron propagator , for instance , states that under a change in the scale of the cutoff @xmath110 @xmath120 where @xmath121 the well - known solution to this equation takes the form @xmath122 where the effective parameters are given by @xmath123 increasing the scale @xmath124 of the cutoff is equivalent to measure the observables of the theory at large distance scale . from the one - loop order results of the previous section we see that @xmath125 is constant at this level , while @xmath126 is not . we expect , therefore , the effective coupling of the electronic interaction @xmath119 to have a nontrivial renormalization group flow in the infrared regime . actually , from equations ( [ zetav ] ) and ( [ betav ] ) @xmath127 so that the asymptotic behaviour of the coupling is @xmath128 we have to bear in mind that , when integrating equation ( [ flow ] ) , we are relying heavily on perturbation theory . if we recall our original estimate of the fermi velocity @xmath48 , we see that the coupling @xmath119 of the nonrelativistic theory is not a small parameter but , in fact , larger than one . nevertheless , we know that when dealing with a renormalized quantum theory the statement that a coupling like @xmath119 is small or large applies to a particular energy scale . we then have to be very precise in establishing the regime in which the integration of the renormalization group equation becomes consistent . let us remark that the solution ( [ sol ] ) of the renormalization group equation is equivalent to performing the sum of leading logarithm terms of the perturbative expansion and curing , therefore , the problem of the infrared instability of the density of states addressed at the end of section 2 . taking the limit @xmath129 in the bare theory corresponds to measure at large distances ( or small energies with respect to the fermi level ) and in this regime perturbation theory becomes more and more accurate . this is another instance of a physical system which has strong coupling at certain energies and an asymptotic regime at which perturbation theory becomes reliable , similar to what is believed to be true ( as supported by experiment ) in the quantum field theory of the strong interactions . we are providing the example of a nonrelativistic theory in which this also happens , exchanging the ultraviolet regime in the case of the strong interactions by the infrared regime here . as for the practical application of the flow ( [ sol ] ) , we remark that the nonrelativistic theory appears to be formally incomplete , since it does not give any hint about what the value of @xmath130 has to be . this is related to the unlimited growth of the fermi velocity , and to the fact that , in the nonrelativistic approximation , there is no signal of a renormalization group fixed point . later we will see that in the complete quantum field theory the fixed point is given by @xmath131 , but the flow in the graphite layer may not be necessarily controlled by this value of the renormalized coupling . we believe , anyhow , that the observation of the phenomenon of scaling in the asymptotic region should be possible in a single graphite sheet . regarding the possible existence of a natural infrared cutoff , it can be shown that in our model the coulomb interaction is not screened , in the usual metallic manner , by quantum effects . this is due to the vanishing of the density of states at the fermi level . thus , a finite screening length is only generated by giving to the system an excess of charge , as a consequence of doping . in order to study this effect , we may place the system away from half - filling by introducing a chemical potential @xmath132 for the conserved charge @xmath133 the correct expression of the free electron propagator at finite @xmath132 is@xcite @xmath134 obviously , the @xmath135-dependent part can not change the ultraviolet behaviour of the loop - integrals and the renormalization coefficients of the quantum field theory are not modified by the nonvanishing chemical potential . this may lead , however , to finite corrections . the fermion loop diagram is now nonvanishing , opposite to what happened before at @xmath136 . the computation of ( [ ncharge ] ) gives @xmath137 which , after multiplying by 2 ( the two independent fermi points ) gives the correct number of particles . regarding the polarization tensor , a nonvanishing value of @xmath138 signals the appearance of a photon effective mass . at small @xmath132 one finds actually the estimate for the inverse of the screening length @xmath139 which gives a measure of the screening effects away from half - filling . we now turn , finally , to the issue of whether the electron observables may get anomalous dimensions in the nonrelativistic limit . we remark that the absence of one - loop wavefunction renormalization in the limit @xmath92 is just an accident . the signal of wavefunction renormalization is given by an ultraviolet divergent self - energy contribution of the form @xmath140 to establish the degree of divergence of the local contributions one can safely implement the nonrelativistic limit in the photon propagator , which leads to the expression ( [ propa ] ) . the @xmath67-dependence in the vertex diagram in figure 3 can be absorbed then by a simple change of variables in the loop integral , so that the divergent part of the diagram can not bear any dependence on it . at the two - loop level , the same independence of the local contributions on @xmath67 holds except for the vertex diagrams in figure 6 . in these diagrams the loop integrals overlap in such a way that the previous argument does not apply . by using the nonrelativistic approximation ( [ propa ] ) and after a very lengthy calculation , one obtains from the sum of the two - loop diagrams @xmath141 where @xmath142 is a nonvanishing quantity , @xmath143 . we see therefore that the electron observables acquire anomalous dimensions in the infrared regime . according to equation ( [ rge ] ) , the electron propagator transforms under a change of scale @xmath124 by the factor @xmath144 we have therefore an electronic system which deviates slightly from fermi liquid behaviour . in particular , we have seen that the vanishing of the density of states at the fermi level suppresses the usual screening of the coulomb interaction . this property is maintained in the quantum theory , at least at the one - loop level , since the fermi level is not shifted by quantum corrections . also a gap does not open at the fermi points to the one - loop order ( as it would be if an effective mass for the spinor had been generated in perturbation theory ) . the system is not an insulator , though it does not exhibit usual metallic properties either . in the previous section we have mentioned that the nonrelativistic theory is formally incomplete since it does not know of any bound in the effective fermi velocity @xmath126 of electrons . as this parameter grows steadily at large distances the nonrelativistic approximation becomes less accurate , up to a point in which the power series expansion in @xmath59 looses its meaning . in this relativistic regime , we have to use the full expression of the beta function @xmath145 to analyze the behaviour of @xmath126 . in particular , we are interested in knowing if a possible fixed point of the renormalization group flow exists , characterized by the condition @xmath146 by quoting ( [ alt ] ) in terms of elementary functions , this equation reads , at the one - loop level , @xmath147 for positive @xmath48 , the only solution to ( [ fix ] ) is @xmath148 , ) as @xmath149 . ] which in our units means @xmath48 equal to @xmath77 . thus , the full quantum field theory has a built - in upper bound of the renormalization group flow , and the speed of light turns out to be the fixed point for the fermi velocity of electrons . in fact , it is not difficult to check that the one - loop beta function has a simple zero at @xmath150 , and to compute the flow in the neighborhood of the fixed point . the renormalization group equation for @xmath126 becomes @xmath151 the asymptotic behaviour of the solution is , taking for @xmath130 the fixed point value , @xmath152 in figure 7 we show the plot of this scaling behaviour , which is practically indistinguishable in the range considered from that obtained with the nonrelativistic approximation . recalling the vanishing of the electric charge beta function @xmath153 at the one - loop level , we may assert that to this perturbative order we have identified an infrared fixed point given by @xmath131 and arbitrary ( though small ) value of @xmath125 . the one - loop result is not satisfactory , however , since the independence of @xmath125 on the flow may be destabilized by higher - loop effects . in the renormalization group approach to fermi liquid theory@xcite , the consideration of the one - loop graph in figure 5 is enough since it encodes , via the renormalization group , information of the series of ladder diagrams . in the usual analysis of the many - body theory with an extended fermi surface these are which dominate in the infrared regime . because of the existence of fermi points in the honeycomb lattice at half - filling , though , there is no argument here to conclude the relevance of one set of diagrams against the rest . in our quantum field theory description it seems that no diagram can be discarded a priori . it is also for this reason that there are diagrams which are irrelevant in fermi liquid theory but give rise in our model to wavefunction renormalization and anomalous dimensions in the infrared regime . regarding the renormalization of the electric charge , we have analyzed the polarization tensor @xmath154 to second order in perturbation theory . as stated above , the renormalization of the photon field @xmath56 is directly related by gauge invariance to the renormalization of the charge @xmath155 . computing in the framework of dimensional regularization we have not found any divergent contribution of order @xmath156 to @xmath154 . the reason is the same operating in the one - loop polarization tensor . a nave dimensional analysis of the diagram shows that it should diverge linearly in momentum space . in the dimensional regularization scheme this leads to the appearance of gamma functions at half - integer values of the argument , instead of poles at @xmath157 . we believe that this scheme is the best suited for the regularization of the gauge theory it certainly is not going to break the gauge invariance in any event . if , however , one insists on understanding the regularization in terms of a cutoff dependence , it can be checked that this does not lead either to a renormalization of the scale of the field @xmath50 . one would have to look for divergent contributions of @xmath158 type accompanying the gauge invariant expression ( [ pol ] ) for @xmath154 . these are not present at the two - loop order and one may conjecture , following the above dimensional argument , that they do not appear to any order in perturbation theory . at or near the fixed point , the two - dimensional electron liquid we are describing is definitely not a fermi liquid . the main reason for this assertion is the anomalous dimension acquired by the electron field @xmath51 , which may be obtained from ( [ zpsi ] ) evaluated now at @xmath150 . at the fixed point the behaviour of the green function under a change of scale is @xmath159 or , equivalently , after introducing the precise value of @xmath160 @xmath161 we arrive in this way to solve the problem of the proliferation of @xmath162 terms in the perturbative computation of the density of states . as is well - known , the renormalization group approach succeeds in summing up all these increasingly divergent contributions , in the form of an anomalous dependence on the electron frequency . we have found that this sum is consistently achieved in the relativistic regime , where the fixed point lies and the electron field has a simple , anomalous scaling behaviour . we have studied a system of electrons in the two - dimensional honeycomb lattice interacting via a coulomb potential . the fact that the fermi surface of this lattice at half - filling consists of two isolated points renders the problem similar to the @xmath3 case and allows the formulation of the low energy effective theory as a standard quantum field theory . we have been able to investigate the infrared behavior of the system by using renormalization group techniques and have seen that , indeed , it resembles more that of one - dimensional electrons than that of a two - dimensional fermi liquid . before entering into the analysis of the phenomenological consequences of our work , there are a few points that we would like to emphasize . let us first comment on the use made through the paper of the electromagnetic field propagator . this use is unconventional both from the condensed matter point of view where the photon is mostly absent , and from the quantum field theory since , as we will strength now , we are not dealing with standard massless qed . expression ( [ em ] ) shows the usual three - dimensional photon of qed . this is what we use in our relativistic computations . the constraint that , in the electronic interactions , the two points lay on the lattice plane , allows us to integrate out the perpendicular dimension . the resulting expression read as a propagator on the plane certainly differs from the standard propagator in qed@xmath0 . in particular its infrared behavior is quite different , it goes like @xmath163 instead of the standard @xmath164 . this is the origin of , for instance , the finiteness of the photon self - energy up to the two - loop level and , hence , of the nonrenormalization of the electric charge . let us now comment on the comparison between the instantaneous versus the retarded propagator . the effect of retardation on the infrared behavior of nonlocal quantities has been exemplified with the computation of the quantum corrections to the density of states at the end of section 2 . in the theory given by the coulomb interaction ( [ hcoul ] ) , no one - loop corrections were found to the free value of the density of states . in the theory with the retarded interaction our rg analysis shows that , despite the infrared divergence found at the one - loop level , the approach is consistent in the sense that the perturbative expansion of the nonrelativistic approximation does not have to make sense in the infrared when the fermi velocity is still far from the fixed - point value . hence , we must resort to the full relativistic theory when computing the infrared behavior of nonlocal observables . there , the rg allows to interpret the one - loop infrared instability shown in section 2 as the manifestation , in perturbation theory , of the anomalous dimension . we have then found that our original system flows in the infrared towards a fixed point of the renormalization group that is a relativistic quantum field theory with anomalous dimension of the fermionic propagator . our analysis has important consequences in condensed matter physics . one of the more interesting is the anomalous screening of the coulomb interactions of the model . in a fermi liquid , formation of virtual electron - hole pairs ( i. e. , quantum corrections to the photon self - energy ) generates a new scale , the screening length . this is the basis of the absence of long - range interactions in the ultralocal approach to interacting electrons ( hubbard model ) . what we have seen is that , due to the vanishing of the density of states at the fermi level and to the absence of renormalization of the photon self - energy , there is no metallic screening in our materials @xcite . the robustness of this result stands on the non - renormalization of the chemical potential ( the fermi level remains a set of two points after the corrections due to the interactions ) . moreover we have seen that , although the low energy density of states goes to zero , no finite gap appears in the spectrum of charge excitations . our results rule out the validity of descriptions in terms of on - site interactions only . screening effects are not strong enough to reduce the couplings to simple on - site terms . the one - particle properties are very different from those of a conventional fermi system . the wavefunction renormalization leads to the disappearance of the quasiparticle peak at low energies . thus , no effective description in terms of independent , and coherent , quasiparticles is possible . the situation resembles the 1d luttinger liquid , where any amount of electron - electron interaction suffices to wipe out the quasiparticle pole , leading to incoherent one - particle excitations . finally , the existence of nontrivial scaling laws implies that anomalous exponents should show up in a variety of properties , like the specific heat , susceptibility , etc . the possibility of having an experimental test of these results remains open . to conclude this summary and to point out possible future work , we would like to remark the interest of investigating the connections of our work with the infrared catastrophe mentioned in the introduction as well as with other unconventional fermi liquid behavior discussed in the literature @xcite . on this respect , this work can be seen as a rigorous realization of the ideas settled in @xcite . * acknowledgements * this work has been partially supported by the cicyt . 99 a. a. abrikosov , l. p. gorkov and i. e. dzyaloshinski , _ methods of quantum field theory in statistical physics _ ( dover , new york , 1975 ) . e. m. lifshitz and l. p. pitaevskii , _ statistical physics _ , vol . 2 ( pergamon press , oxford , 1980 ) . f. d. m. haldane , _ j. phys _ * c14 * ( 1981 ) 2585 . p. w. anderson , _ phys . lett . _ * 64 * ( 1990 ) 1839 ; * 65 * 2306 ; * 66 * 3226 . for a recent review on the infrared catastrophe see the talk of p. w. anderson in _ the physics and the mathematical physics of the hubbard model " _ , nato a.r.w . , san sebastian , october 1993 . for a theoretical physicist s vision of the problem see j. polchinski , _ effective field theory and the fermi surface " _ , hep - th-9210046 . g. benfatto and g. gallavotti , _ j. stat . * 59 * ( 1990 ) 541 ; _ phys . rev . _ * b42 * ( 1990 ) 9967 . r. shankar , _ physica _ * a177 * ( 1991 ) 530 . for an extense review see r. shankar , _ renormalization group approach to interacting fermions " _ _ , to appear . wilson , _ phys . rev . _ * b4 * ( 1971 ) 3174 , 3184 . j. gonzlez , f. guinea , and m. a. h. vozmediano , _ nucl . _ * b406 * ( 1993 ) 771 . h. w. kroto , j. r. heath , s. c. obrien , r. f. curl and r. e. smalley , _ nature _ * 318 * ( 1985 ) 162 . w. krtschmer , l. d. lamb , k. fostiropoulos and d. r. huffman , _ nature _ * 347 * ( 1990 ) 354 . a. f. hebard , m. j. rosseinsky , r. c. haddon , d. w. murphy , s. h. glarum , t. t. m. palstra , a. p. ramirez , and a. r. kortan , _ nature _ * 350 * ( 1991 ) 600 . m. j. rosseinsky _ _ , _ phys . _ * 66 * ( 1991 ) 2830 . k. holczer , o. klein , s .- huang , r. b. kaner , k .- j . fu , r. l. whetten and f. diederich , _ science _ * 252 * ( 1991 ) 1154 . for a review on the properties of fullerenes , see k. holczer and r. l. whetten , _ carbon _ * 30 * ( 1992 ) 1261 . j. gonzlez , f. guinea and m. a. h. vozmediano , _ electrostatic screening in fullerene molecules " _ , preprint cond - mat/9303007 ( to be published in _ mod . phys s. sorella and e. tosatti , _ europhys . * 19 * ( 1992 ) 699 , g. santoro , m. airoldi , s. sorella and e. tosatti , _ phys . b _ * 47 * ( 1993 ) 16216 . p. ramond , _ field theory . a modern primer_. ( benjamin / cummings , london , 1981 ) . j. zinn - justin , lectures delivered at the cargse summer school 1973 ( unpublished ) . k. symanzik , in _ particles , quantum fields and statistical mechanics _ ( edited by m. alexanian and a. zepeda ) ( springer verlag , berlin , 1975 ) . d. j. amit , _ field theory , renormalization group and critical phenomena _ , chap . 8 ( mcgraw - hill , new york , 1978 ) . i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series and products _ , p. 1044 ( academic press , new york , 1980 ) . k. g. wilson , _ phys . * 28 * ( 1972 ) 548 . s. a. chin , _ ann . * 108 * ( 1977 ) 301 . x. g. wen , _ phys . _ * b42 * ( 1990 ) 6623 . c. m. varma _ et al . _ , _ phys . * 63 * ( 1989 ) 1996 . * figure 1 : * the planar honeycomb lattice .

galaxies experience the most dramatic changes during galaxy mergers . mergers can sometimes change the galaxy morphology ( toomre & toomre 1972 ; barnes & hernquist 1992 ) , generate starbursts ( cox et al . 2008 ; kaviraj 2010a ; lpez - snchez 2010 ) , stimulate nuclear activity ( canalizo et al . 2007 ; bennert et al . 2008 ; ellison et al . 2011 ; cotini et al . 2013 ; satyapal et al . 2014 ) , and contribute to stellar mass growth ( lee & yi 2013 ) . therefore , merger history of galaxies is important to understand present - day galaxies . since mergers cause distortions in the morphology of the system , visual inspections of images are often used to trace recent mergers . deep images are therefore essential for detecting evidence of mergers and interactions . deep images reveal that galaxies have assembled substantial fraction of their mass via mergers even in recent epochs ( van dokkum 2005 ; tal et al . 2009 ; kaviraj et al . 2010b ; sheen et al . 2012 ; yi et al . 2013 ; duc et al . 2015 ) . kinematics plays a key role in determining the morphology of galaxies , and therefore , visually observable disturbances are also linked to the underlying kinematics ( rampazzo et al . 2005 ; kronberger et al . 2007 ; yang et al . 2008 ) . integral field spectroscopy ( ifs ) allows us to access 2d spatially - resolved velocity fields , so the analyses that are used to detect features in an image can be applied to ifs data as well . as we visually define peculiar morphologies using images , kinematically distinct features ( e.g. , kinematically distinct cores , counter - rotating cores , and double peaks in velocity dispersion ) can be detected using 2d kinematic maps of galaxies ( e.g. , krajnovi@xmath2 et al . 2011 ) . several techniques quantifying peculiarities in gas and stellar kinematics have been applied to large datasets from various surveys ( shapiro et al . 2008 ; krajnovi@xmath2 et al . 2011 ) . a recent study by bloom et al . ( 2016 , submitted ) shows that morphologically disturbed galaxies correlate with asymmetry in gas kinematics better than `` stellar '' ( i.e. , photometric ) parameters , such as gini ( abraham et al . 2003 ) , m20 ( lotz et al . 2004 ) , or cas ( abraham et al . 1996 ; conselice 2003 ) . the distribution of gas is more easily disturbed than that of stars , making gas an efficient tracer of kinematic disturbances ( stil & israel 2002 , cannon et al . 2004 , koribalski & lpez - snchez 2009 , lpez - snchez et al . gas however can not be used for gas - deficient early - type galaxies . we therefore use stellar kinematics for our cluster galaxy sample that is predominantly early type and gas deficient . kinematic studies from ifs surveys have changed our view of galaxy morphology . the sauron ( de zeeuw et al . 2002 ) and atlas@xmath3 surveys ( cappellari et al . 2011 ) reported that galaxies can be separated into two categories , fast and slow rotators , according to their spin parameter ( @xmath4 ) . in their classification , the majority of early - type galaxies show fast and regular rotation , which implies that they have measurable levels of ordered motion ( e.g. , emsellem et al . 2007 ; cappellari et al . 2007 ; emsellem et al . 2011 ; krajnovi@xmath2 et al . 2013 ) . this however is counterintuitive to the traditional idea of a `` pressure - supported '' early - type and `` ordered - motion supported '' late - type description of galaxies , thereby questioning the origin and evolution of angular momentum in different types of galaxies . the investigation of mergers on angular momentum changes can provide clues to the origin of slow rotators . numerical simulations suggest that galaxies can lose their angular momentum via mergers and become slow rotators ( bois et al . 2011 ; khochfar et al . 2011 ; naab et al . 2014 , and references therein ) . to empirically assess the role of mergers on the evolution of angular momentum , it is necessary to know the intrinsic distribution of angular momentum of stable disks , before a galaxy undergoes an interaction . the tully - fisher relation ( t - f , tully & fisher 1977 ) serves this purpose . the tight correlation between stellar luminosity and hi line width has been well established by many studies on disk galaxies in the local universe . furthermore , the t - f relation is found even among early - type galaxies using hi and co line observations ( williams , bureau & cappellari 2010 ; davis et al . 2011 ; heijer et al . recently , cortese et al . ( 2014 ) found that early - type galaxies exhibit t - f relation similar to spirals , although with markedly larger scatter . these findings imply that the t - f relation is a general trend in `` regular '' rotating systems regardless of galaxy morphology ; and if dynamical interactions change the angular momentum of galaxies , they introduce scatter in the t - f relation ( kassin et al . 2007 ; covington et al . 2010 ; de rossi et al . 2012 ) . several studies from numerical simulations suggest that the direction of angular momentum can be easily changed via gas accretion from cosmic webs or galaxy mergers ( e.g. kimm et al . 2011 ; sharma , steinmetz , & bland - hawthorn 2012 ; dubois et al . the misalignment of orientation between images and stellar kinematics was first introduced in franx , illingworth & de zeeuw ( 1991 ) , and after that several ifs surveys also detected the kinematic misalignment . krajnovi@xmath2 ( 2011 ) reported that 10% of early - type galaxies , some of which show prominent merger features , show misaligned kinematic and morphological orientations . using data from the califa survey ( snchez et al . 2012 ) , barrera - ballesteros et al . ( 2015 ) compared kinematic misalignment of interacting and undisturbed galaxies , and found large misalignments among galaxies that seem to be on - going mergers . on the other hand , the misaligned galaxies in their sample do not always show merger signatures , and misalignment was more often found among galaxies that were rounder or had low angular momentum , making it difficult to conclude on the effect of mergers on kinematic anomalies . it is therefore necessary to use deeper imaging data that reveal the merger status of galaxies clearly to answer this question . here we investigate the impact of mergers on three aspects of kinematics : angular momentum , orientation , and asymmetry . in order to do this we used deep optical images obtained with the blanco 4-m telescope at the cerro tololo inter - american observatory ( ctio , sheen et al . 2012 ) , and stellar kinematics from the sydney - aao multi - object integral field ( sami , croom et al . 2012 ) galaxy survey ( bryant et al . 2015 ) . specifically , we quantitatively compare kinematic properties between morphologically - disturbed and undisturbed galaxies in abell 119 and discuss the merger origin of their kinematic anomalies . throughout the paper , we adopt a standard @xmath5cdm cosmology with @xmath6 , @xmath7 , and @xmath8 km / s / mpc . we obtained deep optical images of abell 119 in @xmath9 , @xmath10 , and @xmath11 bands using the mosaic ii ccd mounted on the blanco telescope . these observations were carefully planned with the goal of detecting low surface brightness features related to galaxy interactions . for abell 119 , seven dithered exposures were taken and combined to make an @xmath11-band stacked image with a total integration time of 5,040 seconds . the large field of view ( fov ) of mosaic ii ccd ( @xmath12 ) covers @xmath13 of abell 119 ( @xmath141.53 mpc ) at @xmath15 0.044 . the surface brightness profiles of galaxies go down to 30 magnitude per square arcsecond in the @xmath11 band , which allows us to detect extremely faint features ( @xmath16 , 1@xmath17 above the sky level ) . magnitudes in this study are corrected for foreground extinction using the reddening maps from schlegel et al . details of data reduction and treatment of photometric data are described in sheen et al . ( 2012 ) . the sami instrument ( croom et al . 2012 ) is mounted on the 3.9-m anglo - australian telescope ( aat ) and fed into the aaomega spectrograph ( sharp et al . it has 13 ` hexabundles ' of 15@xmath18 diameter ( bland - hawthorn et al . 2011 ; bryant et al . 2014 ) ; each is comprised of 61 1.6@xmath18-diameter optical fibers . the hexabundles patrol an fov of 1-degree diameter . the beam is split into blue ( 580 v grating , 3700 - 5700 @xmath19 ) and red ( 1000r grating , 6300 - 7400 @xmath19 ) arms ; the spectral resolution of the blue and red gratings are r = 1730 and r = 4500 , respectively . the main sami survey will target 3,700 nearby galaxies at 0.04 @xmath20 0.095 from the galaxy and mass assembly ( gama ; driver et al . 2011 ) to obtain an unbiased sample of stellar mass ( bryant et al . 2015 ) when completed . in addition , sami contains eight rich clusters at @xmath21 0.1 ( edcc0442 , abell 85 , abell 119 , abell 168 , abell 2399 , abell 3880 , apmcc0917 , and abell 4038 ) to cover the high - density environments missed by the main survey . the raw data were reduced using the @xmath22 package ( croom et al . 2004 ) which carries out bias subtraction , flat fielding , wavelength calibration , and sky subtraction . the pairs of data cubes ( blue and red ) for each galaxy were built after applying telluric correction and flux calibration . see sharp et al . ( 2015 ) and allen et al . ( 2015 ) for a full description of the data reduction . the stellar line - of - sight velocity distribution and velocity dispersion for all galaxies were derived using the ppxf code ( cappellari & emsellem 2004 ) following van de sande et al . ( in prep . ) , and two - dimensional stellar kinematic maps consisting of @xmath23 resampled spatial pixels ( spaxels ) were built from the cubes . in this study , we used stellar velocity fields without binning with a signal - to - noise ( s / n ) cutoff of 5 for the continuum in each spaxel . cluster membership of abell 119 was determined by the redshift information from the sloan digital sky survey ( sdss ; york et al . 2000 ; abazajian et al . 2009 ) , 2-degree field galaxy redshift survey ( 2dfgrs ; colless et al . 2001 ) , and hydra on ctio ( sheen et al . galaxies with @xmath24 and @xmath25 were selected as primary targets for sami observations for this cluster . the sami cluster observations are described in detail in owers et al . ( in prep . ) . of the 254 primary targets in abell 119 , 108 galaxies were observed by sami by the end of 2015 . our final sample contains 63 galaxies brighter than @xmath26 ( @xmath27 ) after cross - matching the 108 galaxies observed by sami with the deep images . our sample contains both early- and late - type galaxies because the parent targets for sami were selected regardless of galaxy morphology ( or color ) . figure 1 shows six sample galaxies . the gray scale background with contours shows the deep image , the colored isophotes are for the specific values of surface brightness as given in the caption , and the colored map at the center shows the sami stellar velocity map . we visually selected galaxies that showed evidence for interactions or mergers in the @xmath11-band deep images . of the 63 galaxies in our sample , 20 galaxies displayed disturbed morphology of varying degree ( figure [ img ] ) . we considered galaxies as morphologically - disturbed when they displayed disturbed features including tidal structure , asymmetric light profile , or unusual dust lanes . if galaxies had close - projected companions , we did not classify them as morphologically - disturbed galaxies unless they showed any of the above - mentioned distortion features . this cluster has been studied for the visual selection of post - merger features in red sequence galaxies by sheen et al . ( 2012 ) . in their study using the same images ( but for a larger number ( 133 ) of member galaxies ) , they found that approximately 30@xmath28 of red sequence galaxies show post - merger features , which is comparable to this study ( 32@xmath28 , 20/63 ) . abell 119 has been imaged by sdss ; therefore , we can directly compare our classifications of the same galaxies at different image qualities . for comparison , we could find only 5 disturbed galaxies using the same visual classification on sdss images . we present the images of all 63 galaxies in the appendix , but it is necessary to inspect the images with varying scale to properly determine the level of disturbance . the identification of morphologically - disturbed galaxies is admittedly subjective , and we further discuss this in section 5.2 . we classify the 63 galaxies in our sample into 53 early- ( e / s0 ) and 10 late - type galaxies . four of the 20 morphologically - disturbed galaxies are late type . considering the high early - type fraction and the dominance of the brightest cluster galaxy ( bcg ) , this puts a119 into a cd cluster category ( way , quintana , & infante 1997 ) . its dominance with early - type galaxies within @xmath13 is also consistent with the typical morphology - density relation found in clusters ( dressler et al . abell 119 is found to be a dynamically young cluster having unrelaxed groups within the virial radius ( lee et al . its dynamical mass has been estimated to be @xmath29 ( sheen et al . 2012 ) . we measured the effective radius ( @xmath30 ) , position angle ( @xmath31 ) , and ellipticity ( @xmath32 ) using the @xmath33 task in iraf stsdas . the center of the galaxies was fixed ; however , we allowed free estimation of the position angle and ellipticity at a given level of light . all the contaminants were masked out to consider only the light from the target galaxy . we used the curve of growth of light from the radial intensity profile to measure the effective ( half - light ) radius @xmath30 . we measured the spatially - resolved stellar velocity observed with sami using the @xmath34 procedure ( krajnovi@xmath2 et al . the @xmath34 procedure calculates the expansion of the surface photometry on velocity fields with high - order moments : @xmath35 where @xmath36 is the semi - major radius of each kinematic ellipse and @xmath37 is the azimuthal angle of the projected major axis . the semi - major axis of each kinematic ellipse was increased by following @xmath38 , where @xmath39 = 1 , 2 , 3 to avoid possible correlations between ellipse elements . the number of ellipse elements was varied between 20 and 100 ( but typically over 60 ) . we measured the ellipse parameters as long as spaxels cover 75% area of the ellipse , and the largest ellipse typically reaches @xmath40 for our sample . the orientation and flattening of the iso - velocity contours are measured along the kinemetric ellipse . the amplitude coefficients for each kinemetric ellipse ( @xmath41 ) can be calculated using the harmonic coefficients ( @xmath42 and @xmath43 ) : @xmath44 we used 6 terms for the harmonic analysis to obtain 3 odd terms ( n=1 , 3 , 5 ) . the coefficient @xmath45 describes the amplitude of the circular velocity , and the higher order terms ( @xmath46 and @xmath47 ) represent complex structure of the velocity field , and deviations from the bulk motion . these coefficients are used to calculate the model rotational velocity in section 4.1 and the kinematic asymmetry in section 4.3 . for a detailed description of the significance of the harmonic coefficients for a thin disk case the reader is referred to wong et al . ( 2004 ) and references therein . profile distribution . the black and red lines indicate the distribution of undisturbed and morphologically - disturbed galaxies . a gradual distribution of @xmath4 from slow to fast is detected , rather than two distinct classes . , scaledwidth=50.0% ] we measured the spin parameter @xmath4 integrating within elliptical apertures with fixed @xmath31 and @xmath48 at the effective radius ( see section 3.2 ) following emsellem et al . ( 2007 ) : @xmath49 where @xmath50 , @xmath51 , @xmath52 , and @xmath53 are the flux , radius , velocity , and velocity dispersion of the @xmath39th spaxel , respectively . for the galaxies whose isophote at the effective radius is highly affected by inner features , such as bars , we measured @xmath4 using the values of @xmath48 and @xmath31 at @xmath54 which corresponds to a semi - major axis @xmath55 . radial profiles of @xmath4 are shown in figure [ lambda ] . our cluster galaxies do show a wide range of spin parameter reaching high values , as has been reported by previous studies . the red curves show the morphologically - disturbed galaxies ; and from a cursory inspection , there does not appear to be a clear distinction between disturbed and undisturbed galaxies . we present the result first in the @xmath56 plane in figure [ fig_lambda2 ] . the demarcation line of @xmath57 where @xmath58 is the coefficient for fiducial radius , has been proposed to select slow and fast rotators by previous studies . we adopt @xmath58 = 0.265 , 0.310 , and 0.363 for r = @xmath59 , @xmath30 , and @xmath60 from emsellem et al . ( 2011 ) and fogarty et al . ( 2014 ) , respectively . considering the radial dependence of kinematics , it is desirable to measure kinematic properties at a consistent radius in all galaxies ( arnold et al . 2014 ; foster et al . we used the diagnostic for the effective radius for all the galaxies in our sample except six . for three galaxies including the bcg , we used the diagnostic for @xmath59 because the sami map did not reach out to @xmath30 . this is not a problem because the two non - bcg galaxies are already classified as fast rotators at @xmath59 , and so if their @xmath4 rises with radius as in most cases , they would still be classified as fast rotators at @xmath30 . the bcg on the other hand is classified as a slow rotator at @xmath59 , and given its radial profile would most likely be a slow rotator at larger radii as well . for the other three galaxies whose @xmath61 arcsecs , we used the diagnostic for @xmath60 instead in order to avoid the beam smearing effect . the seeing effect on @xmath62 measurement is discussed in van de sande et al . ( in prep . ) . out of the 63 galaxies in our sample , 8 galaxies were classified as slow rotators ( 13% ) . this fraction increases to 15% if the sample is limited to early - type galaxies ( 53 galaxies ) and is consistent with previous studies that found slow rotators in cluster environments using the same classification as this study ( cappellari et al . 2011 ; d@xmath63eugenio et al . 2013 ; houghton et al . 2013 ; fogarty et al . 2014 ; scott et al . 2014 ) . of the eight slow rotators three galaxies exhibited disturbed features , which is comparable to previous reports ( duc et al . 2011 ; jimmy et al . we found a slightly higher fraction of slow rotators in morphologically - disturbed galaxies ( @xmath64 ) than in their undisturbed counterparts ( @xmath65 ) . the effective rotational velocity ( @xmath66 ) is defined as the modeled amplitude of the rotational velocity ( @xmath45 ) at @xmath30 as measured from @xmath34 . sami could not recover full rotation curves for most of our sample because of the limited fov ( 15@xmath18 ) and low @xmath67 in outer spaxels . therefore , we could not estimate the maximum flattened velocity for the whole sample . instead , we measured the rotational velocity at @xmath30 . the @xmath66 for the three galaxies whose sami measurements did not reach out to @xmath30 was estimated by the linear extrapolation using the closest 3 points of @xmath45 . we confirm that there is a linear correlation between @xmath66 and the maximum rotational velocity based on 20 galaxies whose velocity profile reaches out to a flattened velocity curve . the @xmath66 was corrected for inclination using the apparent major ( @xmath36 ) and minor ( @xmath68 ) axes at @xmath69 = 25 @xmath70 using the following formula : @xmath71 where intrinsic shape parameter , @xmath72 , is set to be 0.65 for early - type and 0.2 for late - type galaxies , following catinella et al . although the inclination correction is highly uncertain especially for early - type galaxies , it has a negligible effect on our results and conclusions . we present the `` effective '' t - f relation using @xmath66 in figure [ fig_tully ] . our strategic use of velocities from kinemetric modeling significantly reduces the velocity scatter and restores the t - f relation , even for early - type galaxies which have generally been known to lie away from the sequence . as a result , most of the galaxies in our sample are well aligned with the relation . the few outliers seen in figure [ fig_tully ] are very slow for their stellar mass and most of them ( 7 out of 8) are classified as slow rotators in the @xmath73 plane ( figure [ fig_lambda2 ] ) . one slow rotator ( cross inside diamond at @xmath74 and @xmath75 ) lies closely to the t - f relation , possibly due to an inaccurate inclination correction . the undisturbed galaxy that lies beyond the 3@xmath17 ( dashed ) line ( cross at @xmath76 and @xmath77 ) on the other hand , likely suffered from the beam smearing effect and hence the inaccurate estimate of @xmath66 . we estimated the scatter by first finding the robust least - squares fit ( the solid line ) to the unperturbed galaxies and then measuring the standard deviation of residuals in velocity from the fit excluding slow rotators . our t - f relation based on the _ derived _ values of @xmath66 shows scatter that is comparable to that of late - type galaxies ( see kannappan et al . 2002 ) . we looked for a relation between the stellar velocity dispersion at @xmath30 ( @xmath78 ) and the scatter of the t - f relation and found that only slow rotators , which have high velocity dispersion , also have large scatter ( over @xmath79 from the t - f relation ) . fast rotators , regardless of if they are morphologically disturbed or not , typically have a low scatter ( @xmath80 ) from the t - f relation , but show a range of @xmath78 . the possible physical reasons for early - type galaxies lying on the t - f relation will be investigated in a future paper . morphologically - disturbed galaxies show a larger scatter ( 0.11 dex ) in the t - f relation relative to the undisturbed galaxies ( 0.06 dex ) . kassin et al . ( 2007 ) reported that perturbed @xmath81 galaxies deviate from the t - f relation by always showing slower rotation , whereas we found that morphologically - disturbed galaxies scatter in both directions in velocity . flores et al . ( 2006 ) and puech et al . ( 2010 ) reported large scatter on the t - f relation in galaxies with complex kinematics or perturbed rotation , which probably relates to galaxy mergers . galaxy mergers generally enhance luminosity but may either reduce or enhance their rotation speed depending on the merger / interaction geometry . plane for the diagnostic of slow rotators . the black and red symbols indicate morphologically - disturbed and undisturbed galaxies , respectively . slow rotators ( open diamonds ) are defined as galaxies under the demarcation line at @xmath30 from emsellem et al . we used @xmath4 at @xmath59 ( stars ) and @xmath60 ( triangles ) for some galaxies . error bars on @xmath4 were calculated from the analytic method described in houghton et al . approximately , 37% ( 3/8 ) of slow rotators show disturbed features . late - type galaxies ( open squares ) tend to have high @xmath62 . , scaledwidth=50.0% ] as an independent variable ( black symbols ) . the dotted and dashed lines correspond to 2@xmath17 and 3@xmath17 of the relation for undisturbed galaxies , respectively . slow rotators defined using the @xmath56 plane are denoted by open diamonds ( see figure [ fig_lambda2 ] ) , and they deviate significantly from the t - f relation of undisturbed galaxies . scatter from the least - squares fit was calculated on the morphologically - disturbed ( 0.11 ) and undisturbed galaxies ( 0.06).,scaledwidth=50.0% ] . morphologically - disturbed and undisturbed galaxies are denoted by the red and black symbols , respectively . the error bars indicate a @xmath79 confidence level on @xmath82 . the dashed line corresponds to @xmath83 dividing angle of kinematic misalignment . fractions of galaxies showing the kinematic misalignment ( @xmath84 ) in disturbed ( red ) and undisturbed ( black ) galaxies are shown in the top - right corner . morphologically - disturbed galaxies show 4 times higher fraction of kinematic misalignment , and most of misaligned galaxies displayed merger signatures.,scaledwidth=50.0% ] some galaxies with disturbed features show misalignment in orientation between the photometry and kinematics . the global kinematic orientation ( @xmath82 ) , indicating the mean motion of the stellar velocity , was measured using the method described in krajnovi et al . specifically , we made 361 simulated maps by rotating the original map in steps of 0.5@xmath85 and took the angle that minimized @xmath86 . the uncertainty in @xmath82 is defined as the minimum opening angle of simulated maps that satisfied a @xmath79 confidence level ; @xmath87 ( van den bosch & van de ven 2009 ; krajnovi et al . the mean uncertainty value on @xmath82 is 6@xmath85 for the sample galaxies ; so , we used @xmath83 as a cut for `` misaligned '' galaxies . we measured a luminosity - weighted average of position angle ( @xmath88 ) and ellipticity ( @xmath89 ) at @xmath90 as representative parameters in photometry . there are several uncertainties on the measurement of @xmath31 in both photometry and kinematics . first , @xmath31 can not be determined in round galaxies . the central regions of galaxies ( @xmath91 ) are rather circular , and therefore @xmath31 measurements at @xmath30 from optical images have large uncertainties . we therefore measured @xmath31 in outer regions ( @xmath90 ) as routinely exercised ( e.g. cappellari et al . 2007 , krajnovi@xmath2 et al . 2011 ) . second , photometric @xmath31 at specific radius can follow the dominant light from spiral arms or disturbed features . a luminosity - weighted average of @xmath31 in this study is free from this issue . we manually confirmed that @xmath88 does not follow local features . third , we can not determine the kinematic orientation for galaxies with no rotation . however , even slow rotators in this study have a certain level of rotation which is enough to judge their kinematic orientation , except two galaxies whose @xmath66 is below 10 km / s . we confirmed that our results are not affected by excluding them from this analysis . galaxy interactions seem to have played a key role in creating the misalignment between @xmath82 and @xmath88 ( figure [ misalign1 ] ) . galaxies with disturbed features show a 4 times larger fraction ( @xmath92 ) of misaligned galaxies ( @xmath93 ) compared to their undisturbed counterparts ( @xmath94 ) . we could not find significant differences in the misalignment even if we measure @xmath88 at @xmath95 or @xmath96 . krajnovi@xmath2 et al . ( 2011 ) investigated the kinematic misalignment with regards to galaxy morphology and found a higher fraction of kinematic misalignment greater than @xmath83 in their `` interaction '' or `` shell '' class ( 24% ) than in their normal morphological class ( 9% ) . barrera - ballesteros et al . ( 2015 ) also reported that 43% of the `` interacting '' sample showed misalignment larger than the mean value of misalignment angle of the control sample . based on the measurement at @xmath97 , we found that 67% ( 6/9 ) of misaligned galaxies showed disturbed features . in krajnovi@xmath2 et al . ( 2011 ) , however , only 18% of misaligned galaxies were classified as `` interaction '' or `` shell '' class . our images are three magnitudes deeper than those used by krajnovi@xmath2 et al . ( 2011 ) ( from the sdss and isaac newton telescope ) , and thus reveal the link between kinematic misalignment and galaxy interaction more clearly . misalignment has been reported to be more conspicuous in slow rotators ( e.g. , cappellari et al . 2007 , krajnovi@xmath2 et al . 2011 ) . in figure [ fig_misalign2 ] we also observe a similar trend . we found that only four out of the 55 fast rotators are misaligned , compared to five out of the eight slow rotators . in addition , we found that kinematic misalignment is only seen in galaxies with low angular momentum ( e.g. @xmath98 ) . if we only consider galaxies with @xmath98 , a contrast between the morphologically - disturbed and undisturbed galaxies becomes more significant : undisturbed galaxies still have a small fraction ( 0.11 ) of kinematic misalignment ; whereas 60% of the morphologically - disturbed galaxies show misalignment . therefore , we can conclude that mergers seem to affect the kinematic orientation of the galaxies primarily with low angular momentum . . the dashed line indicates @xmath99 , the dividing angle of the kinematic misalignments in this study . the error bars indicate a @xmath79 confidence level on @xmath82 . kinematic misalignment is only seen in galaxies having low angular momentum ( e.g. @xmath100 ) ; over half of the morphologically - disturbed galaxies ( 0.60 ) with low angular momentum are misaligned . undisturbed galaxies show small misalignment fraction ( 0.11 ) even when the spin parameter proxy @xmath62 is small.,scaledwidth=50.0% ] has a tight correlation with rotational velocity @xmath66 . inclination correction was not applied in this figure . the solid- and dashed - black lines show the least - squares fit and 2@xmath17 distribution of undisturbed galaxies ( @xmath101 dex in @xmath102 ) , respectively . the solid - red line indicates the least - squares fit of morphologically - disturbed galaxies . we present the fits of disturbed ( red ) and undisturbed ( black ) galaxies in the bottom left , and the typical errors are shown in the top left . in general , kinematic asymmetry in morphologically - disturbed galaxies is boosted compared to that in undisturbed galaxies of the same @xmath66.,scaledwidth=50.0% ] we measured the kinematic asymmetry in the stellar velocity field ( @xmath103 ) using the first ( @xmath45 ) coefficient and the summation of the odd high order ( @xmath104 ) coefficients from @xmath34 following bloom et al . ( 2016 , submitted ) : @xmath105 this @xmath103 , using only the odd terms , is a modification of kinematic asymmetry used by shapiro et al . ( 2008 ) , who defined kinematic asymmetry using both odd and even terms ( @xmath106 , and @xmath47 ) . we took the mean value of @xmath103 at each spaxel without giving luminosity weights because our study is focused on faint features . previous studies that parameterized kinematic asymmetry used specific values of kinematic asymmetry , e.g. , @xmath107 or @xmath108 , to identify `` non - regular '' kinematics ( e.g. krajnovi@xmath2 et al . 2011 , bloom et al . 2016 , submitted ) ; however , a further consideration of rotational velocity might be necessary . the @xmath102 has a tight correlation with @xmath66 , @xmath45 measured at the effective radius , due to its definition including @xmath45 ( equation ( 6 ) ) . therefore , galaxies with small @xmath66 show large @xmath102 , spontaneously ( figure 8) . morphologically - disturbed galaxies show a large @xmath102 compared with undisturbed galaxies at a given @xmath66 . the least - squares fit to the morphologically - disturbed sample ( the solid - red line ) is roughly parallel to and 2@xmath17 away from the sequence for the undisturbed sample ( the solid - black line ) , which indicates that galaxy interactions generally increase @xmath102 . krajnovi@xmath2 et al . ( 2011 ) reported that only a small fraction of galaxies show signatures of past interactions _ regardless _ of the presence of kinematic anomaly based on the simple cut of @xmath107 = 0.04 . following a similar horizontal cut in figure 8 , we would find the same result . however , our sample spans a larger baseline in @xmath109 and clearly calls for necessity for considering a @xmath109-dependent cut rather than a uniform cut to conclude on the impact of dynamical interactions on kinematic anomaly . we discuss this in the next section . we now discuss how kinematic anomalies relate to each other . first , we define two parameters to quantify scatter in the t - f relation ( @xmath110 ) and the perturbations in @xmath102 ( @xmath111 ) : @xmath112 where log@xmath113 is the fitted rotational velocity for a given luminosity ( the solid line in figure [ fig_tully ] ) , and @xmath114 is the fitted kinematic asymmetry for a given @xmath115 ( the solid - black line in figure [ fig_vasym ] ) . therefore , @xmath110 defines scatter in the rotational velocity for a given luminosity . we use the absolute value because mergers appear to cause scatter in both directions . @xmath111 is the excess of @xmath116 compared with the fitted value of undisturbed galaxies for a given rotation . we present a comparison between three kinematic parameters in figure [ fig_resi ] . for easy comparison , we divided galaxies into three groups according to @xmath111 : low levels of perturbation ( lp ; @xmath117 0 ; the least - squares fit of undisturbed galaxies ) , intermediate perturbation ( ip ; @xmath118 ) , and highly perturbed kinematics ( hp ; @xmath119 ; 2@xmath17 of the undisturbed galaxies ) . except for slow rotators , galaxies with highly perturbed kinematics ( hp ) show a higher mean value of @xmath110 ( @xmath120 ) than galaxies with intermediate or low levels of perturbation ( ip or lp ) whose mean @xmath110 is @xmath121 , although scatter is too large to make this trend statistically significant . the trend , if real , is consistent with previous studies based on visual classification on kinematic morphology ( flores et al . 2006 ; puech et al . 2010 ) . the bottom panel of figure 9 shows a comparison between the kinematic misalignment ( @xmath122 ) and perturbation in @xmath102 ( @xmath111 ) . kinematic misalignment ( @xmath123 ) is only shown in ip and hp group , and we found a higher fraction of misalignment in hp ( 0.36 ) than in ip ( 0.16 ) . the mild yet positive correlations between the three parameters quantifying kinematic anomalies suggest that galaxy interactions and mergers have a key role in causing such asymmetries . ) , kinematic asymmetry ( @xmath111 ) , and angular momentum ( @xmath110 ) . morphologically - disturbed and undisturbed galaxies are denoted by the red and black symbols , respectively . late - type galaxies and slow rotators are indicated by open rectangles and diamonds , respectively . in the top panel , most of the slow rotators are outside the scope of the y - axis . galaxies are divided into 3 groups ( lp , ip , and hp ) according to their @xmath111 . the mean value of @xmath110 ( top panel ) or fractions of misalignment ( bottom panel ; @xmath123 ) for each group is provided in the parentheses . the hp group shows a larger mean value of @xmath110 and a higher fraction of misalignment than the others.,scaledwidth=50.0% ] so far we have relied primarily on visual selection of mergers . this is because we believe that human eyes detect morphological anomalies effectively and simply enough . as a sanity check , we now measure the degree of disturbed features using the cas asymmetry parameter ( @xmath124 ) on photometric data ( abraham et al . 1996 ; conselice 2003 ) . following the classic approach , we subtracted the 180@xmath85 rotated image from the original one and normalized it using the total flux . then , the asymmetry parameter @xmath124 is measured by the summation of residuals excluding contaminated regions by other objects . the central parts of bright galaxies are saturated in our deep images because we were attempting to detect low surface brightness features ; hence , we only considered residuals in the outer regions : beyond @xmath30 to @xmath79 above the sky level . for this reason , there is a need for caution when making direct comparisons against previous studies employing asymmetry parameters . in this study , even the morphologically - disturbed galaxies have @xmath125 ; i.e. , we are indeed dealing with low surface brightness features that are only detectable in deep images ( figure [ fig_asym ] ) . our visually selected disturbed galaxies were well distinguished by @xmath124 , except the late - type galaxies , which naturally have a high value of @xmath124 due to spiral arms . over 80% of the morphologically - disturbed galaxies have @xmath126 as shown by the vertical line in figure [ fig_asym ] . in fact , apart from late - type galaxies ( green squares ) , all but one galaxies in the high value of @xmath124 ( @xmath127 ) region are morphologically disturbed . conselice , rajgor , & myers ( 2008 ) pointed out that it is difficult to identify visually - selected disturbed galaxies using asymmetry parameters if @xmath128 . however , we found a good separation between disturbed and undisturbed galaxies based on low surface brightness features measured in @xmath124 . the ability of @xmath124 to detect peculiarities seems to depend greatly on the depth of the image . in the top panel , the kinematic and photometric asymmetries seem to be correlated with each other , even for slow rotators and late - type galaxies . however , we could not find a good correlation between photometric asymmetry and the scatter in the t - f relation , @xmath110 ( middle panel ) . we found a higher mean value of @xmath110 in galaxies with @xmath126 ( @xmath129 ) , than in galaxies with @xmath130 ( @xmath131 ) . kinematic misalignments ( @xmath123 ) appear more in galaxies with @xmath132 : 19% ( 4/21 ) of galaxies with @xmath126 show misalignment , and 12% ( 5/42 ) of galaxies with @xmath130 are misaligned ( bottom panel ) . the results based on asymmetry index measurements are largely consistent with the result based on the visual inspection presented in the earlier sections . . morphologically - disturbed and undisturbed galaxies are denoted by the red and black symbols , respectively . slow rotators are denoted by open diamonds , and late - type galaxies are indicated by open rectangles . the photometric asymmetry @xmath124 and the perturbation on stellar kinematics ( @xmath111 ) are correlated to each other ( top panel ) . the mean value of @xmath110 ( middle panel ) or fractions of misalignment ( bottom panel ; @xmath123 ) for each group is provided in the parentheses . galaxies with @xmath126 shows a larger mean value of @xmath110 and a higher fraction of misalignment than galaxies with @xmath130 . , scaledwidth=50.0% ] repeated mergers and interactions between galaxies probably reduce the specific angular moment of a galaxy , and hence it is reasonable to expect more frequent merger - related features in slow rotators . our deep images are capable of detecting even ( reasonably important ) minor mergers and faint features in outer regions of galaxies where dynamical time scale is much longer than in galaxy center and therefore we can probe the signatures of past mergers further back . we focus on the fraction of disturbed galaxies among slow rotators . three out of 8 slow rotators appear disturbed , which is only marginally significant compared to the fast rotators ( 17 out of 55 ) . but , the statistical differences do not appear clear enough . recently , duc et al . ( 2015 ) investigated galaxy morphology using similar deep images with ours and kinematics of atlas@xmath3 sample in the field environment . of the 78 galaxies whose morphological class is not `` undetermined '' , 46 galaxies ( @xmath133 ) displayed features involved in galaxy interactions ( `` major merger '' or `` minor merger '' or `` interacting '' in their classification ) . fractions of galaxy interactions in fast and slow rotators were @xmath134 ( 38/66 ) and @xmath135 ( 8/12 ) , respectively in their study . the major difference between their and our results is that they found around 1.8 times higher fractions of galaxy interactions than us . this can be explained by the difference in a sample ; their sample includes only galaxies in the field environment . previous studies using deep images reported that galaxies with post - merger features are more frequently observed in the field ( e.g. van dokkum 2005 ) than in the cluster environment ( e.g. sheen et al . besides , the median distance of their sample is 28 mpc , which is 7 times closer than our abell 119 sample . this makes it easier to detect features of morphological disturbance . apart from that , slow rotators show only slightly higher fractions of galaxy interactions than fast rotators both in duc et al . ( 2015 ) and in this work . this is perhaps because mergers may have occurred in the distant past on those slow rotators that do not show disturbed features . this is consistent with recent galaxy formation models where important mergers are rare after @xmath136 ( khim et al . 2015 ) . as shown in section 4.1 , a single merger ( or interaction ) can either reduce or enhance the spin depending on the merger condition ; that is , both effects are possible , as also demonstrated by simulations ( bois et al . 2011 ; naab et al . this makes it difficult to uniquely determine the effect of mergers on the angular momentum of a galaxy at a specific epoch . we investigated the impact of galaxy interactions on the kinematic anomalies using 63 galaxies in abell 119 based on stellar kinematics from sami . we visually selected 20 galaxies with disturbed features ( @xmath137 ) using our deep images . morphologically - disturbed galaxies in this study have faint features that generally show a slight asymmetry ( @xmath138 ) , which can only be detected in deep images . the use of `` model '' values of rotational velocities reveals a tully - fisher relation with unexpectedly small scatter even for early - type galaxies . the morphologically - disturbed galaxies , on the other hand , showed a scatter twice as large . galaxy interactions are likely the main origin of scatter in the t - f relation , capable of both reducing and enhancing the level of angular momentum , probably depending on the details of galaxy interaction . we classified slow / fast rotators using the spin parameter @xmath4 and found 8 slow rotators in our sample . approximately , @xmath139 of the slow rotators and @xmath140 of the fast rotators displayed disturbed features . the difference hints for a role of mergers on galaxy spin but is not statistically significant . galaxy interactions are one of the key mechanisms that change the orientation of spin of galaxies . likewise , kinematic misalignments are more often found among the morphologically - disturbed galaxies in our sample . kinematic misalignments are also more common among the galaxies with low angular momentum ; this is probably because the same level of interaction can have a larger impact on a galaxy with lower angular momentum . when we measure the level of kinematic perturbation , it seems necessary to account for the underlying relationship between the kinematic asymmetry and the level of rotation . we thus introduced a new parameter ( @xmath111 ) which effectively reveals the impact of galaxy interactions on kinematics and photometric morphology . a causality connection is mildly visible between morphological disturbance and kinematics , but not all the disturbed galaxies have unusual kinematics . several effects may contribute to this . first , we compared the stellar kinematics within @xmath40 with the disturbed features in the outer regions . if we can detect stellar kinematics in the same region of galaxies where the disturbed features are shown , we might find a tighter correlation between photometry and kinematics . we also need to consider the lifetime of features in kinematics and images . some aspects might be more quickly regularized than others through galaxy evolution ( hoffman et al . 2010 ; ji et al . 2014 ) , which could weaken correlations . we anticipate progress in several directions . first , it is desired to obtain and analyze the ifs and photometric data in the same extended regions of galaxies . second , we should perform ifs observations on the numerous cluster galaxies for which deep images are already available . although some advances have been made in this field by numerical simulations , they have so far been limited to a small number of galaxies or to semi - analytic models . but , new hydrodynamic simulations of cosmological scale will soon provide a critical clue to the evolution of spin in the hierarchical merger paradigm ( e.g. vogelsberger et al . 2014 ; schaye et al . 2015 ; choi et al . in prep . the sami galaxy survey is based on observations made at the anglo - australian telescope . the sydney - aao multi - object integral field spectrograph ( sami ) was developed jointly by the university of sydney and the australian astronomical observatory . the sami input catalogue is based on data taken from the sloan digital sky survey , the gama survey and the vst atlas survey . the sami galaxy survey is funded by the australian research council centre of excellence for all - sky astrophysics ( caastro ) , through project number ce110001020 , and other participating institutions . the sami galaxy survey website is http://sami - survey.org/. so thanks dwl for his continued support . sky acknowledges support from the korean national research foundation ( nrf-2014r1a2a1a01003730 ) . smc acknowledges the support of an australian research council future fellowship ( ft100100457 ) . jta acknowledges the award of a sief john stocker fellowship . m.s.o . acknowledges the funding support from the australian research council through a future fellowship fellowship ( ft140100255 ) . jvds is funded under bland - hawthorn s arc laureate fellowship ( fl140100278 ) . sb acknowledges the funding support from the australian research council through a future fellowship ( ft140101166 ) . mahajan acknowledges the funding support from the inspire faculty award ( dst / inspire/04/2015/002311 ) and the fast track fellowship ( sb / ftp / ps-054/2013 ) from the department of science and technology ( dst ) and the science education and research board ( serb ) respectively . ns acknowledges support of a university of sydney postdoctoral research fellowship . acknowledges support from the basic science research program through the national research foundation of korea ( nrf ) , funded by the ministry of education ( nrf-2013r1a6a3a04064993 ) . support for amm is provided by nasa through hubble fellowship grant # hst - hf2 - 51377 awarded by the space telescope science institute , which is operated by the association of universities for research in astronomy , inc . , for nasa , under contract nas5 - 26555 . this work has been supported by the yonsei university future - leading research initiative of 2015 ( rms2 2015 - 22 - 0064 ) . abazajian , k. n. , adelman - mccarthy , j. k. , ageros , m. a. , et al . 2009 , , 182 , 543 - 558 abraham , r. g. , tanvir , n. r. , santiago , b. x. , et al . 1996 , , 279 , l47 abraham , r. g. , van den bergh , s. , & nair , p. 2003 , , 588 , 218 arnold , j. a. , romanowsky , a. j. , brodie , j. p. , et al . 2014 , , 791 , 80 barnes , j. e. , & hernquist , l. 1992 , , 30 , 705 barrera - ballesteros , j. k. , garca - lorenzo , b. , falcn - barroso , j. , et al . 2015 , , 582 , a21 bennert , n. , canalizo , g. , jungwiert , b. , et al . 2008 , , 677 , 846 - 857 bland - hawthorn , j. , ellis , s. c. , leon - saval , s. g. , et al . 2011 , nature communications , 2 , 581 bois , m. , emsellem , e. , bournaud , f. , et al . 2011 , , 416 , 1654 bryant , j. j. , bland - hawthorn , j. , fogarty , l. m. r. , lawrence , j. s. , & croom , s. m. 2014 , , 438 , 869 bryant , j. j. , owers , m. s. , robotham , a. s. g. , et al . 2015 , , 447 , 2857 canalizo , g. , bennert , n. , jungwiert , b. , et al . 2007 , , 669 , 801 cannon , j. m. , mcclure - griffiths , n. m. , skillman , e. d. , & ct , s. 2004 , , 607 , 274 cappellari , m. , & emsellem , e. 2004 , , 116 , 138 cappellari , m. , emsellem , e. , bacon , r. , et al . 2007 , , 379 , 418 cappellari , m. , emsellem , e. , krajnovi , d. , et al . 2011 , , 413 , 813 cappellari , m. , emsellem , e. , krajnovi , d. , et al . 2011 , , 416 , 1680 casteels , k. r. v. , bamford , s. p. , skibba , r. a. , et al . 2013 , , 429 , 1051 catinella , b. , kauffmann , g. , schiminovich , d. , et al . 2012 , , 420 , 1959 colless , m. , dalton , g. , maddox , s. , et al . 2001 , , 328 , 1039 conselice , c. j. 2003 , , 147 , 1 conselice , c. j. , rajgor , s. , & myers , r. 2008 , , 386 , 909 cortese , l. , fogarty , l. m. r. , ho , i .- , et al . 2014 , , 795 , l37 cotini , s. , ripamonti , e. , caccianiga , a. , et al . 2013 , , 431 , 2661 covington , m. d. , kassin , s. a. , dutton , a. a. , et al . 2010 , , 710 , 279 cox , t. j. , jonsson , p. , somerville , r. s. , primack , j. r. , & dekel , a. 2008 , , 384 , 386 croom , s. m. , smith , r. j. , boyle , b. j. , et al . 2004 , , 349 , 1397 croom , s. m. , lawrence , j. s. , bland - hawthorn , j. , et al . 2012 , , 421 , 872 deugenio , f. , houghton , r. c. w. , davies , r. l. , & dalla bont , e. 2013 , , 429 , 1258 davis , t. a. , bureau , m. , young , l. m. , et al . 2011 , , 414 , 968 de rossi , m. e. , tissera , p. b. , & pedrosa , s. e. 2012 , , 546 , a52 de zeeuw , p. t. , bureau , m. , emsellem , e. , et al . 2002 , , 329 , 513 den heijer , m. , oosterloo , t. a. , serra , p. , et al . 2015 , , 581 , a98 dressler , a. , oemler , a. , jr . , couch , w. j. , et al . 1997 , , 490 , 577 driver , s. p. , hill , d. t. , kelvin , l. s. , et al . 2011 , , 413 , 971 dubois , y. , pichon , c. , welker , c. , et al . 2014 , , 444 , 1453 duc , p .- a . , cuillandre , j .- c . , karabal , e. , et al . 2015 , , 446 , 120 duc , p .- a . , cuillandre , j .- c . , serra , p. , et al . 2011 , , 417 , 863 ellison , s. l. , patton , d. r. , mendel , j. t. , & scudder , j. m. 2011 , , 418 , 2043 emsellem , e. , cappellari , m. , krajnovi , d. , et al . 2011 , , 414 , 888 emsellem , e. , cappellari , m. , krajnovi , d. , et al . 2007 , , 379 , 401 flores , h. , hammer , f. , puech , m. , amram , p. , & balkowski , c. 2006 , , 455 , 107 fogarty , l. m. r. , scott , n. , owers , m. s. , et al . 2014 , , 443 , 485 foster , c. , pastorello , n. , roediger , j. , et al . 2016 , , 457 , 147 franx , m. , illingworth , g. , & de zeeuw , t. 1991 , , 383 , 112 hoffman , l. , cox , t. j. , dutta , s. , & hernquist , l. 2010 , , 723 , 818 houghton , r. c. w. , davies , r. l. , deugenio , f. , et al . 2013 , , 436 , 19 ji , i. , peirani , s. , & yi , s. k. 2014 , , 566 , a97 jimmy , tran , k .- v . , brough , s. , et al . 2013 , , 778 , 171 kannappan , s. j. , fabricant , d. g. , & franx , m. 2002 , , 123 , 2358 kassin , s. a. , weiner , b. j. , faber , s. m. , et al . 2007 , , 660 , l35 kaviraj , s. 2010 , , 408 , 170 kaviraj , s. 2010 , , 406 , 382 khim , h .- g . , park , j. , seo , s .- w . , et al . 2015 , , 220 , 3 khochfar , s. , emsellem , e. , serra , p. , et al . 2011 , , 417 , 845 kimm , t. , devriendt , j. , slyz , a. , et al . 2011 , arxiv:1106.0538 koribalski , b. s. , & lpez - snchez , . r. 2009 , , 400 , 1749 krajnovi , d. , alatalo , k. , blitz , l. , et al . 2013 , , 432 , 1768 krajnovi , d. , cappellari , m. , de zeeuw , p. t. , & copin , y. 2006 , , 366 , 787 krajnovi , d. , emsellem , e. , cappellari , m. , et al . 2011 , , 414 , 2923 kronberger , t. , kapferer , w. , schindler , s. , & ziegler , b. l. 2007 , , 473 , 761 lee , j. , & yi , s. k. 2013 , , 766 , 38 lee , y. , rey , s .- c . , hilker , m. , sheen , y .- k . , & yi , s. k. 2016 , , 822 , 92 lpez - snchez , . r. 2010 , , 521 , a63 lpez - snchez , . r. , koribalski , b. s. , van eymeren , j. , et al . 2012 , , 419 , 1051 lotz , j. m. , primack , j. , & madau , p. 2004 , , 128 , 163 naab , t. , oser , l. , emsellem , e. , et al . 2014 , , 444 , 3357 puech , m. , hammer , f. , flores , h. , et al . 2010 , , 510 , a68 rampazzo , r. , plana , h. , amram , p. , et al . 2005 , , 356 , 1177 snchez , s. f. , kennicutt , r. c. , gil de paz , a. , et al . 2012 , , 538 , a8 satyapal , s. , ellison , s. l. , mcalpine , w. , et al . 2014 , , 441 , 1297 schaye , j. , crain , r. a. , bower , r. g. , et al . 2015 , , 446 , 521 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , , 500 , 525 scott , n. , davies , r. l. , houghton , r. c. w. , et al . 2014 , , 441 , 274 shapiro , k. l. , genzel , r. , frster schreiber , n. m. , et al . 2008 , , 682 , 231 - 251 sharma , s. , steinmetz , m. , & bland - hawthorn , j. 2012 , , 750 , 107 sharp , r. , allen , j. t. , fogarty , l. m. r. , et al . 2015 , , 446 , 1551 sharp , r. , saunders , w. , smith , g. , et al . 2006 , , 6269 , 62690 g sheen , y .- k . , yi , s. k. , ree , c. h. , & lee , j. 2012 , , 202 , 8 stil , j. m. , & israel , f. p. 2002 , , 392 , 473 tal , t. , van dokkum , p. g. , nelan , j. , & bezanson , r. 2009 , , 138 , 1417 toomre , a. , & toomre , j. 1972 , , 178 , 623 tully , r. b. , & fisher , j. r. 1977 , , 54 , 661 van den bosch , r. c. e. , & van de ven , g. 2009 , , 398 , 1117 van dokkum , p. g. 2005 , , 130 , 2647 vogelsberger , m. , genel , s. , springel , v. , et al . 2014 , , 444 , 1518 way , m. j. , quintana , h. , & infante , l. 1997 , arxiv : astro - ph/9709036 williams , m. j. , bureau , m. , & cappellari , m. 2010 , , 409 , 1330 wong , t. , blitz , l. , & bosma , a. 2004 , , 605 , 183 yang , y. , flores , h. , hammer , f. , et al . 2008 , , 477 , 789 yi , s. k. , lee , j. , jung , i. , ji , i. , & sheen , y .- k . 2013 , , 554 , a122 york , d. g. , adelman , j. , anderson , j. e. , jr . , et al . 2000 , , 120 , 1579 in figure [ allimg ] , we present the @xmath141band deep images for the 63 galaxies used in this study . our visual classifications are presented on each image .

at the end of their life , stars of low- and intermediate - mass evolve along the asymptotic giant branch , where they experience recurrent thermal instabilities and substantial mass loss . during this phase , the stars undergo a very rich and unique nucleosynthesis . moreover , recurrent dredge - up events enrich the stellar surface with the freshly synthesized nuclides which are then ejected into the interstellar medium through the strong winds . the thermally pulsing agb stars thus play a crucial role in the chemical evolution of galaxies . on the other hand , agb stars are important contributors to the integrated luminosity of stellar systems in a wide range of ages . a coverage of this phase is needed in particular for the treatment of the luminosity functions and color magnitude diagrams of star clusters . last but not least , the corresponding stellar models are basic tools to study carbon stars , oh / ir stars , planetary nebulae and their central objects ( white dwarfs ) . in the last 15 years , extensive observational data has become available that changed and highly constrained the theoretical view of the agb evolution . most of the progresses came from magellanic cloud studies , which revealed for example the absence of very luminous carbon stars ( which were previously predicted to originate from stars with initial masses between 5 and @xmath5 ) , and the unexpected existence of relatively faint and lower mass carbon stars ( blanco et al . 1980 , mould & aaronson 1982 , 1986 ; wood et al . 1983 ; reid & mould 1984 ; aaronson & mould 1985 ; smith & lambert 1989 ) . this so - called `` carbon star mystery '' ( iben 1981 ) has led to the concept of hot - bottom convective envelope burning ( iben 1975 ; sackmann et al . 1974 ; scalo et al . 1975 ; sackmann & boothroyd 1992 ) . this mechanism , which reconverts c - rich to o - rich envelopes and produces @xmath2 , was confirmed by the existence of lithium enriched agb stars ( smith & lambert 1989 , 1990b ; plez et al . 1993 ; smith et al . 1995 ) . on the other hand , the discovery , inside meteorites , of interstellar grains which have been formed in the wind of cool carbon stars ( see e.g. zinner et al . 1991a ) , complemented the stellar abundance analysis , and brought crucial informations on agb star nucleosynthesis . in spite of important improvements , many important issues of agb evolution still remain to be addressed , observationally as well as theoretically . this paper is the first of a series aimed at examining all the physical processes that may induce abundance anomalies in the surface of evolved agb stars . in the present work , we specifically investigate the case of intermediate - mass stars . stars with initial masses between 3 and @xmath0 and with metallicities @xmath6 of 0.02 and 0.005 have been evolved , starting from the pre - main sequence until they had gone through a number of helium - shell flashes along the tp - agb . the @xmath0 ( 6 and @xmath0 ) stars with @xmath7 ( 0.005 ) ignite off - center carbon burning and will not be discussed in detail . for each other stellar mass , full evolutionary models were computed up to the fourth `` full amplitude '' thermal pulse in the so - called `` asymptotic regime '' . synthetic evolution was then used to follow global surface properties and nucleosynthesis up to the planetary nebula ejection . this self - consistent and extended set of models is particularly relevant for population synthesis and chemical evolution models . very complete figures and tables containing informations about the center and surface evolution of the structure and chemical composition for all our models are available . in the present computations , which include the latest input physics , the evolution of the chemical composition is followed by only taking into account nuclear reactions and mixing inside convective zones ( we assume instantaneous mixing , except when hot - bottom burning occurs , which requires a time - dependent convective diffusion algorithm ) . we neglect all other mixing mechanism of non - convective origin which may contribute to transport chemical elements in radiative regions , like rotation - induced mixing or diffusion . in the stellar mass range we consider , this approximation is supported by the observed surface abundances at different evolutionary stages . this is however certainly not the case for lower mass stars , for which some `` non - standard '' mixing processes have to be invoked to explain various chemical anomalies . the physical ingredients of our models and numerical aspects of the computations are presented in sects . 2 and 3 , respectively . the general properties of the stars and their chemical evolution during the phases prior to the tp - agb one are summarized in sects . 4 and 5 , respectively . in sect . 6 , we briefly present the structural features of the thermal pulses and discuss the occurrence of the third dredge - up in our models . we derive a core - mass luminosity relation for our less massive stars that do not undergo hot - bottom burning . we then present the extrapolation procedure that we use to follow both the evolution and the nucleosynthesis in the asymptotic regime " up to the tp - agb tip . we also give the initial mass final mass relation resulting from our models , that appears to be in good agreement with the observed one . in sect . 7 , we review the various nucleosynthesis sites inside tp - agb stars , namely the burning shells ( he , @xmath8 and h ) , the convective tongue that develops during a thermal runaway and the base of hot convective envelopes . we investigate in detail the nucleosynthetic processes that involve the li , cno , f , nena and mgal elements and neutrons . we then discuss which specific nuclear region , inside a tp - agb star , mainly contributes to the surface abundance change of each nuclide when a third dredge - up occurs . in sect . 8 , we finally present our predictions concerning _ ( i ) _ the evolution of all the surface isotopic ratios by including our evolutionary models and synthetic extrapolations up to the convective envelope removal and _ ( ii ) _ the chemical yields ejected at different ages for all the species up to @xmath9 . as in the previous sections , comparisons between our predictions and observations are attentively performed . in this paper version , sects . 2 , 3 , 5 and 8 have been somewhat reduced . our equation of state is analytic . it includes * the ionization of h , he , c , n and o. complete ionization is assumed when oix / oviii @xmath10 ; * the electrostatic corrections through the statistical model of debye - hckel for partially ionized regions and through the interpolation formalism of koester ( 1976 ) , between the debye - hckel and the thomas - fermi models , for the completely ionized matter ; * the electronic degeneracy as soon as its contribution to the total pressure exceeds @xmath11 ; * the production of neutrino pairs through four different processes , namely the photo - neutrinos , plasma neutrinos , @xmath12 pair annihilations and bremsstrahlung neutrinos . energy loss contributions comes from munakata et al . ( 1985 ) , itoh et al . ( 1992 ) and kohyama et al . ( 1993 ) . at low temperatures , i.e. below 8000 k , we use the alexander & ferguson ( 1994 ) opacity tables , very well suited for cool red giant envelopes and atmospheres . at temperatures above 8000 k , we use the opal opacity tables computed by rogers & iglesias ( 1992 ) . the abundance distribution of the heavy elements being scaled from the solar composition given by anders & grevesse ( 1989 ) and updated by grevesse ( 1991 ) , as for low - temperature opacities . for both sets of opacity tables , we first interpolate bi - linearly in @xmath13 and @xmath14 for each table the composition of which is close to the model composition and then linearly in @xmath15 , @xmath6 ( and @xmath16o@xmath17 ) to obtain the right opacity coefficient corresponding to each shell . finally , we used the hubbard & lampe ( 1969 ) program to generate conductive opacity tables corresponding to the same chemical compositions and @xmath18 grid points as for the opal tables . the stellar structure is integrated from the center to a very low optical depth ( @xmath19 ) in the atmosphere . we constraint the atmospheric temperature profile to correspond to those coming from realistic atmosphere models obtained by integrating the radiative transfer equation . more specifically , we use atmosphere models _ ( 1 ) _ from plez ( 1992 ) for @xmath20@xmath21 k , _ ( 2 ) _ from eriksson ( 1994 , private communication , using a physics similar to bell et al . 1976 ) up to 5500 k and _ ( 3 ) _ computed with the kurucz atmosphere program above 5500 k. at each time step , we linearly interpolate a @xmath22 profile corresponding to our model @xmath20 , @xmath23 and @xmath6 from these grids of atmosphere models and then correspondingly modify the radiative pressure and gradient . the structure of the convective regions is computed using the classical mixing - length theory ( mlt ) , as prescribed by kippenhahn et al . our models are standard in the sense that _ ( 1 ) _ the schwarzschild criterion is considered to delimit the convective zones and _ ( 2 ) _ neither overshooting nor semi - convection have been considered in the present computations . the ratio @xmath24 of the mixing - length free parameter over the pressure scale height has been put to a value of 1.5 , i.e. rather close to @xmath25 with which we fit the solar structure . > from the main sequence up to the thermally pulsing agb phase , we used the empirical reimers ( 1975 ) formula @xmath26 with @xmath27 up to the central he exhaustion . > from the beginning of the agb phase , we adopted variable @xmath28 values , depending on the total mass of the star . more specifically , we took @xmath29 , 3 , 3.5 or 4 for our 3 , 4 , 5 or @xmath30 models respectively , whatever @xmath6 . this @xmath28 increase with the total mass was found by bryan et al . ( 1990 ) to give rather good results along the agb . however , later during the thermally pulsing phase , blcker ( 1995 ) recently showed that the bryan et al . ( 1990 ) formula underestimates the observed mass loss rate . this leads us to further increase @xmath28 ( see sect . however , as mentioned by vassiliadis & wood ( 1993 ) , whatever empirical formula , the mass loss rate increase used in the models along the thermally pulsing agb phase still remain uncertain by more than a factor of two . as demonstrated in sect . 8 , this important uncertainty on the mass loss rate formula conditions the agb phase duration and consequently , the surface composition evolution through the successive third dredge - up events . we note that along the thermally pulsing agb phase , the thermal pulse features slightly depend on the way one estimates the rate @xmath31 of gravitational energy release or gain ( see e.g. kippenhahn & weigert 1991 ) . so , let us specify that we wrote @xmath32 where @xmath33 is the specific ( i.e. by unit mass ) internal energy . we follow the abundance evolution , through the whole stars , of 45 nuclides , namely the neutrons , all the 31 stable nuclides up to @xmath34 as well as @xmath35 , @xmath36 , @xmath37 , @xmath38 , @xmath39 , @xmath40 , @xmath41 , @xmath42 , @xmath43 , @xmath44 , @xmath45 , @xmath46 and @xmath47 . these nuclides interact through a network containing 172 nuclear ( neutron , proton and @xmath48 captures ) and decay reactions . many nuclear reaction rates are taken from caughlan & fowler ( 1988 ) . let us just mention the nuclear reactions for which we selected another rate prescription . * @xmath49 from nagai et al . ( 1991 ) , * @xmath50 from caughlan et al . ( 1985 ) , * @xmath51 from raman et al . ( 1990 ) , * @xmath52 from beer et al . ( 1991 ) , * @xmath53 from wiesher et al . ( 1990 ) , * @xmath54 from funck & langanke ( 1989 ) , * @xmath55 , @xmath56 , @xmath57 and @xmath58 from fowler et al . ( 1967 ) , * @xmath59 from brehm et al . ( 1988 ) , * @xmath60 from de oliveira ( 1995 ) , * @xmath61 from allen et al . ( 1971 ) , * @xmath62 from wagoner ( 1969 ) , * @xmath63 and @xmath64 from landr et al . ( 1990 ) , * @xmath65 from thielemann ( 1991 ) , * @xmath66 and @xmath67 from bao & kppeler ( 1987 ) , * @xmath68 from kious ( 1990 ) , * @xmath69 , @xmath70 , @xmath71 , @xmath72 , @xmath73 and @xmath74 from beer et al . ( 1991 ) , * @xmath75 and @xmath76 from hammer ( 1991 ) , * @xmath77 , @xmath78 , @xmath79 and @xmath80 from woosley et al . ( 1978 ) , * @xmath81 and @xmath82 from champagne et al . ( 1988 ) , * @xmath83 and @xmath84 from illiadis ( 1990 ) , * and @xmath85 and @xmath86 from vogelaar ( 1989 ) . on the other hand , the nuclear reaction rates for @xmath87 , @xmath88 , @xmath89 , @xmath90 , @xmath91 , @xmath92 , @xmath93 , @xmath94 as well as the 45 nuclear reactions involving si , p and s are fitted with a procedure developed by rayet ( 1993 , private communication ) let us finally mention recent improvements not yet incorporated in our network : various @xmath95 reactions rates on light nuclei ( nagai et al . 1995 ) and @xmath24 capture rates on @xmath96 by kppeler et al . ( 1994 ) . finally , the nuclear screening factors are parameterized by using the graboske et al . ( 1973 ) formalism , including weak , intermediate and strong screening cases . neutrons can be produced during the thermally pulsing agb phase . they can , in particular , be captured by nuclides heavier than @xmath34 , mainly through @xmath95 reactions . as we do not follow the abundance of such elements , we created an additional nuclide , called @xmath97 , the mass fraction of which is set to @xmath98 this sum counts in fact 253 stable nuclides up to @xmath99 . in order to have a good estimation of the free neutron abundance , we follow a method described by jorissen & arnould ( 1989 ) . if nuclear reactions occur inside a convective zone , most of them proceed in general with a longer time scale than the turn - over time scale associated to the motion of convective cells . in such a case , one usually consider the convective mixing as instantaneous . this allows to treat nucleosynthesis in one shot , by taking mass - weighted averages of the number abundances and nuclear reaction rates over the whole convective region . however , if some key nuclides nuclearly evolve more rapidly than they are mixed , we have to consider the convection transport together with the nuclear reactions , so that nucleosynthesis equations become diffusion equations of the form @xmath100 \ , + \ , \makebox{nuclear terms } \,,\ ] ] where @xmath101 , the ratio of the mass fraction over the atomic mass of each nuclide @xmath102 , and @xmath103 , the convective diffusion coefficient , with @xmath104 and @xmath105 being the mean velocity and turn - over time of the convective cells , respectively . it is very cpu time consuming to treat the nucleosynthesis and convective mixing together through such diffusion equations for each nuclide in each convective shell for all the time steps . consequently , as usual , instantaneous mixing has been assumed by default for our stellar evolution computations , except in two cases that potentially require a time - dependent treatment of the convective mixing : _ ( 1 ) _ the convective tongue associated with thermal pulses and _ ( 2 ) _ the base of the convective envelope when it is hot enough , both along the agb phase . we verified that inside thermal pulses , the elements that are the most affected by a diffusion treatment of the convective mixing are the very unstable nuclides ( like @xmath42 ) , neutrons and protons . however , this do not change very significantly the nucleosynthesis global results ( see however sect . 7.5 ) . on the contrary , the mean turn - over time scale of the convective motions is much longer inside the convective envelope than inside a thermal pulse . if nuclear burning occurs at the base of the convective envelope , some important nuclear reactions can , depending on the bottom temperature , occur faster than the convective mixing so that a diffusion treatment is required . this is in particular the case for some of the reactions involved in the @xmath2 synthesis ( see sect . 7.2 ) . we have built initial models of 3 , 4 , 5 , 6 and @xmath0 allowing us to start our computations at the beginning of the pre - main sequence phase ( along the hayashi track ) , when central temperatures are below @xmath106 k. for our @xmath1 ( 0.02 ) models , the hydrogen mass fraction has been set to @xmath107 ( 0.687 ) and that of helium to @xmath108 ( 0.293 ) . the abundances of the heavier elements have been scaled on the anders & grevesse ( 1989 ) abundance distribution . we solve the five one - dimensional hydrodynamical equations for the evolution of the stellar structure with the mass coordinate as independent variable . we first convert the five non - linear stellar structure differential equations into difference equations following the henyey s scheme by defining discrete shells in mass . we then solve the resulting linearized system of equations by the newton - raphson relaxation procedure ( see e.g. press et al . typically , 600 shells are required for a model on the main sequence phase , while along the agb , 1000 to 1300 shells are needed ( 200 to 500 more shells being added during a thermal pulse ) . on the other hand , typically 200 ( 700 ) models are necessary to model the central hydrogen ( helium ) burning phase . during the agb phase , 2500 to 4500 models are required to model the time separating two successive thermal pulses , while the thermal pulse itself requires 100 to 350 models . let us stress that these rather small numbers of mass shells and time steps have been chosen in order to save computer time and present a very large set of nucleosynthesis predictions . straniero et al . ( 1996 ) very recently showed that this is probably in the disadvantage of the occurrence of the third dredge - up along the tp - agb phase . finally , the mass @xmath109 lost during each time step @xmath110 is suppressed proportionally to the shell masses of the whole envelope above the burning regions . the nucleosynthesis equations are solved using the wagoner ( 1969 ) numerical technique , well suited for our purposes . however , if a diffusion treatment is required to simultaneously compute the time - dependent convective mixing and the nuclear burning , the diffusion eqs . ( 4 ) are then solved with the same newton - raphson algorithm . theoretical hertzsprung - russell diagrams ( hrd ) for the ensemble of our models are shown in fig . [ hrd ] , from the early pms phase up to the beginning of the first thermal pulse ( or he - shell flash ) along the agb phase . the lifetimes are given in table [ times ] for the pms , the h - burning phase ( from the zero - age - main sequence to central h - exhaustion ) , the rgb phase ( from central h - exhaustion to central he - ignition ) , the he - burning phase ( up to central he - exhaustion ) , the early ( e ) agb phase ( up to the first thermal pulse ) and the thermally pulsing ( tp ) agb phase . this last duration includes our extrapolation procedure up to the convective envelope exhaustion ( see sect . 6.3 ) . lccccccc & & + & & + & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath110 [ pms ] @xmath114&1.20@xmath115&5.88@xmath116&3.70@xmath116&2.19@xmath116&1.10@xmath115&5.82@xmath116&3.50@xmath116 + @xmath110 [ ms ] @xmath114&2.80@xmath117&1.34@xmath117&7.69@xmath115&5.09@xmath115&2.59@xmath117&1.30@xmath117&7.88@xmath115 + @xmath110 [ rg ] @xmath114&1.78@xmath115&6.39@xmath116&2.86@xmath116&1.22@xmath116&1.45@xmath115&5.17@xmath116&2.38@xmath116 + @xmath110 [ hb ] @xmath114&1.55@xmath117&5.26@xmath115&2.64@xmath115&1.08@xmath115&9.37@xmath115&4.30@xmath115&2.59@xmath115 + @xmath110 [ e - agb ] @xmath114&6.16@xmath116&2.88@xmath116&1.25@xmath116&5.98@xmath118&4.29@xmath116&1.69@xmath116&6.41@xmath118 + @xmath110 [ tp - agb ] @xmath114&1.22@xmath116&2.26@xmath118&1.33@xmath118&9.12@xmath119&9.03@xmath118&1.59@xmath118&7.97@xmath119 + the occurrence of the first thermal pulse marks the end of the e - agb phase and the beginning of the tp - agb phase . this precise time , marked by the maximum surface luminosity just preceding the first major thermal instability , will be called @xmath120 in the following . table [ t0 ] presents some characteristics of our models at @xmath120 . lccccccc & & + & & + & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath121&4.739@xmath117&2.034@xmath117&1.123@xmath117&6.624@xmath115&3.851@xmath117&1.880@xmath117&1.117@xmath117 + @xmath122 & 3921&13274&19367&25743 & 5420&14761&21479 + @xmath123&182&398&486&554&182&339&387 + @xmath124 \ , ( \rm{r_{\odot}})$ ] & 3&15&17&18 & 3&10&10 + @xmath125&2.92&3.77&4.70&5.72&2.81&3.51&4.31 + @xmath126&2.0@xmath127&1.5@xmath128&2.5@xmath128&3.6@xmath128&3.6@xmath127&1.8@xmath128&2.8@xmath128 + @xmath129&3381&3102&3088&3104&3667&3450&3550 + @xmath130 & 0.38 & -0.19 & -0.26 & -0.29 & 0.36 & -0.08 & -0.10 + @xmath131 & 0.96 & 0.87 & 0.80 & 0.71 & 0.94 & 0.79 & 0.66 + @xmath132 & 0.09 & 0.08 & 0.07 & 0.06 & 0.06 & 0.13 & 0.10 + @xmath133&20&30&37&39&20&32&39 + @xmath134&2.11@xmath117&3.48@xmath117&3.97@xmath117&4.60@xmath117&2.44@xmath117&3.91@xmath117&4.54@xmath117 + @xmath135 & 0.320 & 0.521 & 0.643 & 0.745 & 0.373 & 0.608 & 0.736 + @xmath136 \ , ( \rm{r_{\odot}})$ ] & 0.00008 & 0.00003 & 0.00003 & 0.00002 & 0.00008 & 0.00004 & 0.00003 + @xmath137 \ , ( \rm{m_{\odot}})$ ] & 0.520 & 0.732 & 0.815 & 0.894 & 0.594 & 0.814 & 0.903 + @xmath138 \ , ( \rm{m_{\odot}})$ ] & 0.08 & 0.03 & 0.02 & 0.24 & 0.04 & 0.01 & 0.23 + @xmath139 \ , \rm{(k)}$]&1.43@xmath117&1.61@xmath117&1.64@xmath117&1.68@xmath117&1.37@xmath117&1.64@xmath117&1.68@xmath117 + @xmath140 \ , \rm{(g \ , cm^{-3})}$]&13.8 & 7.7 & 5.9 & 5.9&11.2 & 5.3 & 4.6 + @xmath141 \ , ( \rm{m_{\odot}})$]&4.0@xmath142&4.7@xmath143&1.5@xmath143&6.0@xmath144&3.3@xmath142&3.8@xmath143&1.2@xmath143 + @xmath145 \ , ( \rm{r_{\odot}})$ ] & 0.00016 & 0.00006 & 0.00004 & 0.00003 & 0.00015 & 0.00006 & 0.00005 + @xmath146 \ , ( \rm{m_{\odot}})$ ] & 0.542 & 0.738 & 0.818 & 0.895 & 0.609 & 0.818 & 0.905 + @xmath147 \ , ( \rm{m_{\odot}})$]&1.2@xmath148&3.2@xmath149&1.8@xmath149&5.5@xmath148&9.7@xmath149&3.4@xmath149&2.2@xmath149 + @xmath150 \ , \rm{(k)}$]&5.07@xmath115&6.43@xmath115&7.10@xmath115&7.77@xmath115&5.75@xmath115&7.00@xmath115&7.74@xmath115 + @xmath151 \ , \rm{(g \ , cm^{-3})}$]&2.0@xmath152&1.4@xmath152&1.3@xmath152&1.1@xmath152&2.4@xmath152&1.7@xmath152&1.6@xmath152 + @xmath153 \ , ( \rm{r_{\odot}})$ ] & 0.0042 & 0.0008 & 0.0002 & 0.0001 & 0.0044 & 0.0008 & 0.0002 + @xmath154 \ , ( \rm{m_{\odot}})$ ] & 0.544 & 0.738 & 0.818 & 0.896 & 0.612 & 0.819 & 0.905 + @xmath155 \ , \rm{(k)}$]&2.68@xmath116&5.72@xmath116&1.58@xmath115&4.27@xmath115&2.71@xmath116&6.53@xmath116&2.25@xmath115 + @xmath156 \ , \rm{(g \ , cm^{-3})}$]&1.1@xmath148&4.1@xmath148&5.7@xmath152 & 0.9&1.0@xmath148&6.0@xmath148 & 0.2 + for the main sequence duration , our results for @xmath7 are in better agreement with bressan et al . ( 1993 ) predictions than with the slightly higher values given by schaller et al . the central helium burning lifetime crucially depends on the amount of overshooting . as our stellar models are computed without overshooting , we can only compare our predictions at this phase with those of the padova group calculated in the same conditions ; we obtain very similar results . the other important quantity for the agb phase is the core mass ( i.e. the mass contained up to the hbs ) at the end of the central he - burning phase . indeed , its value determines the surface luminosity ( through the core mass luminosity relation ; see e.g. boothroyd & sackmann 1988a ) , the mass loss rate , and consequently the agb phase duration . last but not least , along the tp - agb phase , the thermal pulses are stronger and the third dredge - up deeper when the core mass is higher . our @xmath111 star with @xmath7 enter the agb phase with a core mass of @xmath157 ( see table [ t0 ] ) , a value very close to that of the comparable model of boothroyd & sackmann ( 1988a ) or lattanzio ( 1986 ) , also computed without overshooting . always for a reference @xmath111 star , the tp - agb phase of the straniero et al . ( 1995 ) models begins with a core mass of @xmath158 ( 2 % lower ) . at a comparable evolution stage however , the @xmath113 tp - agb star of vassiliadis & wood ( 1993 ) , also computed without any kind of extra - mixing , has a core mass @xmath159 % higher than ours . this could be due to the somewhat lower metallicity of their models ( @xmath160 ) . in conclusion , the core mass at the beginning of the tp - agb phase , although very important for this phase , still depends on various uncertainties related to the treatment of convection during the previous evolutionary phases . during the first and second dredges - up , convective mixing and induced dilution modify the surface abundances which were unaltered until this phase ( iben 1964 ) . table [ dup12 ] presents the resulting changes for the principal elements altered by h - burning . let us emphasize the main points . a second dredge - up occurs in our @xmath161 @xmath112 models , also leading to surface abundance changes . for our @xmath161 @xmath113 models , the convective envelope penetrates so deep that it pushes down the h he discontinuity . llccccccc & & & + & & & + & & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath162&_initial_&3.22@xmath163&3.19@xmath163&3.13@xmath163&3.12@xmath163&2.71@xmath163&2.71@xmath163&2.66@xmath163 + & _ top rg_&1.55@xmath149&1.05@xmath149&8.52@xmath163&7.32@xmath163&1.71@xmath149&1.28@xmath149&1.08@xmath149 + & _ at @xmath164_&1.54@xmath149&1.03@xmath149&7.79@xmath163&6.65@xmath163&1.66@xmath149&1.13@xmath149&8.56@xmath163 + @xmath48&_initial _ & 0.2932 & 0.2932 & 0.2932 & 0.2932 & 0.2495 & 0.2495 & 0.2495 + & _ top rg _ & 0.3047 & 0.3039 & 0.2995 & 0.3017 & 0.2585 & 0.2505 & 0.2498 + & _ at @xmath164 _ & 0.3054 & 0.3052 & 0.3367 & 0.3421 & 0.2624 & 0.2629 & 0.3014 + @xmath2&_initial_&9.90@xmath165&9.90@xmath165&9.90@xmath165&9.90@xmath165&2.47@xmath165&2.47@xmath165&2.47@xmath165 + & _ top rg_&1.44@xmath166&1.44@xmath166&1.43@xmath166&1.44@xmath166&3.14@xmath167&3.51@xmath167&3.86@xmath167 + & _ at @xmath164_&1.20@xmath166&9.85@xmath167&7.39@xmath167&7.53@xmath168&1.49@xmath167&1.68@xmath167&3.59@xmath169 + @xmath170/@xmath8&_initial_&83.1&83.1&83.1&83.1&83.1&83.1&83.1 + & _ top rg_&19.3&19.0&18.8&18.5&18.7&18.4&18.3 + & _ at @xmath164_&19.1&18.6&17.8&17.5&18.1&17.5&17.4 + @xmath170/@xmath171&_initial_&2.745&2.745&2.745&2.745&2.745&2.745&2.745 + & _ top rg _ & 0.725 & 0.714 & 0.769 & 0.735 & 0.636 & 0.888&1.100 + & _ at @xmath164 _ & 0.712 & 0.684 & 0.565 & 0.538 & 0.581 & 0.610 & 0.496 + @xmath171/@xmath172&_initial _ & 253 & 253 & 253 & 253 & 253 & 253 & 253 + & _ top rg _ & 1243 & 1281 & 1210 & 1279 & 1397 & 1033 & 856 + & _ at @xmath164 _ & 1270 & 1346 & 1654 & 1774 & 1546 & 1510 & 1841 + @xmath170/@xmath173&_initial _ & 0.316 & 0.316 & 0.316 & 0.316 & 0.316 & 0.316 & 0.316 + & _ top rg _ & 0.205 & 0.207 & 0.210 & 0.211 & 0.194 & 0.208 & 0.229 + & _ at @xmath164 _ & 0.204 & 0.205 & 0.206 & 0.207 & 0.191 & 0.192 & 0.197 + @xmath173/@xmath174&_initial_&2467&2467&2467&2467&2467&2467&2467 + & _ top rg _ & 314 & 409 & 562 & 617 & 258 & 869&1814 + & _ at @xmath164 _ & 311 & 391 & 508 & 579 & 245 & 323 & 358 + @xmath173/@xmath96&_initial _ & 442 & 442 & 442 & 442 & 442 & 442 & 442 + & _ top rg _ & 589 & 582 & 577 & 577 & 616 & 573 & 522 + & _ at @xmath164 _ & 591 & 588 & 588 & 588 & 623 & 619 & 611 + @xmath162 is produced at the beginning of the t tauri phase from the initial @xmath175 burning . on the main sequence , a peak of @xmath162 appears , due to the competition of the reactions @xmath176@xmath177 and @xmath178 . during the first dredge - up , the surface abundance of @xmath162 thus increases . however , the higher the stellar mass , the lower the @xmath162 peak , and the lower the surface enhancement of @xmath162 after the first dredge - up . the first ( and second ) dredge - up phase(s ) lead to a decrease of the @xmath170/@xmath8 , @xmath170/@xmath171 , @xmath172/@xmath171 and @xmath173/@xmath174 isotopic ratios , the extent of which depends on both stellar mass and metallicity . * from an initial value equal to 83 , the @xmath170/@xmath8 ratio decreases to @xmath179 . the higher the stellar mass , the lower the final carbon isotopic ratio . + this ratio slightly further decreases after the second dredge - up . * for @xmath7 , the maximum depth of the convective envelope reaches the deepest step of the @xmath171 profile , for @xmath161 @xmath111 stars . this implies that @xmath170/@xmath171 , as well as @xmath172/@xmath171 have almost the same value in these models after the first dredge - up . on the contrary , for the @xmath1 objects , the lower the stellar mass , the less efficient the dredge - up , and consequently , the less @xmath171 is up - heaved to the surface . this explains the trend for the final @xmath180c/@xmath181n ratio versus stellar mass ( see charbonnel 1994 for more details ) . + after the second dredge - up , whatever @xmath6 , the convective envelope of @xmath161 @xmath112 stars penetrates deeper for increasing stellar mass as attested by both @xmath170/@xmath171 and @xmath172/@xmath171 ratios . * the theoretical post dredge - up value of @xmath173/@xmath174 increases with the stellar mass . moreover , since for @xmath1 the deepening convective envelope of our 4 and @xmath113 stars does not reach the regions where the on cycle operated , the post dredge - up values of @xmath173/@xmath174 are higher than for @xmath7 . + after the second dredge - up however , @xmath174 enriched matter is up - heaved in the convective envelope leading to very low @xmath173/@xmath174 . * the predicted post dredge - up value of @xmath173/@xmath96 slightly decreases with increasing stellar mass . our @xmath162 surface abundance increases are in rather good agreement with the predictions by bressan et al . ( 1993 ) . concerning the cno elements , our results are in general good agreement with the predictions by schaller et al . ( 1992 ) , bressan et al . ( 1993 ) and el eid ( 1994 ) . the only important discrepancy between theoretical predictions of the different groups for intermediate - mass stars concerns the post dredge - up @xmath173/@xmath174 ratio , which value highly depends on the adopted @xmath182 and @xmath183 reaction rates . we use the rates given by landr et al . ( 1990 ) , as do schaller et al . ( 1992 ) , and we obtain @xmath173/@xmath174 ratios very similar to those of the geneva group . on the other hand , bressan et al . ( 1993 ) use the lower @xmath174 proton capture rates given by caughlan & fowler ( 1988 ) , and obtain much lower @xmath173/@xmath174 ratios [ see also el eid ( 1994 ) for a discussion of the influence of the adopted @xmath174 destruction rate on the resulting @xmath173/@xmath174 ratio ] . in our intermediate - mass stars , the theoretical post dredge - up values of the carbon isotopic ratio are slightly lower than the observations by gilroy ( 1989 ) in galactic cluster giants . red giants in this evolutionary phase present @xmath173/@xmath174 ratios between 300 to 1000 and @xmath173/@xmath96 ratios in the range 400 to 600 ( harris et al . 1988 ; smith & lambert 1990a ) . this is in very good agreement with our predictions . in particular , our prediction is in perfect agreement with the observed value in @xmath24 uma ( harris et al . 1988 ) , which estimated mass is roughly @xmath111 . this point is in favor of high @xmath174 destruction rates . this agreement between standard models and observations confirms that no extra - mixing ( diffusion , rotation - induced mixing , ) is expected to occur in intermediate - mass stars , as already discussed in charbonnel ( 1994 , 1995 ) . very complete tables containing informations about the center and surface evolution of the structure and chemical composition for our 3 , 4 , 5 , 6 and @xmath0 models with @xmath1 and 0.02 , from the beginning of the pms phase up to the end of the e - agb phase , can be found through the internet network at the following address and path : http://gag.observ-gr.fr/liens/starevol/evol.html more specifically , we present , for each star , the evolution of * the age , mass , mass loss rate , surface luminosity , bolometric magnitude , effective temperature , total radius , surface density and gravity ; * the central temperature , density , pressure , degeneracy parameter and the mass fraction of @xmath184 , @xmath48 , @xmath170 and @xmath173 ; * the mass and radius positions of the central convective core boundary and of the bottom of the convective envelope as well as its temperature and density ; * the nuclear luminosities of the h- and he - burning shells ; * the mass and radius positions of the h- and he - burning shells ; * the surface mass fractions of @xmath184 , @xmath162 , @xmath48 , @xmath2 , @xmath185 , @xmath170 , @xmath8 , @xmath171 , @xmath172 , @xmath173 , @xmath174 , @xmath96 , @xmath3 , @xmath186 , @xmath187 , @xmath188 , @xmath4 , @xmath189 , @xmath190 , @xmath191 , @xmath43 , @xmath192 as the stars are ascending the agb , the features of the thermal pulses evolve , and some of the main characterizing quantities of these events reach asymptotic values " . the most important feature is the thermal pulse intensity that can be defined in terms of the hebs luminosity @xmath194 reached at the maximum of the he - shell flash . @xmath194 is quite low during the first thermal pulses , but it rapidly increases with core mass ( or flash number ) . the increasing rate is greater for lower metallicities or higher stellar masses . after typically 10 ( 5 ) thermal pulses in our @xmath195 ( 0.005 ) models , the thermal pulse intensity is very high and only slightly increases from pulse to pulse , monotonously with the core mass . this second part of the tp - agb phase is usually called the `` asymptotic regime '' or `` full amplitude '' of the thermal pulses . for each of the modeled stars , we have computed at least four full amplitude thermal pulses . let us briefly discuss the main evolutionary properties of these thermal runaways . the thermal runaways are stronger inside stars of lower @xmath6 , which facilitates dredge - up events ( see sect . the intensities of our he - shell flashes are comparable to other published values for stars of about the same core masses ( see e.g. boothroyd & sackmann 1988b and table 4 ) . while the stars evolve along the tp - agb , the hebs and hbs advance in mass , and the mean inter - shell mass slightly decreases with time . from one pulse to another , the top of the convective tongue at its maximum extension gets closer from the hbs , even if contact between both regions has never been found due to the fact that the hbs is still active at that time . moreover , for all the thermal runaways we have computed , the overlap in mass between two successive thermal pulses is ranged between 0.6 and 0.4 for agb stars of increasing total mass . this overlap is slightly larger for our lower @xmath6 models . the maximum ( minimum ) temperature @xmath196 ( @xmath197 ) and density @xmath198 ( @xmath199 ) reached at the base ( top ) of a thermal pulse both increase ( decrease ) from pulse to pulse . for a given he - shell flash , @xmath196 increases with stellar mass and metallicity , while @xmath198 decreases with stellar mass but increases with metallicity . in the asymptotic regime , the growth rate of @xmath196 also tends to an asymptotic value . by comparison with other works for the @xmath111 ( z = 0.02 ) star , we found that our @xmath196 are typically 4 % higher than those of boothroyd & sackmann ( 1988b ) and straniero et al . . this could be mainly due to _ ( i ) _ different prescriptions to compute @xmath31 , _ ( ii ) _ different envelope masses due to different mass loss rates and _ ( iii ) _ different mlt treatments ( see sect . the temperature and density differences between the base and top of each convective tongue are rather large . this point is important for nucleosynthesis purposes ( see sect . the duration of the convective tongue associated with each thermal pulse ( @xmath200 ) and , to a lower extent , that of the inter - pulse phase ( @xmath201 ) , are crucial quantities for the nucleosynthesis too . the stronger a he - flash , the shorter it is and the quickest it deactivates . both quantities are thus smaller for higher core masses and lower metallicities . moreover , @xmath200 tends to decrease as the stars evolve along the tp - agb . @xmath201 also decreases after the first he - shell flashes , but it then reaches an almost constant value during the asymptotic regime , only very slightly decreasing with the increasing core mass . along the tp - agb phase , the bottom of the convective envelope can penetrate deep inside regions that have been nuclearly processed just after a thermal pulse , if it is strong enough ( in order to give rise to a large enough entropy increase in the inter - shell region ) . contrarily to the first and second dredges - up , this so - called third dredge - up , when it occurs , successively mixes up to the surface _ ( i ) _ material that has experienced h - burning in the thin hbs but also _ ( ii ) _ part of the region where the thermal pulse nucleosynthesis operated just before . both regions give rise to very different and sometimes opposite chemical pollutions of the convective envelope ( see sects . 7.1 and 7.6 ) . in our @xmath7 models , the hbs region is already dredged - up from the 14th , 5th and 3rd thermal pulse for the 3 , 4 and @xmath113 stars , respectively , while for the @xmath1 models , such a mixing already occurs from the 2nd and 1st thermal pulse for the 3 and @xmath112 stars , respectively . the @xmath30 ( @xmath113 ) star with @xmath7 ( 0.005 ) continuously mixes h - burning products up to the surface as it experiences h - burning at the bottom of its convective envelope ( the so - called `` hot - bottom burning '' or hbb ; see sect this kind of dredge - up already modifies the surface composition for some species ( like @xmath43 ; see sects . 7.4 and 7.6 ) . the complete third dredge - up ( one usually simply calls the third dredge - up , 3dup hereafter ) , i.e. reaching the upper part of the previous thermal pulse region , is more difficult to obtain numerically . following models ( wood 1981 ; iben 1983 ; lattanzio 1987 ; boothroyd & sackmann 1988c ) , the dredge - up of thermal pulse material is easier for stars of lower metallicity @xmath6 . this is supported by observations ( see e.g. blanco et al . 1978 ) . in our models , we just found a 3dup from the 9th thermal pulse of the @xmath112 with @xmath1 star . however , based on the evolution of the post - thermal pulse entropy profile from pulse to pulse , we can predict that such a dredge - up will also occur , within the next two thermal pulses , for our 5 and @xmath30 stars with @xmath7 and for the @xmath113 star with @xmath1 ( see table [ endagb ] for the number of computed thermal pulses for each object ) . some additional thermal pulses are still needed for lower mass objects . several years ago , the 3dup was found to occur too late along the tp - agb phase for agb stars of decreasing mass and/or increasing metallicity ( see e.g. iben 1976 , fujimoto & sugimoto 1979 ) . this was a common problem ( see lattanzio 1989 for a review ) . recent works showed that with the new opal opacities , the 3dup can now be obtained even for low - mass pop i objects ( see e.g. frost & lattanzio 1996 or straniero et al . 1995 ) . one of the clearest signature of the 3dup is the significant @xmath170 pollution of the convective envelope , brought into the inter - shell region by the convective tongue of a thermal pulse ( see sect . as the thermal pulse intensity ( and consequently the probability of 3dup occurrence ) increases with time , as do the core mass and the total luminosity , the models tend predict the formation of carbon ( c ) stars ( for which the surface @xmath170/@xmath173 becomes greater than unity thanks to the repetitive 3dup events ) at too high luminosities , i.e. to late during the tp - agb phase , especially for solar metallicity stars . solutions to increase the depth of the convective envelope during a dredge - up , e.g. by treating in some way semi - convection ( see e.g. lattanzio 1986 ) or by invoking undershooting ( see alongi et al . 1991 ) have been suggested . they indeed help the 3dup to occur but do not yet solve completely the problem . fundamentally , this could be due to our bad knowledge of convection inside stars . very recently , straniero et al . ( 1996 ) suggested that by considerably increasing the number of mass shells and especially the number of time steps to model thermal pulses , the 3dup occurs earlier during the tp - agb phase . this consists in a real numerical progress in that field . the fact we find 3dup latter than them along the agb phase can be due to our lower number of time steps ( typically by a factor of 3 ) . it was first discovered by paczyski ( 1975 ) that during the asymptotic tp - agb phase , there is a linear relation between the surface luminosity @xmath202 and the core mass @xmath203}$ ] ( see table [ t0 ] ) which is only slightly dependent on the total mass of the agb star . blcker & schnberner ( 1991 ) however found that this kind of relation does not exist for the most massive agb stars for which the bottom of the convective envelope goes down to the hbs so that part of the h - burning occurs inside the convective envelope ( hbb ) . > from our models for the last four thermal pulses of the 3 and @xmath112 with @xmath7 and the @xmath111 with @xmath1 , we have derived a core mass luminosity relation in the form @xmath204 \,,\end{aligned}\ ] ] where @xmath205 , @xmath206 , @xmath207 , @xmath208 , @xmath209 and @xmath210 . it is in good agreement with other @xmath211 relations , for @xmath212 and @xmath213 ( see e.g. lattanzio 1986 or boothroyd & sackmann 1988a ) . the more massive stars experience hbb and do not fit , consequently , to the same relation . in table [ endagb ] , we however give a linear relation of the form @xmath214 for the @xmath113 ( @xmath112 ) agb star with @xmath7 ( 0.005 ) that do not experience a very strong hbb . let us stress that during the tp - agb phase , @xmath202 suffers very important variations during and after each thermal pulse . frequently , the core mass luminosity relation is derived by taking solely account of the maximum surface luminosity preceding each thermal pulse . the probability to observe an agb star just at that time is very weak ( @xmath215 ; see table [ endagb ] ) . we thus preferred to derive our relations with a more realistic surface luminosity that is averaged over each pulse - inter - pulse cycle . on the other hand , it is well known that the rate of mass advance of a burning shell ( here @xmath216 for the hbs ) is related to the rate of nuclear energy production by @xmath217 with @xmath218 being the luminosity produced by the hbs , @xmath219 the mean hydrogen mass fraction over this burning shell and @xmath220 the amount of energy produced by the conversion of 1 g of @xmath184 to @xmath48 through the cno bi - cycle . along the tp - agb , most of the surface luminosity @xmath202 is produced by the hbs . as indicated in table [ t0 ] , the hebs contributes much less than the hbs to the nuclear luminosity ( except during a thermal runaway ) . gravitation also constitutes a relatively small energy source , especially for the less massive agb stars . in conclusion , we can write @xmath221 with @xmath222 depending on @xmath193 . by eliminating @xmath218 in eq . ( 6 ) with the core mass luminosity relation , one finds that the core mass @xmath216 increases exponentially with time during the asymptotic tp - agb phase . the time scale for the core mass growth is @xmath223 ( @xmath224 ) yr for a @xmath111 ( @xmath30 ) agb star , that is , interestingly , comparable to the total duration of the tp - agb phase . table [ mcl ] demonstrates that eq . ( 6 ) predicts rates of hbs advance in mass that are very similar to those coming from our numerical models , except for the @xmath30 ( @xmath113 ) object with @xmath7 ( 0.005 ) , probably due to the strong hbb . lccccccc & & + & & + & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath225&9.94@xmath118&1.13@xmath118&4.56@xmath119&2.32@xmath119&4.39@xmath118&5.53@xmath119&2.13@xmath119 + @xmath226&19&13&13&14&8&9&9 + @xmath122 & 9750&18200&24150&33650 & 9250&19700&27600 + @xmath227&1.662&2.718&3.642&4.575&1.869&2.547&3.124 + @xmath228&0.6052&0.7563&0.8290&0.9022&0.6411&0.8277&0.9101 + @xmath24&see rel.&see rel.&264550&&see rel.&383550& + @xmath229&see rel.&see rel.&0.73729&&see rel.&0.77564& + @xmath230&2.1@xmath115&1.2@xmath115&5.1@xmath116&1.9@xmath116&2.4@xmath115&6.2@xmath231&1.1@xmath116 + @xmath232 & 38000 & 9200 & 3700 & 1600 & 45000 & 7500 & 2500 + @xmath233 & 95 & 24 & 14 & 9 & 104 & 20 & 13 + @xmath234 & 19 & 10 & 7 & 6 & 20 & 9 & 8 + @xmath235 & 0.0115 & 0.0038 & 0.0021 & 0.0011 & 0.0119 & 0.0031 & 0.0015 + @xmath236 & 0.0066 & 0.0019 & 0.0010 & 0.0005 & 0.0069 & 0.0017 & 0.0008 + 3dup ? & no & no & no & no & no & yes & no + lccccccc & & + & & + & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath219 & 0.49&0.54&0.52&0.64&0.54&0.58&0.69 + @xmath237 & 3450&10690&14480&17250&4600&10320&12890 + eq . ( 6 ) prediction & 6.7@xmath238&1.9@xmath127&2.7@xmath127&2.6@xmath127&8.1@xmath238&1.7@xmath127&1.8@xmath127 + @xmath239 & 6.4@xmath238&1.7@xmath127&2.5@xmath127&3.1@xmath127&7.8@xmath238&1.8@xmath127&2.4@xmath127 + one of our ambitions is to extrapolate our predictions coming from the complete evolution models not only for the global surface properties but also for the chemical composition ( 5 ) and ( 6 ) already allow to extrapolate rather accurately the core mass and surface luminosity from our last computed models towards the end of the tp - agb phase . for the @xmath30 ( @xmath113 ) star with @xmath7 ( 0.005 ) , independent fits of @xmath202 and @xmath216 as a function of time have been calculated . we also need fits for the surface radius [ in order to follow the mass loss rate increase through eq . ( 1 ) ] as well as the temperature and density at the base of the convective envelope ( to follow the hbb nucleosynthesis ) . we found , on grounds of the last four computed thermal pulses for each star , very good fitting relations for these quantities , which are presented in table [ fits ] . note that they are linear with time during the last part of the tp - agb phase , contrarily to @xmath216 and @xmath202 that increase exponentially during this phase . the main interest of such an extrapolation procedure is that very detailed stellar evolution computations ( including a very complete description of the nucleosynthesis ) would take prohibitive computation time . however , we had to make two important approximations ( compared to full evolution models ) that must be mentioned . * the successive 3dup events progressively pollute the convective envelope in @xmath48 , @xmath170 and heavier elements to the detriment of @xmath184 . this increases the mean molecular weight ( or @xmath6 ) and the mean opacity in the sub - photospheric region , probably leading to an overestimation of the surface luminosity . * very few computations are existing for the very end of the tp - agb phase ( vassiliadis & wood 1993 ; blcker 1995 ) , that can not maintain to give accurate predictions concerning the evolution of the surface isotopic composition . one of their interests is to indicate that the last few thermal pulses seem to have different properties than the full amplitude ones ( e.g. longer inter - pulse durations ) . on the other hand , it has been noted by boothroyd & sackmann ( 1988a ) that the increasing mass loss rate and @xmath6 possibly prevent the 3dup to occur after these last thermal pulses . as a consequence , assuming the full amplitude regime up to the agb tip probably makes our following predictions quite indefinite concerning the very end of the tp - agb phase . llccccccc & & & + & & & + & & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath240&@xmath24 & 3.17@xmath149&1.16@xmath148&1.76@xmath148&1.95@xmath148&9.03@xmath163&1.33@xmath148&3.48@xmath148 + & @xmath229 & 8.53@xmath241&4.11@xmath242&5.11@xmath242&6.35@xmath242&2.37@xmath242&3.57@xmath242&4.06@xmath242 + @xmath243&@xmath24 & 1.81@xmath244&2.88@xmath241&5.13@xmath242&4.21@xmath242&3.53&2.81@xmath242&1.07@xmath245 + & @xmath229 & 4.05@xmath246&1.14@xmath247&2.05@xmath247&6.34@xmath247&3.16@xmath246&8.82@xmath246&4.50@xmath247 + @xmath248&@xmath24 & 1.16@xmath166&2.51@xmath127&1.56@xmath163&3.32@xmath163&4.71@xmath249&3.86@xmath128&1.24@xmath149 + & @xmath229 & 1.72@xmath148&2.28@xmath152&5.40@xmath152&2.47&9.82@xmath149&1.04@xmath148&3.53@xmath244 + such extrapolation procedures have already been realized by groenewegen & de jong ( 1993 and following papers ) . these synthetic evolution models are very powerful to compare observations with computed models concerning the evolution of the global surface properties of agb stars . however , they contain shortcomings that do not allow to predict in detail the evolution of the surface isotopic composition . mainly , _ ( i ) _ only the most abundant elements are followed ( @xmath48 , @xmath170 , @xmath8 , @xmath171 and @xmath173 ) and _ ( ii ) _ the hbb nucleosynthesis is not followed in detail whereas it becomes a very influent process concerning the convective envelope composition towards the tp - agb tip . recent progresses have been made by busso et al . ( 1995 ) and lambert et al . ( 1995 ) to follow in detail the nucleosynthesis as late as possible along the agb phase . in sect . 8 , we both present the nucleosynthesis encountered in our full evolution models and the extension of the predictions concerning the convective envelope composition towards the end of the agb phase corresponding to the seven computed stars . for this extrapolation , we proceed as follows . 1 . from the last computed inter - pulse phase for each star , we integrate , during the successive inter - pulse phases , the evolution of the nucleosynthesis occuring inside the convective envelope by solving eqs . ( 4 ) , suited for the hbb , for each nuclide while imposing a temperature and density increases at the bottom of the convective envelope given by the fits presented in table [ fits ] . we took the inter - pulse durations of table [ endagb ] ; indeed , they are only very slightly decreasing with the core mass increase along the asymptotic tp - agb phase . 2 . since the 3dup is almost ready to occur in our models ( see sect . 6.2 ) , we mix the envelope with the hbs and inter - shell regions recurrently , every @xmath201 . to quantify the adopted depth of each 3dup , let us remember the definition of the usual parameter + @xmath250 + where @xmath251 represents the amount of mass that is dredged - up ( difference between the normal depth of the convective envelope and its maximum downwards penetration during a dredge - up ) and @xmath252 the increase in mass of the hbs between two successive thermal pulses . we adopted a 3dup depth corresponding to @xmath253 for all the stars . this value has been suggested by groenewegen & de jong ( 1993 ) because it reproduces rather well the carbon star luminosity function of the lmc ( that can not be fitted on grounds of self - consistent evolution models only ; see sect . + during such a dredge - up , we assumed that the material coming from the convective tongue of a thermal pulse has abundance profiles corresponding to end of the last computed thermal pulse . of course , we can not do better without further computing thermal pulses with detailed nucleosynthesis up to the end of the tp - agb phase that is , again , not yet realistic . however , this appears to be rather justified for many important chemical elements as they reach almost constant abundances after the few last computed thermal pulses ( see discussion in sect . 7 ) . taking a constant value for @xmath254 is probably not strictly right . as far as the surface composition is concerned , increasing @xmath254 more efficiently pollutes the convective envelope from pulse to pulse . there is another quantity that greatly influences the evolution of the envelope composition : the mass loss rate . indeed , decreasing the mass loss rate allows to increase the total number of thermal pulses ( and consequently the total number of 3dup episodes ) before the end of the tp - agb phase . we had to increase our @xmath28 values in eq . ( 1 ) for each star , in order to reach the highest mass loss rates ( called super - winds ) observed among very evolved agb stars ( see the discussion in sect . 2.1.5 ; see also zuckerman et al . 1986 ; wannier & sahai 1986 ; knapp et al . 1989 ) : we adopted @xmath255 , 10 , 15 and 20 for the 3 , 4 , 5 and @xmath30 agb models , respectively , whatever @xmath6 . as the exact way following which the mass loss rate increases with time along the asymptotic tp - agb phase is still questionable by more than a factor of two ( see sect . 2.1.5 ) , the resulting uncertainty on the evolution of the surface composition up to the agb tip dominates the effect of @xmath254 ( see our discussion in sect . 8.1 , where our mass loss rates are varied by a factor of four ) . in table [ mlos ] , we present the global mass loss increases we obtained , for the seven tp -agb stars , from the beginning of the tp - agb phase up to its end ( i.e. including our extrapolation computations ) . llccccccc & & & + & & & + & & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath126&_at @xmath164 _ & 2.0@xmath127&1.5@xmath128&2.5@xmath128&3.6@xmath128&3.6@xmath127&1.8@xmath128&2.8@xmath128 + & _ agb tip _ & 7.6@xmath128&4.3@xmath163&9.4@xmath163&1.5@xmath149&5.6@xmath128&4.0@xmath163&1.0@xmath149 + one of the most important constraint for evolutionary models towards the agb tip is the reproduction of the observed correlation between the initial mass of stars ( @xmath256 ) and their final one ( @xmath257 ) , i.e. their core mass ( become a hot white dwarf ) when the convective envelope has been completely removed after the planetary nebula ( pn ) ejection . we present in fig . [ mimf ] the comparison between observations analyzed by weidemann & koester ( 1983 ) and weidemann ( 1987 ) and the final core masses we obtained after our evolutionary plus extrapolations computations . we note that agreement with observations is slightly better than for vassiliadis & wood ( 1993 ) , especially around @xmath258 @xmath111 where they only reproduce the upper limits . on the contrary , the final core mass obtained by boothroyd & sackmann ( 1988b ) is rather low compared to observations . note we have not taken into account of the core mass reduction following each 3dup event . however , as explained below , such a reduction , even if it increases the agreement with the observed relation , is rather negligible compared to other uncertainties ( core mass at the first thermal pulse , thermal pulse number depending on the adopted mass loss rate , ) . we observe that changing the mass loss rate by a factor of four during the asymptotic tp - agb phase ( see also sect . 8.1 concerning surface abundance predictions ) does not significantly change the agreement with observations . we claim that this is due to the fact that while almost 50% of the c o core mass is built during the central he - burning phase , roughly 35 to 45% ( in mass ) are added during the e - agb phase . more precisely , during the whole tp - agb phase , we found that our core masses increased by 14 ( 12 ) , 7 ( 4 ) , 5 ( 3 ) and 4% for the 3 , 4 , 5 , and @xmath30 stars with @xmath7 ( 0.005 ) , respectively . the final core mass is consequently almost decided before the super - wind occurs and this observed relation is not very compulsive for the empirical mass loss rate to adopt during the tp - agb phase . > from the pioneer work of iben ( 1975 ) , one realized that inside tp - agb stars , a very rich nucleosynthesis occurs . during the recurrent third dredge - up events , the synthesized nuclides are mixed inside the convective envelope and then ejected into the interstellar medium ( ism ) through the strong mass loss . in conclusion , tp - agb stars play a crucial role for the chemical evolution of galaxies , especially for some elements like @xmath162 , @xmath2 , @xmath3 , @xmath259 , nuclides heavier than @xmath260 , it is the reason why we will give , in sect . 8.3 , the yields resulting from our intermediate - mass star computations . let us analyze the various nucleosynthesis processes occuring inside intermediate - mass tp - agb stars . in fact , three very different sites have to be distinguished . as indicated in table [ t0 ] , most of the nuclear energy release on the agb is due to the hbs , mainly through the cno bi - cycle . more precisely , the six most important nuclear reactions that energetically supply the star are @xmath261 ( @xmath262 ) , @xmath263 ( @xmath264 ) , @xmath265 ( @xmath266 ) , @xmath40 @xmath267 ( @xmath268 ) , @xmath39 @xmath267 ( @xmath269 ) and @xmath270 ( @xmath271 ) , all occuring inside the hbs . these numbers are quite similar ( within a few % ) for all our modeled stars . whatever the agb star total mass , the hbs slightly depletes @xmath170 and @xmath172 and significantly depletes @xmath174 , @xmath96 and @xmath3 , while @xmath171 is produced . the case of @xmath8 is more difficult : while its abundance is very slightly increased in the hbs of @xmath111 stars ( whatever @xmath6 ) , it is somewhat decreased in the hbs of @xmath30 stars . the rather high temperature inside the hbs also allows the activation of the nena and mgal chains , even if they are much less influent for the global energetics . the most interesting consequences of the nena chain are the production of @xmath188 and @xmath4 in @xmath272 @xmath113 stars and their slight destruction in more massive agb stars . the mgal chain is very sensitive to the agb total mass . in @xmath273 @xmath113 objects , @xmath190 is destroyed to the benefit of @xmath259 ( roughly half in @xmath43 and half in @xmath44 that quite instantaneously decays in @xmath191 ) ; a small quantity of @xmath191 is also destroyed in @xmath192 . in more massive agb stars , the hbs temperature is high enough to substantially produce @xmath190 through the @xmath274 reaction . this in turn leads to a much important @xmath43 production . however , as the hbs mean temperature always increases with time , @xmath43 begins to be partially destroyed by @xmath275 reactions late along the tp - agb . we refer to forestini et al . ( 1991 ) for the first detailed study of @xmath43 production inside agb stars , by radiative proton burning ( i.e. in the hbs ) . in guelin et al . ( 1995 ) , more details can be found , including the possibility to synthesize @xmath43 by proton burning at the bottom of the convective envelope ( i.e. by hbb ) . it was however a parametric study . first predictions concerning the enrichment of the agb star surfaces , on grounds of full evolutionary models are presented in sects . 7.6 and 8.1 . the comparatively small amount of nuclear energy supplied by the hebs comes from the @xmath276 ( @xmath277 ) and @xmath50 ( @xmath278 ) reactions . consequently , the hebs mainly produces @xmath170 and , to a lower extent , @xmath173 , while during the previous central he - burning stage , the @xmath173 production was preponderant . at the beginning of a thermal pulse , it is the @xmath276 reaction that becomes the dominant energy source ( see also sect . 7.1.3 for the thermal pulse energetics ) . during the long inter - pulse phase that follows a 3dup episode , the @xmath8 coming out of the hbs can be burnt in the inter - shell region , close to the hebs top , mainly through the @xmath279 reaction . this reaction is highly energetic . the first published work reporting the existence of this inter - shell @xmath8 burning is straniero et al . the produced neutrons are then captured by light species up to @xmath260 mainly , but part on them are also used to synthesize @xmath280 elements . by decreasing order of amplitude , the nuclear energy liberated in the inter - shell region during the inter - pulse phase comes from these @xmath95 reactions up to @xmath260 ( @xmath281 ) , @xmath282 ( @xmath283 ) , the so - called @xmath284 process reactions ( @xmath285 ) , @xmath286 ( @xmath269 ) and @xmath287 ( @xmath288 ) . note the important nuclear energy released by the neutron captures . this energy production creates a third burning shell between the hebs and hbs , even if its nuclear activity is considerably lower than that of both the hebs and hbs . it continues to exist until @xmath8 has been destroyed there . this happens , in general , almost when the following thermal pulse begins . as a consequence , when the convective tongue of a thermal pulse engulfs the inter - shell matter , the amount of @xmath8 that is mixed is reduced compared to what was previously thought , i.e. @xmath289 x$](@xmath8 ) @xmath290 $ ] , where @xmath290 $ ] designates , as before , the inter - shell region . we show , in fig . [ 13csh ] , the nuclear energy production rate ( @xmath291 ) profile in the case of our @xmath111 agb star . this clearly demonstrates the three - burning shell structure . such a structure is also found in the more massive agb stars . let us emphasize that this third burning shell could at least partially explain the production of the @xmath280 nuclides ( see sect . 7.5 ) in a radiative zone . this has been suggested by straniero et al . ( 1995 ) , even if their amount of @xmath8 in the inter - shell was artificially increased in order to study the @xmath284process nucleosynthesis . if the temperature @xmath292 at the base of the deep convective envelope is high enough ( i.e. @xmath293 k ) , i.e. if the base of the convective envelope reaches the hbs , proton burning occurs inside it . this hbb occurs in models with initial masses @xmath294 @xmath113 for @xmath7 and @xmath294 @xmath112 for @xmath1 . these lower mass limits and this @xmath6 dependence are common features of all the agb models ( see e.g. boothroyd et al . 1993 ) . as mixing up to the cool surface is very efficient ( the mean turn - over time @xmath105 of the convective envelope is roughly 0.5 yr ) , the effective nuclear reaction rates are lower than in radiative h - burning , leading to specific nucleosynthesis signatures . for @xmath295 k already , @xmath2 can be produced from @xmath162 following a scenario that can not operate in radiative h - burning ( see sect . 7.2 for more details ) . this li synthesis has been qualitatively suggested by cameron & fowler ( 1971 ) and quantitatively first demonstrated on grounds of evolutionary models by sackmann & boothroyd ( 1992 ) . for @xmath296 k , the cn cycle begins to operate partially , i.e. @xmath170 is converted to @xmath8 that accumulates . as first investigated by boothroyd et al . ( 1993 ) , this conversion can efficiently reduce the @xmath170/@xmath173 ratio and eventually , if hbb is strong enough , this can prevent the formation of c stars ( see also sects . 8.1 and 8.2 ) . if @xmath292 @xmath297 @xmath298 k typically , the cn cycle almost operates at equilibrium through the whole envelope , leading to a @xmath170/@xmath8 isotopic ratio lower than 10 ( depending on @xmath292 ) . in our models , this happens for the @xmath30 ( @xmath113 ) agb star with @xmath7 ( 0.005 ) and also , later along the tp - agb , for our @xmath113 star with @xmath7 ( see sect . 8.1 ) . at very high @xmath292 , other reactions begin to occur ; in particular , in extreme cases ( i.e. the end of the tp - agb phase of our @xmath30 models with @xmath7 ) , @xmath43 can eventually be directly produced inside the convective envelope through the @xmath83 reaction on @xmath190 supplied by @xmath274 ( see sect . 8.1 ) . the thermal runaway , inside the hebs , leading to form a convective tongue is principally due to the @xmath276 reaction excitement . when this convective tongue finally penetrates the inter - shell region , the hbs ashes are ingested , i.e. not only @xmath48 but also @xmath8 , @xmath171 and @xmath43 mainly . as these elements are partially burnt in the bottom of the inter - shell region during the inter - pulse phase ( globally , we find that half of the @xmath8 is so destroyed ) , they mainly survive in the upper part of the inter - shell . they are consequently engulfed by the convective tongue when it reaches its maximum extension . at that time , the @xmath276 reaction is already dominant for the energetics . we consequently find that the energetics associated with the @xmath8 burning through the @xmath282 reaction as well as the neutron captures that follow is rather negligible ( less than 1 % ) . this is also found by straniero et al . ( 1995 ) . this conclusion could however be modified if more @xmath8 was synthesized in the inter - shell region , as required for the @xmath284process ( see sect . the important energetic effects reported by bazan & lattanzio ( 1993 ) concerning the @xmath8 burning inside thermal pulses was probably due to _ ( i ) _ the fact they did not give account of partial inter - shell @xmath8 burning and _ ( ii ) _ the very high @xmath8 amount artificially ingested into their thermal pulses in order to study the consequent @xmath284process . part of produced neutrons participate to @xmath299 reactions , mainly on @xmath43 and , to a lower extent , on @xmath171 . the resulting protons , at these high temperatures characterizing he - burning during a thermal pulse ( @xmath300 k at the base of the convective tongue ) , are rapidly captured by various elements ( see sects . 7.3 to 7.5 for more details ) . let us stress that the simultaneous presence , inside a mixed he - burning region ( the convective tongue ) , of neutrons and protons leads to a very specific ( and unique ) nucleosynthesis . we will analyze in more details the nucleosynthesis processes associated with the thermal pulses in sects . 7.3 to 7.5 . sackmann & boothroyd ( 1992 ) were the first ones to quantitatively demonstrate how the cameron & fowler ( 1971 ) scenario leads to @xmath2 synthesis in intermediate - mass agb stars . this major result was obtained by using a time - dependent convective diffusion algorithm suited for hot - bottom convective envelopes . indeed , at the base of such envelopes , some nuclear reactions involved in @xmath2 production occur faster than mean turn - over time associated with the convective motions ( see fig . [ reacthp ] ) . we also use a time - dependent treatment to study the @xmath2 synthesis [ see eqs . ( 4 ) in sect . we show in fig . [ libe ] the typical resulting @xmath35 and @xmath2 profiles inside the convective envelope of our @xmath30 agb models with @xmath7 . when hbb occurs ( see sect . 7.1.2 ) , @xmath162 first begins to burn through the @xmath301 and @xmath302 reactions , the second one being significantly slower than the first one . as in deep stellar interiors , the @xmath35 destruction by electron capture and proton capture are much longer than the time scale associated with the convective mixing , the produced @xmath35 is efficiently up - heaved towards cooler regions where it progressively decays in @xmath2 . let us emphasize that without fast convective mixing , @xmath2 could not accumulate at the surface . indeed , at the base of the convective envelope , _ ( i ) _ @xmath35 would be partially destroyed by electron and proton captures and _ ( ii ) _ the @xmath2 resulting from @xmath35 decay would be quite instantaneously destroyed by proton capture too . let us now compare our predictions with sackmann & boothroyd ( 1992 ) results and with observations . a direct comparison with their figs . 2a and 2b reveals nearly identical features . in fig . [ lli ] , we present the evolution of the surface @xmath2 abundance during the tp - agb phase of all our modeled stars . in our @xmath111 models , the temperature at the base of the convective envelope is never high enough for @xmath35 to be produced efficiently , and the surface @xmath2 abundance remains almost constant at a very low value resulting from the first dredge - up . we also see that our @xmath112 star with @xmath1 begins to produce some @xmath2 when it enters the asymptotic regime , but the surface abundance of this light element always stays at least one order of magnitude below the interstellar medium ( or cosmic ) value @xmath303(@xmath2 ) @xmath304 . note the @xmath2 destruction inside the @xmath112 star with @xmath7 attesting that its convective bottom temperature is high enough to burn @xmath2 through @xmath305 but not enough to produce @xmath35 by @xmath162 burning . finally , our 5 and @xmath30 models present real high @xmath2 surface enhancements , compared to the cosmic value . the maximum abundances we obtain range between @xmath303(@xmath2 ) @xmath306 to 5.1 , depending on the stellar mass and metallicity . our super - lithium - rich stars are predicted to appear in a luminosity range between @xmath307 and @xmath308 , i.e. for @xmath309 between -6.25 and -6.65 . these values are in good agreement with those found by sackmann & boothroyd ( 1992 ) . as the evolution proceeds , the surface @xmath2 abundances finally decrease again very rapidly . this indicates that @xmath162 has been almost completely burned in the envelope [ remember that the @xmath301 is dominant ] , so that the @xmath2 production can no more be supported and it is then destroyed by proton capture . this fact clearly puts an upper limit in luminosity to observe super - lithium - rich agb stars . this last feature is also present in sackmann & boothroyd ( 1992 ) , even if it appears somewhat later during the tp - agb phase of their most massive objects . this could be due to different mass loss rate prescriptions . how do these predictions compare with observational data ( see fig . [ lli ] ) ? in the magellanic clouds , the large majority of lithium - rich giants are observed within a narrow luminosity range , @xmath310 to @xmath311 ( smith & lambert 1989 , 1990b ; plez et al . 1993 ; smith et al . 1995 ) , in very good agreement with the predictions . the observed abundance range is also well reproduced , even if no star has been observed with @xmath303(@xmath2 ) higher than 4.5 in the clouds . moreover , the mass estimate from pulsation theory for some lithium - rich giants of both clouds is consistent with the predicted mass range for surface @xmath2 enhancement ( smith et al . 1995 ) . comparisons with @xmath2 in galactic giants are more delicate because of the larger observational uncertainties on luminosity and mass . galactic super - lithium - rich agb stars can exhibit @xmath303(@xmath2 ) values as high as 5.4 ( abia et al . 1991 ) , but these carbon stars have @xmath309 laying between -5 and -6.2 , i.e. rather lower than the predicted luminosities . this could be due to a bad luminosity determination , as it seems difficult to synthesize @xmath2 inside the convective envelope of agb stars with such relatively low luminosities . however , most of the galactic super - lithium - rich agb stars have lower masses than the ones we model here , and they will be addressed in a forthcoming paper . we now review all the species ( from c ) by indicating in which nuclear reactions they are implicated inside the convective tongue of the successive thermal pulses , as a function of the initial mass @xmath193 and metallicity @xmath6 of the computed tp - agb models . for the discussion , we refer to figs . 7 to 9 of the electronic version of the paper . the mean mass fractions @xmath312 are calculated by taking the average of the mass fraction profile of each nuclide @xmath313 over the maximum region covered by the convective tongue of each thermal pulse , just after its disappearance , i.e. @xmath314 we also refer to fig . [ xenv302 ] that presents the mean mass fractions of key nuclides ( calculated as for the convective tongue ) in each burning region ( i.e. the hbs and inter - shell regions ) compared to their convective envelope content , for the @xmath111 star with @xmath7 . similar figures for the other six agb stars are only available in the electronic version . similarly , fig . [ xthp302 ] presents the abundance variations of the same nuclides inside typical convective thermal pulses . again , similar figures for the other six agb stars can solely be found in the electronic form . all these figures are supporting the following nucleosynthesis descriptions . during the growth of the convective tongue , the mean @xmath170 mass fraction first decreases due to the @xmath48 mixing from the inter - shell region . then , at the thermal pulse maximum , @xmath170 is produced by the @xmath276 reaction and also enhanced due to the downwards penetration of the convective tongue into the he - burnt shell . although this penetration is commonly found , its amplitude seems somewhat more important in our case , bringing more @xmath170 into the thermal pulses . note this increased amount of @xmath170 makes the 3dup episodes more efficient to convert an agb star into a c star . during the first thermal pulses , the amount of @xmath170 that is produced increases due to the higher temperatures reached at the base of the convective tongue . this bottom temperature increase from pulse to pulse is a common feature of all the stars and it traduces that the thermal pulse intensity increases . the last computed thermal pulses of the @xmath111 object with @xmath7 are even hot enough to partially destroy @xmath170 by the @xmath50 reaction . most of the engulfed @xmath8 is destroyed very efficiently through @xmath282 that produces neutrons . the trend of the remaining mean @xmath8 mass fraction to increase from pulse to pulse is due to the fact that stronger thermal pulses develop convective tongues reaching regions always closer to the hbs , i.e. engulfing more material from the inter - shell zone . as mentioned in sect . 6.1 , the bottom of the hbs can not be reached however . @xmath38 is only produced by the capture of part of the neutrons on the engulfed @xmath171 , through @xmath59 . note that this reaction contributes to the production of protons . for initial masses above @xmath112 ( @xmath111 ) with @xmath7 ( 0.005 ) however , the convective tongue finally becomes hot enough ( base temperature @xmath315 k ) to significantly destroy the produced @xmath38 by the @xmath54 reaction . the relatively low @xmath38 production inside the @xmath111 with @xmath7 object is due to the relatively lower amount of @xmath8 that is ingested ( partially due to the third @xmath8 burning shell ; see sect . 7.1.1 ) , leading to a smaller quantity of neutrons . however , the last thermal pulses of this star ingest increasing amounts of @xmath8 and @xmath171 from the inter - shell , leading to an increasing production of @xmath38 . due to its short lifetime ( @xmath316 yr ) , @xmath38 partially decays during the inter - pulse phases of our less massive agb models . this depletion is however reduced by the @xmath38 production inside the radiative @xmath8 burning shell . for the most massive agb stars that have inter - pulse durations shorter that @xmath317 , @xmath38 mostly behaves like a stable nuclide , even if strongly destroyed by @xmath24 captures . consequently , as explained with more details in forestini et al . ( 1996 ) , intermediate - mass agb stars are globally not important producers of @xmath38 . apart from the first two thermal pulses ( whatever the initial mass or metallicity ) , @xmath171 is almost completely destroyed inside thermal pulses through @xmath318 . this reaction , as first noted by iben ( 1976 ) , also significantly contributes to the energetics of a thermal pulse . only a small part of @xmath171 is destroyed through the @xmath287 reaction . indeed , the inter - shell mean mass fraction of @xmath171 is ranged between @xmath319 for the @xmath111 star with @xmath7 and @xmath320 for the @xmath113 star with @xmath321 , while the production level of @xmath38 is @xmath322 ( @xmath323 ) , in mass fraction , for the @xmath7 ( 0.005 ) models ( except for the @xmath111 with @xmath7 star ; see above ) . as are @xmath8 and @xmath171 , @xmath172 is engulfed in a thermal pulse from the inter - shell region . the only significant reaction that produces this element inside the convective tongue itself is @xmath324 ( see below for the @xmath96 production ) . on the other hand , @xmath172 is mainly destroyed by the @xmath60 reaction . its rate is lower than that of @xmath318 so that @xmath172 is efficiently destroyed in all the stars only in the asymptotic regime , i.e. when thermal pulses become very intense . remember that it arrives earlier for lower @xmath6 objects . note that the @xmath111 stars ( whatever @xmath6 ) maintain a relatively higher @xmath172 abundance , especially during the full amplitude thermal pulses . indeed , while these thermal pulses efficiently destroy @xmath172 whatever the initial mass of the star , its initial abundance inside the convective tongue is higher for @xmath111 objects . this is due to the fact that the engulfed inter - shell region of @xmath111 stars , that contains @xmath172 , is significantly thicker ( in mass ) compared to that of more massive ones . @xmath173 being partially destroyed in the hbs , the mixing of the inter - shell region reduces its mean mass fraction . on the other hand , at the maximum extent of the thermal pulses , it is enhanced , like @xmath170 , due to the downwards penetration of the convective tongue . from pulse to pulse , it also becomes significantly produced by @xmath50 . this is especially visible for @xmath1 stars , the initial @xmath173 content of which is lower . this primary @xmath173 production by the thermal pulses finally stabilizes around @xmath319 ( in mass fraction ) . the decrease observed during the last thermal pulses of the @xmath111 agb star with @xmath7 traduces the more important role of mixing of inter - shell material ( already emphasized for @xmath171 and @xmath172 ) . the @xmath174 previously contained in the inter - shell region is also mixed in the convective tongue . it is not produced inside thermal pulses . on the contrary , it can be destroyed mainly through @xmath325 and @xmath326 , the first reaction being in general at least ten times faster than the second one . this weakly contributes to produce neutrons inside the thermal pulses . this @xmath174 destruction already operates from the first thermal pulses , whatever the initial @xmath193 and @xmath6 . in later thermal pulses , the @xmath174 mass fraction stabilizes around 4 to @xmath327 , that represents a balance between the amount of @xmath174 brought from the inter - shell region and destroyed by the thermal pulse . again , our @xmath111 models with @xmath7 indicate a still dominant role of the inter - shell mixing . it is less true for the @xmath1 models because at lower @xmath6 , thermal pulses are somewhat hotter , that favors @xmath174 depletion . contrarily to the case of @xmath174 , the inter - shell mass fraction of @xmath96 is very low ( @xmath328 typically ) . its presence inside thermal pulse consequently results from its production by the @xmath329 reaction . note that the rapid @xmath41 @xmath330decay to @xmath96 also significantly contributes to the thermal pulse energetics . this element is destroyed by three reactions that are , by decreasing order of nuclear time scales , @xmath331 , @xmath332 and @xmath333 . this last reaction can even become faster than @xmath332 when significant amounts of protons are produced . this especially arrives for the late thermal pulses of the @xmath111 star with @xmath7 . the convective tongues being hotter for more massive agb stars and/or later along the tp - agb phase of a given star , the @xmath96 destruction , which is very sensitive to temperature , becomes more and more efficient from pulse to pulse . this point was already noted by boothroyd & sackmann ( 1988b ) . the thermal pulse equilibrium mean mass fraction between production and destruction is ranged between @xmath334 and @xmath335 depending on @xmath193 and @xmath6 . as for @xmath96 , @xmath3 is destroyed in the hbs . it is however produced inside thermal pulses . as first demonstrated by forestini et al . ( 1992 ) , the @xmath3 synthesis involves many nuclear reactions , so that it constitutes a good tracer of the nucleosynthesis conditions prevailing inside thermal pulses . @xmath3 is produced by the @xmath60 reaction . the rather rapid increase of the mean @xmath3 mass fraction during the first thermal pulses of each agb star is due to the increasing importance of the inter - shell @xmath172 destruction . when convective tongues become hot enough to completely destroy the engulfed @xmath172 , the following thermal pulses only ingest @xmath172 from matter that has not experienced the preceding thermal pulse ( i.e. @xmath336 of the convective tongue extension in mass typically ; see table [ endagb ] ) . however , as we have explained above , @xmath172 can be produced by the chain @xmath329@xmath337@xmath333 . neutrons are mainly coming from the @xmath282 reaction , and to a lower extent , from @xmath332 , @xmath76 , @xmath338 ( see below ) and @xmath325 . protons are coming from @xmath299 reactions , mostly on @xmath43 and @xmath171 ( see below ) . however , _ ( i ) _ only part of the neutrons are available to produce protons as many @xmath95 reactions also occur ( see sect . 7.5 ) and _ ( ii ) _ a small ( @xmath339 ) part of the protons are also captured by other nuclides than @xmath96 [ mainly by @xmath340 . the success of this reaction chain mainly depends on the rate of @xmath96 destruction , mainly by @xmath24 captures ( see above ) . this destruction becomes more efficient from pulse to pulse so that only a small fraction of @xmath96 is available for the @xmath341 reaction . on the other hand , when convective tongues become very hot ( base temperature @xmath342 k ) , @xmath3 begins to be destroyed by @xmath343 and @xmath66 ( if significant amounts of neutrons are available ) . in conclusion , three facts concur to considerably decrease the @xmath3 production after the first thermal pulses ( whatever @xmath193 and @xmath6 ) . indeed , from pulse to pulse and especially for the most massive agb stars , * the amount of @xmath8 engulfed in the convective tongue of a thermal pulse decreases ( due to the partial destruction of @xmath8 in the inter - shell region during the inter - pulse phase ; see sect . 7.1.1 ) ; * the amount of @xmath172 ingested from the inter - shell region also decreases ; * the @xmath24 captures on @xmath96 are more frequent ; * the @xmath3 destruction becomes significant . this can in particular lead to a partial destruction of the @xmath3 synthesized in the previous thermal pulses for stars with @xmath344 @xmath112 . the mean @xmath186 abundance is roughly unchanged in the hbs as well as by the thermal pulse nucleosynthesis . in number , the contribution of @xmath345 is rather negligible . @xmath187 is significantly destroyed in the hbs . inside the convective tongues of quite hot thermal pulses ( base temperatures above @xmath346 k ) , it is produced by the small proportion of @xmath96 that is destroyed through the @xmath332 reaction . there is also a small contribution by @xmath326 . only a very small @xmath187 fraction can be destroyed by the @xmath338 at the end of very hot thermal pulses ( base temperatures above @xmath347 k ) . @xmath188 is somewhat produced in the hbs . its abundance is considerably enhanced inside thermal pulses ( by typically a factor of 100 ) due to its production by @xmath331 . however , its abundance from pulse to pulse becomes constant ( or even slightly decreases inside full amplitudes thermal pulses of the most massive stars and/or those with lower @xmath6 ) . at their maximum extent , such convective tongues are effectively hot enough ( base temperatures @xmath348 k ) to activate the @xmath349 and @xmath350 reactions . however , as noted below , only a small part of the @xmath188 nuclides are converted into @xmath191 or @xmath190 , respectively ( typically one hundredth ) . nevertheless , this consists in a non - negligible neutron source for these thermal pulses , as first noted by kppeler et al . ( 1990 ) and emphasized by straniero et al . ( 1995 ) . concerning @xmath4 , one has to distinguish between the 3 , 4 , @xmath113 with @xmath7 and @xmath111 with @xmath1 stars on the one hand , and the @xmath30 with @xmath7 , 4 and @xmath113 with @xmath1 stars on the other . inside the former ones , @xmath4 is significantly produced at the bottom of the hbs and very slightly produced by the hebs [ during the inter - pulse phase , through the @xmath351 reaction ] . furthermore , whatever @xmath193 and @xmath6 , @xmath4 is almost unchanged by the thermal pulse nucleosynthesis . as a consequence , its mean mass fraction inside the convective tongue does not evolve significantly from pulse to pulse . at the opposite , inside the latter ones , the bottom of the hbs is hot enough to somewhat destroy @xmath4 while it is produced in its upper part . however , the hebs substantially produces this nuclide during the inter - pulse phase . consequently , when the convective tongues of such stars penetrate the inter - shell region , they engulf material that is impoverished in @xmath4 ; this explains its mean mass fraction decrease from pulse to pulse . the @xmath189 abundance is almost unchanged by the thermal pulse nucleosynthesis too . furthermore , the hbs of @xmath111 stars is not hot enough to destroy it significantly . its mean mass fraction in the convective tongue region of such stars consequently remains almost constant from pulse to pulse . the situation is quite different for more massive stars ( whatever @xmath6 ) as they destroy @xmath189 at the bottom of their hbs . dilution of inter - shell material by the successive convective tongues thus traduces in a mean mass fraction decrease with time . intermediate nuclide of the mgal chain , @xmath190 is modified by the proton burning inside the hbs . we refer to sect . 7.1.1 for more details ( dependence on @xmath193 and @xmath6 ) . however , the evolution of its mean mass fraction inside the convective tongues of thermal pulses is dominated by its production through the @xmath76 reaction [ operating at a quite comparable rate than @xmath75 , following our present knowledge of the corresponding nuclear reaction rates ] . indeed , even if this reaction is very slow inside thermal pulses of relatively low mass agb stars ( especially for their first thermal pulses ) , @xmath188 is much more abundant than @xmath190 ( by typically a factor of one hundred ) in that region , that explains the sensible increase of the @xmath190 mean mass fraction . @xmath191 is slightly ( significantly ) depleted in the hbs of the @xmath195 ( 0.005 ) models . on the other hand , it is substantially produced by the thermal pulse nucleosynthesis _ ( i ) _ by the @xmath75 reaction ( same remark as for @xmath190 above ) and , to a lower extent , _ ( ii ) _ by the @xmath352 reaction ( see below ) . the slope with which its mean mass fraction increases form pulse to pulse is consequently higher for hotter convective tongues , i.e. steeper for full amplitude thermal pulses of massive agb stars . @xmath43 is one of the most important by - product of the hbs ( see above ) . with a lifetime @xmath353 yr , i.e. much longer than the inter - pulse duration of intermediate - mass agb stars , it essentially behaves like a stable nuclide in the inter - shell region . when engulfed inside the convective tongue of a thermal pulse , it is mainly destroyed through the @xmath352 reaction and , at least ten times slower , by @xmath354 . as a result , @xmath43 is the most important proton source inside thermal pulses , whatever @xmath193 and @xmath6 . its destruction is however partial , due to the relatively low neutron abundance ( see sect . 7.5 below ) . note that @xmath352 is also the principal neutron source in the radiative @xmath8 burning shell , as reported by wasserburg et al . ( 1994 ) too . @xmath192 is significantly produced in the hbs , especially in the most massive agb stars . it does not significantly participate to the thermal pulse nucleosynthesis . as a consequence , it is engulfed by the convective tongue from the inter - shell region and its corresponding mean mass fraction slightly increases from pulse to pulse . note however that with time , most of the h - burning of the @xmath113 agb star with @xmath1 occurs inside its convective envelope . this considerably reduces the @xmath192 production . si , p and s are almost not concerned by charged particle nuclear reactions inside the convective tongue of thermal pulses . on the other hand , they are also rather unchanged in the hbs . the only role played by these elements concerns neutron captures that we now briefly discuss . many evolved agb stars show considerable overabundances in elements heavier than iron ( see e.g. smith & lambert 1986 ) . the original discovery by merrill ( 1952 ) of tc in the spectra of some s stars definitively proved that these heavy elements are synthesized inside agb stars . iben ( 1975 ) first demonstrated how such species can indeed be produced inside the thermal pulses by slow neutron captures ( the so - called @xmath284process ) , those neutrons coming from the @xmath76 reaction . however , iben s computations concerned a very massive agb star ( @xmath0 ) , while ever since , all the observed stars showing @xmath280 elements enrichment have been identified as being of lower mass ( i.e. @xmath273 @xmath30 and for many of them @xmath273 @xmath111 ) . more recent observations ( aaronson & mould 1985 ) and evolutionary models ( malaney & boothroyd 1987 ) then strongly suggested that the major neutron source had to be the @xmath282 reaction . @xmath8 is engulfed by the convective tongues of thermal pulses from the inter - shell region where the hbs ashes accumulate . however , the amount of @xmath8 spread out by the hbs is by far ( i.e. at least a factor of 10 or more ) insufficient to explain the @xmath284process that requires more than one neutron by seed @xmath260 . this so - called `` @xmath284process mystery '' ( sackmann & boothroyd 1991 ) is a common failure of all the recent stellar evolution models of tp - agb stars . it is moreover reinforced by the fact that other lighter nuclides also capture part of the available neutrons ( mainly , by increasing order of atomic mass , @xmath170 , @xmath171 , @xmath174 , @xmath3 , @xmath189 , @xmath190 , @xmath43 , @xmath9 , @xmath355 , @xmath356 , @xmath357 , @xmath358 , @xmath34 , ) . one has to note that while the neutron abundance has to be increased to solve the @xmath284process problem , the present computations probably slightly overestimate the abundances of the above mentioned nuclides . more specifically , in our thermal pulse models , the ratio of the engulfed mass of @xmath8 to the mass of @xmath260 inside the convective tongue at its maximum extent ranges between 0.002 to 0.03 . this upper limit is reached during the asymptotic regime of all our modeled stars and thus appears to be rather independent of the initial mass or metallicity . inside their first thermal pulses however , @xmath111 stars already produce neutrons ( ratio roughly equal to 0.02 ) while at the opposite , the corresponding thermal pulses of @xmath30 stars do not engulf substantial amounts of @xmath8 . furthermore , we noted ( see sect . 7.4 ) that the @xmath188 neutron source operates after the maximum of very intense thermal pulses , at a rate that is growing with the maximum temperature of the convective tongue , i.e. with initial mass and/or pulse number . let us stress that such low quantities of @xmath8 ingested by thermal pulses are partly explained by the partial @xmath8 radiative burning , operating at the bottom of the inter - shell region during the inter - pulse phase ( see sect . consequently , it seems clear that the @xmath284process occuring inside tp - agb stars has two distinct origins ( in time and space ) : * during the inter - pulse phase , _ radiative _ @xmath284process occurs in a radiative zone that is the lower part of the inter - shell region ; the only neutron source is @xmath8 ; * during and inside the convective tongue associated with thermal pulses , the _ classical _ @xmath284process takes place with two possible neutron sources : _ ( i ) _ the remainder amount of @xmath8 that is engulfed by the convective tongue and , to a lower extent ( maximum a few percents ) , _ ( ii ) _ the @xmath188 , close to the end of hot thermal pulses . let us emphasize that the @xmath359 and accompanying @xmath95 reactions have non - negligible energetic effects on the structural evolution of the convective tongues . we compute again some thermal pulses by treating together nucleosynthesis and time - dependent convective mixing through eqs . we observed an abundance gradient inside the convective tongue for a few nuclides that are involved in very fast nuclear reactions ( neutrons , protons and some unstable nuclides like @xmath41 or @xmath42 ) . globally however , the final abundances resulting from the thermal pulse nucleosynthesis were similar to within @xmath360 compared to standard computations . more specifically , neutrons were systematically found to be much more abundant ( by typically five orders of magnitude ) at the base of the convective tongues , meaning that the ingested @xmath8 is first transported down before it is destroyed by @xmath282 . the liberated neutrons are very rapidly captured , as attested e.g. by a greater proton abundance at the base of the convective tongue too . this leads us to the conclusion that matter that is actually irradiated by the neutron flux is only the bottom part of each convective tongue . the number of free neutrons by seed @xmath260 is so enhanced . such an approach has already been suggested by malaney et al . the conclusion , however , remains that in order to reproduce the observed distribution of @xmath280 elements in the primitive solar system , higher neutron fluxes are still needed . the most natural way to conciliate observations and theoretical models concerning the production of the @xmath280 elements should be to increase the amount of @xmath8 in the inter - shell region . this can be done if protons are transported down , e.g. during the 3dup . @xmath8 could then be produced by the @xmath270 reaction . such a scenario would enhance the @xmath284process in both sites where it can occur . in our opinion , the present failure to build this extra - amount of @xmath8 could be related to our bad treatment of convective boundaries , especially when they rapidly penetrates very inhomogeneous regions , like it is the case during the 3dup events . the possible occurrence of such a slow - particle transport process able to transport protons down into the inter - shell region and its exact efficiency will be quantitatively investigated in the context of low - mass agb stars . these stars have indeed longer inter - pulse and 3dup durations that could allow to transport significant amounts of hydrogen downwards ( see sect . 9 ) . straniero et al . ( 1995 ) have recently shown that such enhanced amounts of @xmath8 in the inter - shell region could indeed allow the radiative @xmath284process to occur . most interestingly , they found that the resulting signatures are rather similar to those coming from classical @xmath284process computations and are even reached faster . this work is very important as it clearly demonstrates that this so - called `` @xmath8 pocket '' allows to explain the observed @xmath280 element synthesis in agb stars . in order to summarize our detailed discussion of nucleosynthesis in the previous sections , we now identify which specific nuclear region , inside a tp - agb star , mainly contributes to the surface abundance change of each nuclide , when the third dredge - up occurs . concerning the light elements , let us just recall ( see sect . 7.2 ) that @xmath2 can be produced by hbb in the convective envelope of @xmath297 @xmath361 ( @xmath112 ) agb stars with @xmath7 ( 0.005 ) . when a third dredge - up occurs , the convective envelope first penetrates the inter - shell region that is depleted in @xmath170 and then reaches a deeper region previously enriched in @xmath170 by the convective tongue . as a result , @xmath170 globally increases in the convective envelope . the situation is reversed for @xmath8 . after a 3dup , the envelope is somewhat impoverished in @xmath8 as this nuclide is almost completely destroyed inside the convective tongue . consequently , the isotopic ratio @xmath170/@xmath8 significantly increases at each 3dup , all the more as the convective envelope mass is reduced . however , in our @xmath294 @xmath113 ( @xmath112 ) tp - agb stars with @xmath7 ( 0.005 ) , hbb is strong enough to partially convert @xmath170 in @xmath8 through the cn cycle so that the surface @xmath170/@xmath8 drastically decreases . for these objects , the 3dup and hbb have opposite effects . last but not least , @xmath38 , somewhat produced by thermal pulses and the third burning shell , is up - heaved to the surface during 3dup events , where it however substantially decays in the less massive agb stars . @xmath171 is produced in the inter - shell region but is strongly destroyed inside thermal pulses . @xmath172 is destroyed in the inter - shell region while it is somewhat produced by the first few thermal pulses and then destroyed inside full amplitude thermal pulses , i.e. those that can be followed by a 3dup . as a consequence , the surface isotopic ratio @xmath171/@xmath172 is expected to slightly increase from 3dup to 3dup . again , as the cn cycle operates inside the convective envelope of the most massive agb stars , @xmath171 is produced from @xmath170 and @xmath172 is destroyed , thus leading to very high @xmath171/@xmath172 ratios . if hbb occurs at very high temperatures ( i.e. @xmath362 k ) a rather long time enough ( i.e. for the lowest mass loss rate ) , the approach towards the cno bi - cycle equilibrium leads to an increase of the @xmath172 abundance and consequently , a decreasing @xmath171/@xmath172 ratio . this mainly concerns our @xmath30 ( @xmath113 ) agb star with @xmath7 ( 0.005 ) . @xmath173 , like @xmath170 , is depleted in the inter - shell region , but is produced inside the thermal pulses . during 3dup events , its surface abundance very slightly increases . @xmath174 is always considerably depleted in both the inter - shell and thermal pulse regions , so that a 3dup decreases its surface abundance . finally , at the end of full amplitude thermal pulses , @xmath96 has a lower abundance as inside the convective envelope . as it is also completely destroyed in the inter - shell region , its abundance drastically decreases from 3dup to 3dup . consequently , both @xmath173/@xmath174 and @xmath173/@xmath96 surface isotopic ratios significantly increase during each dredge - up . if hbb is strong enough however , the operation of the on cycle increases the amount of @xmath174 to the expense of @xmath173 and substantially destroys @xmath96 . so , in massive agb stars , @xmath173/@xmath174 can decrease while @xmath173/@xmath96 further increases much more than in stars that do not experience hbb . @xmath3 is destroyed in the hbs but it is significantly produced by thermal pulses . as a consequence , 3dup events must lead to correlated @xmath3 and @xmath170 surface enhancements , since both elements are produced in the same region . as already explained , @xmath186 is almost unchanged in the hbs and thermal pulse nucleosynthesis . @xmath187 is destroyed inside the inter - shell region but produced inside thermal pulses , this production slightly increasing from pulse to pulse , whatever @xmath193 or @xmath6 . third dredges - up consequently slightly decreases the @xmath186/@xmath187 isotopic ratio . in our @xmath30 ( @xmath113 ) agb models with @xmath7 ( 0.005 ) however , @xmath187 is partially destroyed by hbb . on the other hand , the @xmath186/@xmath188 ratio substantially decreases from 3dup to 3dup due to large @xmath188 production in the hbs and especially inside the thermal pulses . during 3dup events , the surface abundance of @xmath4 slightly increases for agb stars with initial masses up to @xmath112 ( @xmath111 ) and @xmath7 ( 0.005 ) , due to its production in the hbs . the hotter hbs of more massive agb stars do not produce @xmath4 , globally . our @xmath30 ( @xmath113 ) models for the same respective @xmath6 even slightly destroy @xmath4 by hbb . @xmath189 is somewhat destroyed in the hbs of our most massive agb stars , especially those with @xmath1 , while its abundance does not change significantly after the thermal pulse nucleosynthesis . @xmath190 is depleted in the inter - shell region of our @xmath111 models with @xmath195 as well as in the @xmath1 models of all masses . it is however produced in the hbs of @xmath297 @xmath112 agb stars with @xmath7 . nevertheless , its surface abundance evolution following 3dup events is mainly conditioned by its large abundance ( compared to the envelope one ) in the region where thermal pulses occur , especially inside the most massive agb stars . therefore , the surface @xmath189/@xmath190 isotopic ratio necessarily decreases in all agb stars . the same is true for the @xmath189/@xmath191 ratio that roughly behaves in a similar way . reasons are also rather similar , even if @xmath191 is less modified than @xmath190 in the hbs . at the surface of our @xmath30 ( @xmath113 ) with @xmath7 ( 0.005 ) agb stars , the abundances of both @xmath190 and @xmath191 can be enhanced to the expense of @xmath189 , if hbb is strong enough , leading to very low @xmath189/@xmath190 and @xmath189/@xmath191 ratios . as we will see in sect . 8.1 however , this only concerns the end of the tp - agb phase in extreme situations allowing hbb to occur a long time enough ( e.g. in case of relatively low mass loss rates ) . only a very small amount of @xmath43 is present in the convective envelope of e - agb stars as a result of the first ( and eventually the second ) dredge(s)-up . along the tp - agb phase , @xmath43 is largely produced in the hbs and most of it accumulates in the inter - shell region . inside full amplitude thermal pulses however , it is somewhat destroyed . for its part , @xmath192 is slightly produced in the hbs , especially inside the most massive agb stars . consequently , @xmath43/@xmath192 significantly increases with time during the tp - agb phase , due to the repetitive 3dup episodes . of course , dilution by convective mixing being more important for increasing @xmath193 , the highest ratios are expected at the surface of the less massive agb stars . however , the same extreme cases mentioned in the case of the mg isotopic ratios can lead to very high @xmath43/@xmath192 ratios for the most massive agb stars . > from si , the 3dup does not change significantly the surface isotopic ratios . however , as stressed in sect . 7.5 , when models will be able to reproduce the @xmath280 element enhancements at the surface of evolved agb stars , si and s isotopic ratios could be slightly modified by the high neutron flux . in this last section , we present our predictions concerning _ ( i ) _ the evolution of all the surface isotopic ratios along the tp - agb phase ( by including our evolutionary models and extrapolations up to the convective envelope removal ) and _ ( ii ) _ the total amount of material ejected at different ages for all the species up to @xmath9 . we refer to the previous sections for detailed explanations . thus , in the following , we will just comment these predictions . figs . [ xsurf302a ] and [ xsurf302b ] display , for the @xmath111 agb star with @xmath7 , the evolution of the isotopic ratios as a function of the total remaining mass that decreases due to mass loss . similar figures for the other six agb stars are only displayed in the electronic version . let us stress that we stop our calculations at the last thermal pulse that occurs before the convective envelope mass is reduced below @xmath363 . this stage indeed roughly corresponds to the pn ejection that we do not maintain to model here . this explains the quite large remaining total stellar masses in the figures , especially for those computed with the highest mass loss rates . the first comment concerns the formation of carbon stars . @xmath111 agb stars with @xmath7 and 3 and @xmath112 agb stars with @xmath1 become carbon stars during their tp - agb phase . the surface @xmath170/@xmath173 becomes greater than unity more rapidly for _ ( i ) _ decreasing total mass and/or _ ( ii ) _ decreasing @xmath6 . this metallicity dependence is confirmed by the increase of c stars compared to m stars from the galactic bulge to lmc and smc ( blanco et al . 1978 ) . our @xmath30 ( @xmath113 ) agb stars with @xmath7 ( 0.005 ) can also become c stars with our lowest mass loss rates . in such cases indeed , in spite of very efficient hbb transforming @xmath170 in @xmath8 and @xmath171 , many 3dup events can still occur with very reduced envelope masses , so that the relatively short inter - pulse phases are not long enough to substantially destroy @xmath170 . we will see however , together with other isotopic ratios , that these very low mass loss rates seem actually extreme . let us emphasize that the only super - lithium - rich star ( see sect . 7.2 ) we found to be also a carbon star is the @xmath113 with @xmath321 one ( with our medium mass loss rate ) . however , as already mentioned in sect . 6.2 , observations indicate that @xmath7 carbon stars are formed easier ( i.e. at lower luminosities ) than predicted by all the models . let us also stress the very high @xmath170/@xmath8 isotopic ratios that can be expected at the surface of @xmath272 @xmath112 agb stars , whatever @xmath6 . on the other hand , more massive agb stars experience hbb that progressively leads the @xmath170/@xmath8 isotopic ratio close to its cn cycle equilibrium value ( formation of j stars ) . the case of @xmath112 stars with @xmath1 very clearly shows the competition between the 3dup episodes ( increasing @xmath170/@xmath8 ) and hbb ( decreasing it ) . for lower mass loss rates , the 3dup events are more numerous for the same total mass decrease , leading to a more rapid @xmath170/@xmath8 increase . however , the corresponding longer tp - agb phase allows the base of the convective envelope to reach higher temperatures that finally destroy @xmath170 through the cn cycle . this efficient transformation of @xmath170 to @xmath8 also reduces the @xmath170/@xmath173 ratio , i.e. the possibility to form carbon stars . it happens if the temperatures at the base of the convective envelopes become higher than @xmath364 k. such stars have corresponding surface luminosities @xmath297 @xmath365 ( @xmath366 ) for @xmath7 ( 0.005 ) , i.e. @xmath367 ( -6.45 ) . we consequently predict that c stars can not be formed at higher luminosities than this threshold , due to strong hbb . this is in agreement with boothroyd et al . ( 1993 ) predictions . this also corresponds to the observed threshold for agb stars of the lmc ( cohen et al . 1981 ) . finally , note the surface pollution in @xmath38 that is more important , compared to that of @xmath8 , for decreasing total masses . for @xmath111 agb stars , the surface @xmath8/@xmath38 ratio is quite independent of the mass loss rate . this is due to the inter - pulse duration that is significantly greater than the @xmath38 lifetime , so that the @xmath38 decay is the dominant factor for the @xmath8/@xmath38 ratio evolution at the surface of such stars . the quite complicated behavior of the @xmath8/@xmath38 isotopic ratio at the surface of our most massive models is again due to the competition between the 3dup that adds @xmath38 into the convective envelope and hbb that can partially destroy it . by extrapolation , it is reasonable to expect even lower @xmath8/@xmath38 ratios in low - mass agb stars . as mentioned in forestini et al . ( 1996 ) , @xmath38 could become detectable for @xmath273 @xmath111 evolved agb stars . due to hbb , those stars showing very low surface @xmath170/@xmath8 ratios are also expected to have very high @xmath171/@xmath172 ones . it has to be stressed indeed that in our most massive agb stars , the cn cycle operating inside the convective envelope produces large amounts of @xmath171 . the @xmath173/@xmath174 and @xmath173/@xmath96 surface isotopic ratios also evolve as expected from our discussions in sect . note that quite low @xmath173/@xmath174 and very high @xmath173/@xmath96 ratios , which are clear signatures of the operation of the on cycle at the base of the convective envelope ( hbb ) , appear rather simultaneously with the cn cycle signatures in our most massive objects . also remark the progressive @xmath173/@xmath96 decrease in our @xmath30 ( @xmath113 ) agb star with @xmath7 ( 0.005 ) . for these objects , the temperature at the base of the convective envelope finally becomes high enough for @xmath173 to partially burn . the @xmath3 surface enhancements we predict are displayed in a special form , corresponding to the observation presentations adopted by jorissen et al . compared to their fig . 8 , two conclusions can be stated . * there is a clear observed correlation between the @xmath3 excess and the degree of @xmath170 surface pollution , that is perfectly understandable from the identified nucleosynthetic processes and the mixing episodes ( see sect . furthermore , our models clearly indicate a trend : the less massive the agb star , the more efficient the @xmath3 surface enhancement . in fact , our most massive models even show a surface @xmath3 depletion due to its partial destruction by hbb . * our @xmath7 models well reproduce only the less @xmath3 enriched stars in the [ @xmath3/@xmath173 ] versus @xmath170/@xmath173 diagram . based on the trend just mentioned , we predict that the agb stars showing very large @xmath3 surface enhancements have initial total masses @xmath273 @xmath111 . we plan to verify this statement , i.e. that low - mass agb stars are the most efficient @xmath3 galactic producers ( see sect . 9 ) . as expected , the @xmath189/@xmath190 and @xmath189/@xmath191 surface isotopic ratios are much less sensitive to the 3dup episodes and hbb . more precisely , only our most massive models with the lowest mass loss rates show large @xmath189 depletions due to a very strong hbb . again , these are however quite extreme situations . last but not least , we predict substantial @xmath43 surface enhancements due to the repetitive dredges - up of the hbs and inter - shell regions . the @xmath43/@xmath192 isotopic ratio tends to decrease with increasing total mass due to _ ( i ) _ the greater dilution as mixing occurs in a more massive convective envelope and _ ( ii ) _ the thinner hbs and inter - shell regions ( in mass ) . nevertheless , @xmath43/@xmath192 could well be increasing again if @xmath43 was produced inside the convective envelope itself by strong hbb . note however that such a very high @xmath43 production level is actually expected to be accompanied by very low mg isotopic ratios , as @xmath43 is produced by the mgal chain in the extreme situations mentioned above . let us finally mention that wasserburg et al . ( 1994 ) presented computations in which the evolution of the @xmath43 surface abundance has been followed self - consistently , by including the effect of neutron irradiation due to an enhanced amount of @xmath8 in the inter - shell region ( to engender a @xmath284process ) . our results compare very well with those obtained by boothroyd et al . ( 1994 , 1995 ) concerning the o isotopic ratios in intermediate - mass stars . they also mention the problem to reproduce the @xmath173/@xmath174 ratio , especially in low - mass carbon agb stars . we can not go further in this comparison without having yet computed the tp - agb phase of low - mass stars . boothroyd et al . ( 1995 ) already suggested that to conciliate all the data with observations of low - mass stars , one probably has to take into account of slow - particle transports inside radiative zones , what they called `` cool - bottom precessing '' . that is what we plan to do ( see sect . evolutionary models of tp - agb stars can be strongly constrained by the confrontation between predicted and observed surface isotopic ratios , especially if different ratios are observed for the same stars . for most of the observed agb stars , only the c and o isotopic ratios have been determined ( see e.g. harris & lambert 1984 ; harris et al . 1985 ; lambert et al . 1986 ; harris et al . 1987 ; harris & lambert 1987 ; kahane et al . globally , we found agreements . general features come out . * as the temperatures required at the base of the convective envelope to allow @xmath2 production ( see sect . 7.2 ) are significantly lower than those allowing efficient @xmath170 destruction by the cn cycle , we predict that luminous c stars enriched in @xmath2 can exist , especially for lower @xmath6 objects ( that have lower @xmath173 content ) . there are indeed a few % of the observed c stars that are also super - lithium - rich and there are more numerous in the lmc compared to our galaxy ( see e.g. smith & lambert 1990b for the lmc and feast 1974 for the galaxy ) . * there is no observed c star with @xmath170/@xmath8 @xmath369 . most of the j stars show @xmath170/@xmath8 ratios between 3 et 4 , i.e. very close to the cn cycle equilibrium value . * j stars also show rather low @xmath173/@xmath174 isotopic ratios while only upper limits have been found for the @xmath173/@xmath96 ratio . most of the other agb stars present @xmath173/@xmath174 ratios ranged between 300 and 4000 and @xmath173/@xmath96 ratios ranged between 500 and 5000 . only stars with rather high @xmath173/@xmath174 ratios are difficult to explain with our intermediate - mass agb models ( see also boothroyd et al . also note that the corresponding o isotopic ratios in the ism are @xmath173/@xmath174 @xmath370 and @xmath173/@xmath96 @xmath371 . again , there is a problem with @xmath174 . let us however stress that important uncertainties remain concerning the nuclear cross sections of proton captures on @xmath174 . on the other hand , low - mass agb stars are present among those observations . presently , there is only one star for which all the isotopic ratios have been determined . it is cw leo . this c star is one of the nearest evolved agb star ever observed . its distance is estimated to @xmath372 pc ( see e.g. claussen et al . 1987 ; keady et al . 1988 ; griffin 1990 ) . this corresponds to a surface luminosity @xmath373 @xmath374 and a present mass loss rate of roughly @xmath375 @xmath376@xmath377 . such a high wind makes it quite optically thick . a very extended circumstellar envelope has been detected to surround cw leo , irc+10216 . irc+10216 shows evidences for a bi - polar structure ( see e.g. gulin et al . 1993 ) , attesting that cw leo is very evolved and close to the end of its tp - agb phase . consequently , its present convective envelope mass is probably very reduced ( @xmath378 @xmath379 typically ) . the circumstellar envelope of cw leo allowed very accurate determinations of the isotopic ratios ( see kahane et al . 1988 and 1992 ; gulin et al . 1995 ; forestini et al . 1996 ) : @xmath170/@xmath8 @xmath380 , @xmath8/@xmath38 @xmath381 , @xmath171/@xmath172 @xmath382 , @xmath173/@xmath174 @xmath383 , @xmath173/@xmath96 @xmath384 , @xmath189/@xmath190 @xmath385 , @xmath189/@xmath191 @xmath386 and @xmath43/@xmath192 @xmath387 ( to be confirmed ) . furthermore , the si isotopic ratios appear to have interstellar medium values , i.e. unaltered by the star . as derived in gulin et al . ( 1995 ) and forestini et al . ( 1996 ) , almost all these isotopic ratios seem to be consistent ( within the error bars ) , following our predictions , with an agb star that had a main sequence total mass between 4 and @xmath113 . however , @xmath173/@xmath174 is again not well explained . as various surface isotopic ratios ( especially @xmath171/@xmath172 and @xmath173/@xmath96 ) are significantly changing between 4 and @xmath113 agb models , we are presently computing the tp - agb phase of a @xmath361 star with @xmath7 in order to improve this last prediction . finally , another very interesting way to constraint stellar evolution models of agb stars is to compare the predicted surface isotopic ratios with those determined in primitive meteorites ( mainly through grains included inside carbon rich chondrites ) . most of the @xmath170/@xmath8 , @xmath171/@xmath172 and @xmath43/@xmath192 isotopic anomalies measured in grains ( see e.g. ott 1991 ; zinner et al . 1991a ) are consistent with those coming from the surface of evolved tp - agb stars of intermediate - mass . this is of course in favor of the presence , when the solar system has been formed , of one or a few agb stars in its surroundings . such grains are indeed well known to be formed in the cool atmosphere of evolved agb stars ( i.e. when mass loss rates become rather high ) . concerning the measured o isotopic ratios , the situation is not so clear as many grains have ratios significantly out of the corresponding predicted ranges . some of these grains at least can have been formed in low - mass agb stars . again , as mentioned in sect . 8.1 , boothroyd et al . ( 1995 ) recently suggested that in such stars , cool - bottom processing can help to reduce these discrepancies . on the contrary , some other grains perfectly agree with our predictions . this is for example the case of a @xmath388 @xmath389 grain of the bishunpur ll3.1 chondrite by huss et al . the derived ratios , @xmath173/@xmath174 @xmath390 , @xmath173/@xmath96 @xmath391 and @xmath43/@xmath192 @xmath392 , are in perfect agreement with ratios corresponding to a @xmath393 @xmath112 agb star of nearly solar metallicity . finally , other sic @xmath43-rich grains raise problems , like that discovered by zinner et al . ( 1991b ) , exhibiting a very high @xmath43/@xmath192 ratio of @xmath394 . it seems hard to explain such a value from our models . however , this grain also exhibits an anomalously large @xmath395/@xmath396 isotopic ratio that seems to exclude its agb star origin . tables [ yield1 ] to [ yield3 ] give the total mass ejected by the wind ( in @xmath376 ) for various elements _ ( i ) _ at the top of the first rgb , _ ( ii ) _ at the beginning of the tp - agb phase and _ ( iii ) _ at the end of the tp - agb phase ( thus including our extrapolated nucleosynthesis calculations ) . as such data mainly concerns the chemical evolution of galaxies , we also present the net yield ( positive or negative ) for each element . the data of tables [ yield1 ] to [ yield3 ] correspond to the medium mass loss rates ( see sect . 8.1 ) , i.e. the solid lines of figs . [ xsurf302a ] and [ xsurf302b ] . brief comments are made about these predictions in the electronic version . llccccccc & & & + & & & + & & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath125&_at rg top_&2.9978&3.9985&4.9970&5.9980&2.9996&3.9999&4.9997 + & _ at @xmath164_&2.9151&3.7732&4.7009&5.7167&2.8053&3.5067&4.3149 + & _ at agb tip_&1.0989&1.0143&1.0750&1.3735 & 0.9471&1.2519&1.3818 + @xmath162&_at rg top_&2.59@xmath397&1.51@xmath397&2.93@xmath397&1.59@xmath397&7.73@xmath398&9.70@xmath165&3.54@xmath398 + & _ at @xmath164_&1.30@xmath144&2.35@xmath144&2.41@xmath144&1.96@xmath144&3.26@xmath144&5.70@xmath144&6.52@xmath144 + & _ at agb tip_&2.87@xmath143&3.03@xmath143&1.57@xmath143&5.00@xmath144&3.30@xmath143&3.04@xmath143&9.19@xmath144 + & _ net yield _ & 2.25@xmath143 & 2.07@xmath143 & 3.46@xmath144&-9.44@xmath144 & 2.74@xmath143 & 2.29@xmath143&-4.27@xmath399 + @xmath48&_at rg top_&6.53@xmath143&4.28@xmath143&8.69@xmath143&5.84@xmath143&1.15@xmath143&1.89@xmath144&8.10@xmath144 + & _ at @xmath164_&2.59@xmath400&6.91@xmath400&9.59@xmath400&9.13@xmath400&5.11@xmath400&0.13&0.18 + & _ at agb tip_&0.59&0.91&1.32&1.59&0.55&0.72&1.28 + & _ net yield _ & 3.76@xmath400 & 3.20@xmath400 & 0.17 & 0.23 & 3.56@xmath400 & 3.70@xmath400 & 0.37 + @xmath2&_at rg top_&1.59@xmath167&1.02@xmath167&2.49@xmath167&1.45@xmath167&1.42@xmath401&2.67@xmath402&1.26@xmath401 + & _ at @xmath164_&2.63@xmath167&3.35@xmath167&5.18@xmath167&3.87@xmath167&4.47@xmath403&1.15@xmath167&1.96@xmath167 + & _ at agb tip_&2.92@xmath166&6.35@xmath167&7.33@xmath398&1.08@xmath397&8.23@xmath167&4.60@xmath166&3.57@xmath167 + & _ net yield_&-1.85@xmath398&-2.95@xmath398 & 3.44@xmath398 & 6.23@xmath398&-5.00@xmath165&-6.34@xmath165&-8.92@xmath165 + @xmath170&_at rg top_&6.44@xmath399&4.25@xmath399&9.11@xmath399&5.84@xmath399&2.14@xmath397&4.01@xmath398&1.90@xmath397 + & _ at @xmath164_&1.72@xmath143&4.53@xmath143&5.90@xmath143&5.56@xmath143&9.01@xmath144&2.34@xmath143&3.43@xmath143 + & _ at agb tip_&1.26@xmath400&9.71@xmath142&5.97@xmath142&2.74@xmath142&1.06@xmath400&4.40@xmath142&9.68@xmath143 + & _ net yield _ & 6.53@xmath142 & 1.33@xmath143&-6.62@xmath142&-1.21@xmath400 & 9.00@xmath142 & 2.19@xmath142&-1.93@xmath142 + @xmath8&_at rg top_&1.26@xmath397&8.52@xmath398&1.48@xmath397&1.14@xmath397&1.14@xmath398&2.17@xmath165&1.04@xmath398 + & _ at @xmath164_&8.75@xmath399&2.41@xmath144&3.23@xmath144&3.09@xmath144&4.90@xmath399&1.31@xmath144&1.91@xmath144 + & _ at agb tip_&1.95@xmath143&3.12@xmath143&1.00@xmath142&6.67@xmath143&5.05@xmath144&9.09@xmath144&1.93@xmath143 + & _ net yield _ & 1.22@xmath143 & 1.97@xmath143 & 8.50@xmath143 & 4.89@xmath143 & 3.06@xmath144 & 6.43@xmath144 & 1.58@xmath143 + @xmath38&_at rg top_&4.89@xmath404&4.77@xmath405&3.22@xmath406&5.69@xmath407&2.99@xmath408&9.59@xmath409&3.23@xmath406 + & _ at @xmath164_&6.92@xmath410&4.77@xmath405&3.62@xmath406&5.56@xmath169&1.68@xmath411&4.92@xmath412&3.38@xmath413 + & _ at agb tip_&6.20@xmath165&8.93@xmath165&2.90@xmath165&1.16@xmath165&8.54@xmath166&1.43@xmath165&4.35@xmath403 + & _ net yield _ & 6.20@xmath165 & 8.93@xmath165 & 2.90@xmath165 & 1.16@xmath165 & 8.54@xmath166 & 1.43@xmath165 & 4.35@xmath403 + @xmath171&_at rg top_&3.41@xmath399&2.22@xmath399&3.92@xmath399&2.99@xmath399&3.25@xmath397&4.47@xmath398&1.69@xmath397 + & _ at @xmath164_&2.35@xmath143&6.53@xmath143&9.26@xmath143&9.01@xmath143&1.53@xmath143&3.60@xmath143&4.66@xmath143 + & _ at agb tip_&5.24@xmath142&8.52@xmath142&1.84@xmath400&2.89@xmath400&1.58@xmath142&2.07@xmath142&1.04@xmath400 + & _ net yield _ & 3.01@xmath142 & 5.03@xmath142 & 1.38@xmath400 & 2.35@xmath400 & 9.75@xmath143 & 1.26@xmath142 & 9.33@xmath142 + @xmath172&_at rg top_&8.89@xmath165&5.83@xmath165&1.28@xmath398&8.02@xmath165&2.40@xmath166&4.36@xmath167&2.02@xmath166 + & _ at @xmath164_&1.92@xmath397&4.93@xmath397&6.33@xmath397&5.89@xmath397&1.01@xmath397&2.56@xmath397&3.70@xmath397 + & _ at agb tip_&4.13@xmath399&6.32@xmath399&1.97@xmath399&2.66@xmath399&1.02@xmath399&9.13@xmath397&5.13@xmath399 + & _ net yield_&-4.65@xmath399&-7.47@xmath399&-1.62@xmath144&-1.87@xmath144&-1.35@xmath399&-2.26@xmath399 & 9.53@xmath397 + llccccccc & & & + & & & + & & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath173&_at rg top_&2.23@xmath144&1.47@xmath144&3.00@xmath144&2.00@xmath144&1.10@xmath399&1.92@xmath397&8.22@xmath397 + & _ at @xmath164_&8.33@xmath143&2.20@xmath142&2.84@xmath142&2.67@xmath142&4.68@xmath143&1.21@xmath142&1.69@xmath142 + & _ at agb tip_&1.92@xmath400&2.86@xmath400&3.61@xmath400&3.96@xmath400&5.34@xmath142&6.74@xmath142&4.17@xmath142 + & _ net yield_&-1.36@xmath143&-1.66@xmath142&-3.72@xmath142&-7.33@xmath142 & 1.27@xmath143&-2.29@xmath143&-5.01@xmath142 + @xmath174&_at rg top_&2.18@xmath398&1.15@xmath398&1.51@xmath398&1.26@xmath398&4.19@xmath165&2.09@xmath166&4.38@xmath166 + & _ at @xmath164_&2.62@xmath399&5.56@xmath399&5.45@xmath399&4.51@xmath399&1.87@xmath399&3.63@xmath399&3.17@xmath399 + & _ at agb tip_&5.87@xmath144&7.26@xmath144&7.14@xmath144&1.17@xmath143&1.94@xmath144&2.03@xmath144&1.05@xmath144 + & _ net yield _ & 5.09@xmath144 & 6.03@xmath144 & 5.52@xmath144 & 9.80@xmath144 & 1.73@xmath144 & 1.75@xmath144 & 6.78@xmath399 + @xmath96&_at rg top_&4.74@xmath398&3.13@xmath398&6.61@xmath398&4.29@xmath398&1.79@xmath165&3.36@xmath166&1.59@xmath165 + & _ at @xmath164_&1.42@xmath399&3.76@xmath399&4.87@xmath399&4.57@xmath399&7.55@xmath397&1.97@xmath399&2.86@xmath399 + & _ at agb tip_&3.10@xmath144&4.84@xmath144&2.09@xmath144&5.32@xmath399&7.66@xmath399&1.06@xmath144&3.81@xmath399 + & _ net yield_&-1.26@xmath144&-2.01@xmath144&-6.91@xmath144&-1.01@xmath143&-4.11@xmath399&-5.16@xmath399&-1.69@xmath144 + @xmath3&_at rg top_&9.54@xmath166&6.23@xmath166&1.27@xmath165&8.46@xmath166&4.88@xmath167&8.38@xmath403&3.55@xmath167 + & _ at @xmath164_&3.70@xmath398&9.54@xmath398&1.22@xmath397&1.13@xmath397&2.07@xmath398&5.27@xmath398&7.34@xmath398 + & _ at agb tip_&1.87@xmath399&1.68@xmath399&2.07@xmath399&1.75@xmath399&4.31@xmath397&3.41@xmath397&1.62@xmath397 + & _ net yield _ & 1.06@xmath399 & 3.96@xmath397 & 3.84@xmath397&-2.33@xmath397 & 2.11@xmath397 & 4.65@xmath398&-2.26@xmath397 + @xmath186&_at rg top_&3.79@xmath399&2.49@xmath399&5.07@xmath399&3.39@xmath399&1.92@xmath397&3.24@xmath398&1.39@xmath397 + & _ at @xmath164_&1.46@xmath143&3.89@xmath143&5.12@xmath143&4.85@xmath143&8.34@xmath144&2.11@xmath143&2.93@xmath143 + & _ at agb tip_&3.25@xmath142&5.05@xmath142&6.69@xmath142&7.90@xmath142&8.64@xmath143&1.17@xmath142&1.58@xmath142 + & _ net yield_&-1.87@xmath399&-6.51@xmath144&-3.03@xmath144&-2.98@xmath144&-1.51@xmath144&-4.53@xmath399 & 2.59@xmath144 + @xmath187&_at rg top_&9.81@xmath165&6.45@xmath165&1.31@xmath398&8.76@xmath165&5.31@xmath166&8.75@xmath167&3.71@xmath166 + & _ at @xmath164_&3.83@xmath397&1.03@xmath399&1.35@xmath399&1.27@xmath399&2.37@xmath397&5.86@xmath397&7.85@xmath397 + & _ at agb tip_&1.01@xmath144&1.42@xmath144&9.29@xmath399&2.71@xmath399&2.83@xmath399&3.36@xmath399&1.15@xmath399 + & _ net yield _ & 1.79@xmath399 & 1.21@xmath399&-7.85@xmath399&-1.75@xmath144 & 5.92@xmath397 & 3.60@xmath397&-2.80@xmath399 + @xmath188&_at rg top_&2.99@xmath397&1.97@xmath397&4.06@xmath397&2.69@xmath397&1.38@xmath398&2.54@xmath165&1.11@xmath398 + & _ at @xmath164_&1.06@xmath144&2.81@xmath144&3.66@xmath144&3.45@xmath144&5.85@xmath399&1.54@xmath144&2.21@xmath144 + & _ at agb tip_&9.45@xmath143&6.81@xmath143&7.75@xmath143&8.94@xmath143&2.40@xmath143&1.39@xmath143&1.44@xmath143 + & _ net yield _ & 6.83@xmath143 & 2.69@xmath143 & 2.35@xmath143 & 2.56@xmath143 & 1.69@xmath143 & 4.44@xmath144 & 1.98@xmath144 + @xmath4&_at rg top_&8.40@xmath398&5.48@xmath398&1.07@xmath397&7.41@xmath398&5.66@xmath165&7.33@xmath166&2.92@xmath165 + & _ at @xmath164_&4.11@xmath399&1.13@xmath144&1.55@xmath144&1.49@xmath144&2.60@xmath399&6.08@xmath399&7.76@xmath399 + & _ at agb tip_&9.47@xmath144&1.49@xmath143&2.16@xmath143&2.51@xmath143&2.87@xmath144&3.50@xmath144&2.80@xmath144 + & _ net yield _ & 2.76@xmath144 & 4.30@xmath144 & 7.72@xmath144 & 8.80@xmath144 & 1.06@xmath144 & 1.07@xmath144&-3.98@xmath399 + llccccccc & & & + & & & + & & @xmath111&@xmath112&@xmath113&@xmath30&@xmath111&@xmath112&@xmath113 + @xmath189&_at rg top_&1.20@xmath399&7.90@xmath397&1.61@xmath399&1.08@xmath399&6.09@xmath398&1.03@xmath398&4.40@xmath398 + & _ at @xmath164_&4.62@xmath144&1.23@xmath143&1.63@xmath143&1.54@xmath143&2.64@xmath144&6.68@xmath144&9.29@xmath144 + & _ at agb tip_&1.03@xmath142&1.60@xmath142&2.12@xmath142&2.31@xmath142&2.74@xmath143&3.70@xmath143&2.28@xmath143 + & _ net yield_&-1.35@xmath399&-2.55@xmath144&-2.09@xmath144&-2.09@xmath143&-6.05@xmath399&-4.73@xmath399&-2.65@xmath143 + @xmath190&_at rg top_&1.60@xmath397&1.05@xmath397&2.14@xmath397&1.42@xmath397&8.26@xmath165&1.40@xmath165&5.96@xmath165 + & _ at @xmath164_&6.12@xmath399&1.63@xmath144&2.11@xmath144&1.99@xmath144&3.57@xmath399&9.04@xmath399&1.24@xmath144 + & _ at agb tip_&1.42@xmath143&2.22@xmath143&2.84@xmath143&5.17@xmath143&3.88@xmath144&5.13@xmath144&1.73@xmath143 + & _ net yield _ & 6.17@xmath399 & 7.72@xmath399 & 3.19@xmath399 & 1.85@xmath143 & 2.00@xmath399 & 2.06@xmath399 & 1.08@xmath143 + @xmath191&_at rg top_&1.82@xmath397&1.20@xmath397&2.44@xmath397&1.63@xmath397&9.31@xmath165&1.56@xmath165&6.69@xmath165 + & _ at @xmath164_&7.00@xmath399&1.87@xmath144&2.51@xmath144&2.39@xmath144&4.05@xmath399&1.03@xmath144&1.44@xmath144 + & _ at agb tip_&1.68@xmath143&2.57@xmath143&3.48@xmath143&4.06@xmath143&4.50@xmath144&5.98@xmath144&5.43@xmath144 + & _ net yield _ & 1.23@xmath144 & 1.22@xmath144 & 2.52@xmath144 & 2.57@xmath144 & 2.84@xmath399 & 3.39@xmath399&-1.99@xmath144 + @xmath414&_at rg top_&1.20@xmath169&2.83@xmath402&1.01@xmath402&1.93@xmath401&1.29@xmath168&4.29@xmath406&1.11@xmath406 + & _ at @xmath164_&6.15@xmath169&1.39@xmath166&5.63@xmath165&2.44@xmath398&3.10@xmath401&6.12@xmath166&2.73@xmath165 + & _ at agb tip_&2.03@xmath397&8.87@xmath398&4.03@xmath397&3.67@xmath399&4.34@xmath398&9.02@xmath165&7.69@xmath399 + & _ net yield _ & 2.03@xmath397 & 8.87@xmath398 & 4.03@xmath397 & 3.67@xmath399 & 4.34@xmath398 & 9.02@xmath165 & 7.69@xmath399 + @xmath192&_at rg top_&1.36@xmath397&8.91@xmath398&1.82@xmath397&1.22@xmath397&6.89@xmath165&1.16@xmath165&4.97@xmath165 + & _ at @xmath164_&5.21@xmath399&1.39@xmath144&1.84@xmath144&1.74@xmath144&2.99@xmath399&7.57@xmath399&1.05@xmath144 + & _ at agb tip_&1.18@xmath143&1.83@xmath143&2.45@xmath143&2.94@xmath143&3.27@xmath144&4.38@xmath144&7.42@xmath144 + & _ net yield _ & 1.49@xmath399&-5.26@xmath398 & 3.78@xmath399 & 1.00@xmath144 & 1.24@xmath399 & 1.60@xmath399 & 1.87@xmath144 + @xmath9&_at rg top_&1.53@xmath399&1.00@xmath399&2.05@xmath399&1.37@xmath399&7.76@xmath398&1.31@xmath398&5.60@xmath398 + & _ at @xmath164_&5.87@xmath144&1.57@xmath143&2.07@xmath143&1.96@xmath143&3.36@xmath144&8.52@xmath144&1.18@xmath143 + & _ at agb tip_&1.31@xmath142&2.04@xmath142&2.70@xmath142&3.18@xmath142&3.49@xmath143&4.73@xmath143&6.07@xmath143 + & _ net yield_&-1.47@xmath399&-2.76@xmath144&-1.35@xmath144&-1.84@xmath144&-5.95@xmath399&-2.20@xmath399&-1.85@xmath144 + we tried to confirm and improve our present knowledge of the structural and nucleosynthetic evolution of intermediate - mass agb stars . to do that , we have first presented in detail _ ( i ) _ the physics and numerical aspects of our stellar evolution code , _ ( ii ) _ results concerning evolution phases prior to the agb one and _ ( iii ) _ comparisons with other works . these informations are essential to better understand and appraise the large set of predictions we discuss about the thermally pulsing agb stage . we also compare our predictions with various observations ( at different evolutionary stages ) in order to clearly identify the processes that should be included in future models . * concerning the structure ( intensities , temperatures ) of the thermal pulses as well as the thermal properties at the base of the convective envelope , the most decisive quantities are the total mass and core mass . both change with time in a way that is governed by the increasing mass loss rate . this translates into a great sensitivity of the various nucleosynthesis processes to the adopted mass loss rate . consequently , the evolution of the thermal pulse characteristics crucially depend on this last quantity , unfortunately still rather badly known . in addition , the lower the mass loss rates , the longer the tp - agb phase , the more third dredge - up events have time to occur and the greater are the surface abundance changes due to hbb . this is clearly demonstrated by the high sensitiveness of the predicted surface evolution of almost all the isotopic ratios to the mass loss rate dependence on time . * on the other hand , at a given total stellar mass , core masses are greater for lower metallicities , as a result of the central burning phases . as a consequence , for the reason mentioned at the former point , tp - agb stars with @xmath1 mostly behave like @xmath7 ones with a total mass greater by @xmath415 @xmath416 . however , due to important differences in initial compositions that also influence the nuclear reactions , this is not completely true as far as isotopic ratios are concerned . at the end of the tp - agb phase indeed , some yields are quite different for both @xmath6 we have studied . * the neutron captures on intermediate - mass and heavy nuclides occur in two distinct sites inside tp - agb stars , namely _ ( i ) _ at the base of the inter - shell zone during the inter - pulse phase and , as already known , _ ( ii ) _ inside the convective tongues associated with thermal pulses . along the asymptotic tp - agb , the amount of produced neutrons is rather independent of the initial total mass and metallicity ( at least for intermediate - mass objects ) . * concerning chemical evolution of galaxies , intermediate - mass stars , at the end of their existence , appear to be rather significant producers of @xmath162 , @xmath48 , @xmath2 ( most massive and/or higher @xmath6 stars ) , @xmath8 , @xmath171 , @xmath174 , @xmath3 , @xmath188 , @xmath4 , @xmath191 and @xmath192 . a found trend , compared to observations , however clearly supports the idea that most of the @xmath3 present in the interstellar medium has to come from low - mass agb stars ( that point has to be confirmed by forthcoming evolutionary models ) . + intermediate - mass agb stars also substantially produce radio - nuclides , namely @xmath38 and @xmath259 . the former one mostly comes from lower mass stars while it is the contrary for the last one . + these stars partially deplete the interstellar medium content in @xmath2 ( less massive stars ) , @xmath172 , @xmath173 , @xmath96 , @xmath186 , @xmath189 and @xmath9 . * we have not yet identified the @xmath8 additional source needed to quantitatively explain the @xmath284process . this process is however required in order to account for many observations unquestionably indicating that agb stars are responsible for the main component of the solar system heavy element distribution . some suggestions have been made . radiative diffusion operating during the tp - agb phase of low - mass stars could substantially help . although already invoked , such a process has never been yet included in agb star evolution models . * the rather high level of @xmath2 production detected at the surface of some evolved but relatively faint agb stars of our galaxy still remains unaccountable . thus , the cameron - fowler mechanism does not explain all the super - lithium - rich agb stars . as these galactic objects have relatively low initial masses , the explanation could also come from the radiative slow - transport processes above mentioned . * last but not least , some c stars are observed , especially in the galactic bulge , with low luminosities at which the models fail at predicting the occurrence of the third dredge - up . this is probably due to our bad knowledge ( and treatment ) of finely tuned convection motions in boundaries of stellar regions having strong chemical composition gradients . clearly , some kind of extra - mixing ( deeper inside nuclearly processed regions ) has to be invoked . however , no self - consistent mechanism has been settled and tested through complete evolution models up to now . other confrontations with observations require the modeling of low - mass agb stars . a clear distinction between low- and intermediate - mass stars is justified . indeed , numerous observations indicate that `` non - standard '' particle transport processes are acting inside low - mass stars at different phases of their evolution , that substantially modify the chemical structure compared to what is obtained in classical models . during the various dredge - up episodes , matter up - heaved to the surface has consequently a different composition . this is not the case in intermediate - mass stars . in conclusion , low - mass agb stellar models , maintaining to make detailed predictions concerning the evolution of surface isotopic ratios , have to include such slow - particle transport processes . they are currently being calculated and will be presented in a next - coming paper . we are first indebted to maurizio busso , the referee , for its very careful reading of our manuscript that helped us to significantly improve and clarify it . we also thank him and roberto gallino for the very instructive discussions we already had together about these very complex agb stars . let us finally thank lionel siess for his always valuable contributions to improve the stellar evolution code and related utilities . part of the computations presented in this paper ( roughly representing 7 months of cpu time ) were performed at the `` center de calcul intensif de lobservatoire de grenoble '' . most of them have been realized at `` imag '' on a ibm sp1 computer financed by the mesr , cnrs and rgion rhne - alpes . this work was supported by grants from the gdr `` structure interne des etoiles et des plantes gantes '' ( cnrs ) . aaronson , m. , mould , j. 1985 , apj , 288 , 551 abia , c. , boffin , h.m.j . , isern , j. , rebolo , r. 1991 , a&a , 245 , l1 abia , c. , boffin , h.m.j . , isern , j. , rebolo , r. 1993 , a&a , 272 , 455 alexander , d. r. , ferguson , j. w. 1994 , apj , 437 , 879 allen , b.j . , machlin , r.l . , gibbons , j.h . 1971 , adv . , 4 , 205 alongi , m. , bertelli , g. , bressan , a. , chiosi , c. 1991 , in formation and evolution of star clusters , ed . j. kanneth ( asp conf . 13 ) , 223 anders , e. , grevesse , n. 1989 , geochim . acta , 53 , 197 bao , z.y . , kppeler , f. 1987 , atomic data nucl . data tables , 36 , 411 bazan , g. , lattanzio , j.c . 1993 , apj , 409 , 762 beer , h. , wiescher , m. , kppeler , f. , gorres , j. , koehler , p.e . 1991 , preprint bell , r.a . , eriksson , k. , gustafsson , b. , nordlund , a. 1976 , a&a , 23 , 37 blanco , b.m . , blanco , v.m . , mccarthy , m.f . 1978 , nat , 271 , 638 blanco , v.m . , mccarthy , m.f . , blanco , b.m . 1980 , apj , 242 , 938 blcker , t. 1995 , a&a , 297 , 727 blcker , t. , schnberner , d. 1991 , a&a , 244 , l43 boothroyd , a.i . , sackmann , i .- j . 1988a , apj , 328 , 641 boothroyd , a.i . , sackmann , i .- j . 1988b , apj , 328 , 653 boothroyd , a.i . , sackmann , i .- j . 1988c , apj , 328 , 671 boothroyd , a.i . , sackmann , i .- j . , ahern , s.c . 1993 , apj , 416 , 762 boothroyd , a.i . , sackmann , i .- , wasserburg , g.j . 1994 , apj , 430 , l77 boothroyd , a.i . , sackmann , i .- j . , wasserburg , g.j . 1995 , apj , 442 , l21 brehm , k. , becker , h.w . , rolfs , c. , kppeler , f. , ratynski , w. 1988 , phys . a. atomic nuclei , 330 , 167 bressan , a. , fagotto , f. , bertelli , g. , chiosi , c. 1993 , a&as , 100 , 647 bryan , g.l . , volk , k. , kwok , s. 1990 , apj , 365 , 301 busso , m. , lambert , d.l . , beglio , l. , gallino , r. , raiteri , c.m . , smith , v.v . 1995 , apj , 446 , 775 cameron , a.g.w . , fowler , w.a . 1971 , apj , 164 , 111 caughlan , g.r . , fowler , w.a . 1988 , atomic data nucl . data tables , 40 , 283 caughlan , g.r . , fowler , w.a . , harris , m.j . , zimmerman , b.a . 1985 , atomic data nucl . data tables , 32 , 197 champagne , a.e . , cella , c.h . , konzes , r.t . , loury , r.m . , magnus , p.v . , smith , m.s . , mao , z.q . 1988 , nucl . a , 487 , 433 charbonnel , c. 1994 , a&a , 282 , 811 charbonnel , c. 1995 , apj , 453 , l41 claussen , m.j . , kleinmann , s.g . , joyce , r.r . , jura , m. 1987 , apjs , 65 , 385 cohen , j.g . , frogel , j.a . , persson , s.e . , elias , j.h . 1981 , apj , 249 , 481 de oliveira , f. 1995 , ph . d. thesis ( universit de paris xi , france ) el eid , m.f . 1994 , a&a , 285 , 915 feast , m.w . 1974 , in stellar instability and evolution , eds . p. ledoux , a. noels and a.w . rodgers ( dordrecht , reidel ) , 93 forestini , m. , gulin , m. , cernicharo , j. 1996 , a&a , _ in press _ forestini , m. , paulus , g. , arnould , m. 1991 , a&a , 252 , 597 forestini , m. , goriely , s. , jorissen , a. , arnould , m. 1992 , a&a , 261 , 157 fowler , w.a . , caughlan , g.r . , zimmerman , b.a . 1967 , ara&a , 5 , 525 frost , j.c . , lattanzio , j. 1996 , apj , in press fujimoto , m.y . , sugimoto , d. 1979 , pasj , 31 , 1 funck , c. , langanke , k. 1989 , apj , 344 , 46 gilroy , k.k . 1989 , apj , 347 , 835 graboske , h.c . , dewitt , h.e . , grossman , a.s . , cooper , m.s . 1973 , apj , 181 , 457 grevesse , n. 1991 , in evolution of stars : the photospheric abundance connection , eds . g. michaud and a. tutukov ( dordrecht , kluwer ) , 63 griffin , i.p . 1990 , mnras , 247 , 591 groenewegen , m.a.t . , de jong , t. 1993 , a&a , 267 , 410 gulin , m. , lucas , r. , cernicharo , j. 1993 , a&a , 280 , l19 gulin , m. , forestini , m. , valiron , p. , ziurys , l.m . , anderson , m.a . , cernicharo , j. , kahane , c. 1995 , a&a , 297 , 183 hammer , j.w . 1991 , preprint harris , m.j . , lambert , d.l . 1984 , apj , 281 , 739 harris , m.j . , lambert , d.l . 1987 , apj , 318 , 868 harris , m.j . , lambert , d.l . , smith , v.v . 1985 , apj , 299 , 375 harris , m.j . , lambert , d.l . , smith , v.v . 1988 , apj , 321 , 761 harris , m.j . , lambert , d.l . , hinkle , k.h . , gustafsson , b. , eriksson , k. 1987 , apj , 316 , 294 hubbard , w.b . , lampe , m. 1969 , apjs , 18 , 297 huss , g.r . , fahey , a.j . , gallino , r. , wasserburg , g.j . 1994 , apj , 430 , l81 iben , i. 1964 , apj , 140 , 1631 iben , i. 1975 , apj , 196 , 525 iben , i. 1976 , apj , 208 , 165 iben , i. 1981 , apj , 246 , 278 iben , i. 1983 , apj , 275 , l65 illiadis , c. 1990 , nucl . phys . a , 512 , 509 itoh , n. , mutoh , h. , hikita , a. , kohyama , y. 1992 , apj , 395 , 622 jorissen , a. , arnould , m. 1989 , a&a , 221 , 161 jorissen , a. , smith , v.v . , lambert , d.l . 1992 , a&a , 261 , 164 kahane , c. , gmez - gonzles , j. , cernicharo , j. , gulin , m. 1988 , a&a , 190 , 167 kahane , c. , cernicharo , j. , gmez - gonzles , j. , gulin , m. 1992 , a&a , 256 , 235 kppeler , f. , gallino , r. , busso , m. , picchio , g. , raiteri , c.m . 1990 , apj , 354 , 630 kppeler , f. et al . 1994 , apj , 437 , 396 keady , j.j . , hall , d.n.b . , ridgway , s.t . 1988 , apj , 326 , 832 kious , m. 1990 , ph.d . thesis ( universit de paris xi , france ) kippenhahn , r. , weigert , a. 1991 , stellar structure and evolution ( springer - verlag ) , 23 kippenhahn , r. , weigert , a. , hofmeister , e. 1968 , methods of computational physics , 7 , 129 knapp , g.r . 1989 , apj , 336 , 822 koester , d. 1976 , a&a , 52 , 415 kohyama , y. , itoh , n. , obama , a. , mutoh , h. 1993 , apj , 415 , 267 lambert , d.l . , gustafsson , b. , eriksson , k. , hinkle , k.h . 1986 , apjs , 62 , 373 lambert , d.l . , smith , v.v . , busso , m. , gallino , r. , straniero , o. 1995 , apj , 450 , 302 landr , v. , prantzos , n. , aguer , p. , bogaert , g. , lefbvre , a. , thibaud , j .- 1990 , a&a , 240 , 85 lattanzio , j. 1986 , apj , 311 , 708 lattanzio , j. 1987 , apj , 313 , l15 lattanzio , j. 1989 , in evolution of peculiar red giants , eds . johnson and b. zuckerman ( cambridge university press , cambridge ) , 161 malaney , r.a . , boothroyd , a.i . 1987 , apj , 320 , 866 malaney , r.a . , savage , m.j . , boothroyd , a.i . 1988 , apj , 324 , 948 merrill , p.w . 1952 , apj , 116 , 21 mould , j.r . , aaronson , m. 1982 , apj , 263 , 629 mould , j.r . , aaronson , m. 1986 , apj , 303 , 10 munakata , h. , kohyama , y. , itoh , n. 1985 , apj , 296 , 197 nagai , y. , igashira , m. , takeda , k. , mukai , n. , motoyama , s. , uesawa , f. , kitazawa , h. , fukuda , t. 1991 , apj , 372 , 683 nagai , y. et al . 1995 , in nuclei in the cosmos iii , eds . m. busso , r. gallino and c.m . raiteri ( aip , new york ) , 201 ott , u. 1991 , protostars and planets iii ( university of arizona press , tuscon ) , 144 paczyski , b. 1975 , apj , 202 , 558 plez , b. 1992 , a&as , 94 , 527 plez , b. , smith , v.v . , lambert , d.l . 1993 , apj , 418 , 812 press , , w.h . , flannery , b.p . , teukolsky , s.a . , vetterling , w.t . 1986 , numerical recipes ( cambridge university press ) , 588 raman , s. , igashira , m. , dozono , y. , kitazawa , h. , mizimoto , m. , lynn , j.e . 1990 , phys . c , 41 , 458 reid , n. , mould , j. 1984 , apj , 284 , 98 reimers , d. 1975 , mem . soc . lige , 6th ser . , 8 , 369 rogers , f.j . , iglesias , c.a . 1992 , apjs , 79 , 507 sackmann , i .- j . 1980 , apj , 235 , 554 sackmann , i .- j . , boothroyd , a.i . 1991 , in iau symp . 145 , eds . g. michaud and a. tutukov ( dordrecht , reidel ) , 275 sackmann , i .- j . , boothroyd , a.i . 1992 , apj , 392 , l71 sackmann , i .- smith , r.l . , despain , k.h . 1974 , apj , 187 , 555 scalo , j.m . , despain , k.h . , ulrich , r.k . 1975 , apj , 196 , 805 schaller , g. , schaerer , d. , meynet , g. , maeder , a. 1992 , a&as , 96 , 269 schwarzschild , m. , hrm , r. 1965 , apj , 142 , 855 smith , v.v . , lambert , d.l . 1986 , apj , 311 , 843 smith , v.v . , lambert , d.l . 1989 , apj , 345 , l75 smith , v.v . , lambert , d.l . 1990a , apjs , 72 , 387 smith , v.v . , lambert , d.l . 1990b , apj , 361 , l69 smith , v.v . , plez , b. , lambert , d.l . 1995 , apj , 441 , 735 straniero , o. , gallino , r. , busso , m. , chieffi , a. , raiteri , c.m . , limongi , m. , salaris , m. 1995 , apj , 440 , l85 straniero , o. et al . 1996 , in torino workshop on nucleosynthesis of s- and r - nuclei i ( in press ) thielemann , f .- k . 1991 , in nuclei in the cosmos i , eds . h. oberhummer and c. rolfs ( springer - verlag , berlin ) , 171 vassiliadis , e. , wood , p.r . 1993 , apj , 413 , 641 vogelaar , r.b . 1989 , ph.d . thesis ( california institute of technology , california ) wagoner , r.v . 1969 , apjs , 162 , 247 wannier , p.g . , sahai , r. 1986 , apj , 311 , 335 wasserburg , g.j . , busso , m. , gallino , r. , raiteri , c.m . 1994 , apj , 424 , 412 weidemann , v. 1987 , a&a , 188 , 74 weidemann , v. , koester , d. 1983 , a&a , 121 , 77 wiesher , m. , gorres , j. , thielemann , f .- d . 1990 , apj , 363 , 340 wood , p.r . 1981 , in physical processes in red giants , eds . i. iben and a. renzini ( dordrecht , reidel ) , 135 wood , p.r . , bessell , m.s . , fox , m.w . 1983 , apj , 272 , 99 woosley , s.e . , fowler , w.a . , holmes , j.a . , zimmerman , b.a . 1978 , atomic data nucl . data tables , 22 , 371 zinner , e. , amari , s. , lewis , r. 1991b , in meteoritic society meeting ( monterey , california ) , 17 zinner , e. , amari , s. , anders , e. , lewis , r. 1991a , nat , 349 , 51 zuckerman , b. , dyck , h.m . , claussen , m.j . 1986 , apj , 304 , 401

the path integral ( pi ) has been used in quantum physics since the revolutionary work of feynman @xcite , although the basic observation goes back to dirac @xcite who appreciated the rle of the lagrangian in short - time evolution of the wave function , and even suggested the time - slicing procedure for finite , i.e. , non - infinitesimal , time lags . since then the pi approach yielded invaluable insights into the structure of quantum theory @xcite and provided a viable alternative to the traditional operator - formalism - based canonical quantization . during the second half of the 20th century , the pi became a standard tool in quantum field theory @xcite and statistical physics @xcite , often providing the easiest route to derivation of perturbative expansions and serving as an excellent framework for ( both numerical and analytical ) non - perturbative analysis @xcite . feynman pi has its counterpart in pure mathematics , namely , in the theory of continuous - time stochastic processes @xcite . there the concept of integration over a space of continuous functions ( so - called fluctuating paths or sample paths ) had been introduced by wiener @xcite already in 1920 s in order to represent and quantify the brownian motion . interestingly enough , this so - called wiener integral ( or integral with respect to wiener measure ) was formulated 2 years before the discovery of the schrdinger equation and 25 years before feynman s pi formulation . the _ local time _ for a brownian particle ( in some literature also called _ sojourn time _ ) has been of interest to physicists and mathematicians , since the seminal work of paul lvy in 1930 s @xcite . in its essence , the local time characterizes the time that a sample trajectory @xmath1 of a given stochastic process spends in the vicinity of an arbitrary point @xmath2 . this in turn defines a sample trajectory @xmath3 of a new stochastic process . a rich theory has been developed for local - time processes that stem from diffusion processes ( see , e.g. , ref . @xcite and citations therein ) . for later convenience , we should particularly highlight the ray knight theorem which states that the local time of the wiener process can be expressed in terms of the squared bessel process @xcite . in contrast to mathematics , the concept of the local time is not uniquely settled in physics literature . various authors define essentially the same quantity under different names ( local time , occupation time , traversal time , etc . ) , and with different applications in mind . for example , in ref . @xcite the _ traversal time _ is used to study quantum scattering and tunneling processes , in @xcite the small - temperature behavior of the equilibrium density matrix is analyzed with a help of the _ occupation time _ , while in ref . @xcite the large - time behavior of path integrals that contain functionals of the _ local time _ is discussed . the aim of this paper is to derive a local - time pi representation of the bloch density matrix , i.e. , the matrix elements @xmath4 of the gibbs operator . this can serve not only as a viable alternative to the commonly used feynman kac representation but also as a powerful tool for extracting both large and small - temperature behavior . apart from the general theoretical outline , our primary focus here will be on the low - temperature behavior which is technically more challenging than the large - temperature regime . in fact , the large - temperature expansion was already treated in some detail in our previous paper @xcite . last but not least , we also wish to promote the concept of the local time which is not yet sufficiently well known among the path - integral practitioners . the structure of the paper is as follows . to set the stage we recall in the next section some fundamentals from the feynman pi which will be needed in later sections . in section [ sec : heuristic ] we provide motivation for the introduction of a local time , and construct a heuristic version of the local - time representation of pi s . the key technical part of the article is contained in section [ sec : lt ] , where we derive by means of the replica trick the local - time representation of the bloch density matrix . relation to the sturm liouville theory is also highlighted in this context . a local - time analog of the feynman salam formula @xcite is presented in section [ sec : functionals ] and its usage is illustrated with a computation of the one - point distribution of the local time . since a natural arena for local - time pi s is in thermally extremal regimes , we confine our attention in section [ sec : asymp ] on large- and small-@xmath5 asymptotic behavior of the bloch density matrix . there we also derive an explicit leading - order behavior in large-@xmath5 ( i.e. , low - temperature ) expansion . the analysis is substantially streamlined by using the rayleigh ritz variational principle . finally , section [ sec : conclusion ] summarizes our results and discusses possible extensions , applications , and future developments of the present work . for the reader s convenience the paper is supplemented with appendix which clarifies some finer technical details . consider a non - relativistic one - dimensional quantum - mechanical system described by a time - independent hamiltonian @xmath6 where @xmath7 . throughout this paper we will study the matrix elements @xmath8 of the gibbs operator @xmath9 , where @xmath10 is the inverse temperature and @xmath11 is the boltzmann constant . the matrix @xmath12 , known also as the bloch density matrix , is a fundamental object in quantum statistical physics , as the expectation value of an operator @xmath13 at the temperature @xmath14 can be written in the form @xmath15 where @xmath16 is the partition function of the system . in case of need , ensuing quantum mechanical transition amplitudes can be obtained from ( [ rhodef ] ) via a wick rotation which formally amounts to the substitution @xmath17 , converting thus the gibbs operator @xmath9 to the quantum evolution operator @xmath18 . the matrix elements in eq.([rhodef ] ) can be represented via the path integral as @xcite @xmath19 \right\}\ , .\ ] ] this represents a sum " over all continuous trajectories @xmath20 , @xmath21 $ ] , connecting the initial point @xmath22 with the final point @xmath23 . it should be noted that the integral @xmath24 $ ] is the classical euclidean action integral along the path @xmath20 with @xmath25 . in the following we will denote the euclidean action as @xmath26 . the integrand in @xmath26 , i.e. , @xmath27 , can be identified with the classical hamiltonian function , in which the momentum @xmath28 is substituted for @xmath29 . one can also regard ( [ pix ] ) as an expectation value of the functional @xmath30 $ ] over the ( driftless ) brownian motion with the diffusion coefficient @xmath31 , and duration @xmath32 , that starts at point @xmath33 , and terminates at @xmath34 . the latter stochastic process is also known as brownian bridge . the purpose of this section is twofold . firstly , we would like to motivate a need for reformulation of pi s in the language of local - time stochastic process . in particular , we point out when such a reformulation can be more pertinent than the conventional `` sum over histories '' prescription . secondly , we wish to outline a heuristic construction of the local - time representation of pi s . more rigorous and explicit ( but less intuitive ) formulation of pi s over ensemble of local times will be presented in the subsequent section . to provide a physically sound motivation for the local - time representation of pi s we follow exposition of paulin _ . @xcite . to this end we first consider the diagonal elements of the bloch density matrix , i.e. , @xmath35 ( often referred to as the boltzmann density ) . upon shifting @xmath36 , and setting @xmath37 , @xmath38 ( @xmath39 is the thermal de broglie wavelength ) , the pi ( [ pix ] ) can be reformulated in terms of dimensionless quantities @xmath40 and @xmath41 as @xmath42 \right\ } .\ ] ] note in particular , that in contrast to @xmath43 the measure @xmath44 does not explicitly depend on @xmath5 , and thus @xmath5-dependent parts in the pi are under better control . such a rescaled representation is particularly useful when discussing large- and/or small-@xmath5 behavior of the path integral in question . path fluctuations in the potential are controlled by @xmath45 , and the factor @xmath5 quantifies the significance of the potential @xmath46 with respect to the kinetic term . for small @xmath5 ( i.e. , high temperature ) , typical paths @xmath47 stay in the vicinity of the point @xmath33 , as depicted in fig . [ fig : paths ] , are plotted as functions of the dimensionless time @xmath40 . the solid green path , representing a typical trajectory with a high value of @xmath5 , exhibits large fluctuations , whereas the dashed red path , corresponding to small @xmath5 , stays in the vicinity of the initial and final point @xmath33 . on the right , two local - time profiles @xmath48 are shown . the broad one ( solid green ) arises from the violently fluctuating path @xmath49 , whereas the narrow one ( dashed red ) corresponds to the path with small fluctuations . on the left , we depict a generic potential @xmath50 . ] and therefore a systematic wigner kirkwood expansion can be readily developed by taylor - expanding the potential part of the action @xcite . when @xmath5 is large ( i.e. , low temperature ) , the trajectories @xmath49 fluctuate heavily around the value @xmath33 , and the potential term @xmath46 dominates over the kinetic one . from the statistical physics point of view , the most important contribution to the low temperature behavior of the path integral ( [ pixresc ] ) should come from those paths that spend a sizable amount of time near the _ global _ minimum of the potential @xmath50 . for this reason , it is important to be able to keep track of the time which a given path spends in an infinitesimal neighborhood of an arbitrary point @xmath0 . let us define , for each wiener trajectory @xmath20 present in the feynman path integral ( [ pix ] ) the ensuing _ local time _ as @xmath51 since the local time @xmath52 is a functional of the random trajectory @xmath53 for @xmath54 , it represents a random variable . from the definition ( [ ltdef ] ) we can immediately see that @xmath55 for all @xmath56 , @xmath57 and that @xmath3 has a compact support . in addition , it can be proved @xcite that local - time trajectories @xmath3 are , with probability one , continuous curves which ( similarly as trajectories in the underlying wiener process ) are nowhere differentiable . in fig . [ fig : paths ] we depict two examples of representative local - time trajectories . an extensive mathematical discussion of properties of the local time can be found , e.g. in refs . @xcite . with the definition ( [ ltdef ] ) the potential part of the euclidean action can be recast into form @xmath58 . a local - time representation of the bloch density matrix @xmath12 is then given by @xmath59 \ \!\delta \ ! \left ( \int_\mathbb{r } dx \ \ ! l^x - \beta\hbar \right ) \exp \left[-\frac{1}{\hbar } \int_\mathbb{r } dx \ \ ! l^x v(x ) \right ] , \ ] ] where the pi `` sum '' is taken over all local - time trajectories @xmath60 , with @xmath0 being the independent variable ( not to be mistaken with the wiener trajectory @xmath20 ) . the @xmath61-function enforces the normalization constraint mentioned above . basically , transition to the local - time description represents a change ( or a functional substitution ) of stochastic variables @xmath62 . the weight factor @xmath63 appearing in ( [ heurform ] ) can be formally written in the form @xmath64 \ = \ \exp \left[-\frac{1}{\hbar}\int_{0}^{\beta\hbar}\!\ ! d\tau \ \ ! \frac{m}{2}\dot{x}^2 \right ] \ \ ! \det\left(\frac{\delta l^x(\beta\hbar)}{\delta x(\tau ) } \right)^{-1}\ , . \label{iii.7.a}\end{aligned}\ ] ] it is a function of @xmath33 , @xmath34 ( which are implicitly present in @xmath65 ) and @xmath5 , and a functional of the local time @xmath60 . of course , these cavalier manipulations do not have more than a heuristic nature , and it is , indeed , a non - trivial task to determine @xmath63 directly from ( [ iii.7.a ] ) . for this reason , we will in the following section tackle this problem indirectly . in this section , we present a derivation of the local - time representation of the bloch density matrix ( [ rhodef ] ) . initially , we limit ourselves to considering the case of the diagonal part , @xmath66 , and arrive at the key result ( [ diaglapl ] ) , which expresses the matrix elements in the ensuing laplace picture with respect to @xmath5 . this intermediate outcome is shown to agree with the sturm liouville theory . in the next step , we generalize the latter result to the off - diagonal elements ( @xmath67 ) . the inverse laplace transform will then yield the sought local - time representation of pi [ cf . ( [ offdiag ] ) ] . it follows from the definition ( [ rhodef ] ) that the function @xmath12 satisfies the heat equation @xmath68 \rho(x_a , x_b,\beta ) \ = \ 0\ , , \ ] ] with the initial condition @xmath69 . this is merely a wick - rotated ( @xmath70 ) analogue of the schrdinger equation . the feynman kac formula @xcite then ensures that the pi ( [ pix ] ) can be calculated by solving corresponding parabolic differential equation ( [ diffusioneq ] ) . in the laplace picture eq . ( [ diffusioneq ] ) takes the form @xmath71 \widetilde{\rho}(x_a , x_b , e ) \ = \ \delta(x_a - x_b)\ , , \ ] ] with @xmath72 . ( [ lapdifeq ] ) implies that @xmath73 is nothing but the green function of the operator @xmath74 . with the benefit of hindsight , we represent the green function @xmath73 as a path integral over fluctuating fields so - called functional integral @xcite . this is rather standard strategy in quantum field theory @xcite , and in our case it yields @xmath75 } } { \int_{\psi(x_{-})=0}^{\psi(x_{+})=0 } \mathcal{d}\psi(x ) \ \ ! e^ { -\frac{1}{2 } \mathcal{a}^e[\psi ] } } \ , , \ ] ] where @xmath76 & \ \equiv \ \int_{x_{-}}^{x_{+ } } \!\!\!dx \ \ ! \psi(x ) \left[-\frac{\hbar^2}{2 m } \frac{d^2}{d x^2 } + v(x ) + e \right ] \psi(x ) \nonumber\\[2 mm ] & \ = \ \int_{x_{-}}^{x_{+ } } \!\!\!dx \left [ \frac{\hbar^2}{2 m } \psi'(x)^2 + ( v(x ) + e ) \psi(x)^2 \right],\end{aligned}\ ] ] is the euclidean action functional of the field - theoretic path integral . the super - index `` @xmath77 '' in @xmath26 indicates the shift in the potential @xmath50 by the amount @xmath77 . here , we have confined our quantum - mechanical system within a finite box @xmath78 $ ] , with @xmath79 and @xmath80 . a real scalar field @xmath81 satisfies dirichlet boundary conditions @xmath82 so as to ensure the validity of the operations being performed . since we will ultimately want to invert the laplace transform to regain from @xmath83 the original bloch density matrix @xmath12 , we can not treat the denominator in ( [ fieldrepre ] ) as an irrelevant normalization constant ( which is the usual practice in quantum field theory ) but we have to take care of its @xmath77-dependence . to this end we take advantage of the formula @xmath84 which is a simple version of the so - called _ replica trick_. [ the usual replica - trick formula ( cf . e.g. , ref . @xcite ) can be obtained from ( [ rep.1a ] ) by integrating both sides with respect to @xmath85 and subsequently dividing by @xmath86 . ] with the help of ( [ rep.1a ] ) we can rewrite ( [ fieldrepre ] ) as a multidimensional functional integral @xmath87 } \ , , \ ] ] where the multiplet @xmath88 is a @xmath89-component replica " field in @xmath90 dimensions , @xmath91 denotes the scalar product @xmath92 , and we have rescaled the fields by a factor of @xmath93 in passing . the factor @xmath94 results from a pi generalization of the well know mean - value identity @xmath95 valid for any two vectors in @xmath89-dimensional statistically isotropic environments . as a side remark , note that we may now invert the laplace transform , using the trivial identity @xmath96 to obtain the representation @xmath97 } \ , .\ ] ] upon scaling @xmath98 , this agrees with the formula ( 2.9 ) in ref . @xcite . though the result ( [ replrepre ] ) generalizes to arbitrary number of dimensions of the @xmath0-space , i.e. , @xmath99 , our further development will be illustrated ( for simplicity s sake ) only on the one - dimensional case . in order to derive the weight factor @xmath63 ( cf . ( [ iii.7.a ] ) ) we have arrived at the representation ( [ replrepre ] ) with @xmath89 replica fields . this form is still not very transparent , and a further simplification step is needed to get rid of an explicit dependence of the measure on @xmath89 . to this end , we note that @xmath100 $ ] is in fact the action of a @xmath89-dimensional harmonic oscillator with the `` time '' variable @xmath0 , `` position '' variable @xmath101 , mass @xmath102 , and time - dependent `` frequency '' @xmath103 . considering for a moment only diagonal matrix elements , @xmath66 , spherical symmetry in the replica field space allows to reduce the path integral ( [ afterrepl ] ) to its radial part . due to the boundary conditions , @xmath104 , only the zero - angular - momentum ( @xmath40-wave ) contribution is non - vanishing ( generally a weighted sum over radial pi s with different angular momenta would be required ) . this will be rigorously justified at the end of this section . the corresponding _ radial _ pi representation for ( [ afterrepl ] ) reads @xcite @xmath105 } \ , .\ ] ] here , the radial part @xmath106 of the @xmath89-dimensional replica field @xmath107 is always non - negative , i.e. , @xmath108 ; the area of a unit sphere in @xmath89 dimensions , @xmath109 , may be replaced by its small-@xmath89 asymptotic form @xmath110 ; and @xmath111 have been introduced to regularize the origin of the @xmath107-space . the new action functional @xmath112 \ \equiv \ \mathcal{a}^e[\eta ] \ + \ \int_{x_-}^{x_+ } \!\ ! dx \ \ ! \frac{(d\!-\!1)(d\!-\!3)}{8 \eta^2(x ) } \ , , \ ] ] is the euclidean action functional of the _ radial harmonic oscillator _ it contains an additional centrifugal potential term ( edwards gulyaev or langer term @xcite ) , which emerges from bessel function @xmath113 present in the finite sliced form of the radial pi ( [ diaglapl ] ) . at this point we should stress that in contrast to the quantum - mechanical radial pi , one can use safely the asymptotic expansion for the bessel function @xmath113 ( see , e.g. ref . @xcite ) : @xmath114 > 0)\ , , \nonumber \\[2 mm ] & & y_j \ = \ ( m/\varepsilon \hbar)r_jr_{{j-1}}\ , , \label{asymptot - b}\end{aligned}\ ] ] in the euclidean pi sliced form . here the infinitesimal `` time '' slice @xmath115 is related to the number of slices @xmath116 via the relation @xmath117 . in quantum mechanic this is a problematic step because ( [ asymptot - b ] ) requires @xmath118 > 0 $ ] while there @xmath118 = \mbox{re}[(m / i\varepsilon \hbar)r_jr_{{j-1 } } ] = 0 $ ] . fortunately , the pi for radial harmonic oscillator is exactly solvable even in the case of @xmath0-dependent oscillator frequency . the solution reads @xcite @xmath119 \right\ } \nonumber \\[2 mm ] & = & \ \frac{\hbar^2}{m } \frac{\sqrt{\eta_1 \eta_2}}{g(x_1 ) } i_{d/2 - 1}\left(\frac{\hbar^2}{m } \frac{\eta_1 \eta_2}{g(x_1)}\right ) \exp\left[- \frac{\hbar^2}{2 m } \left(\frac{f'(x_2)}{f(x_2 ) } \eta_2 ^ 2 - \frac{g'(x_1)}{g(x_1 ) } \eta_1 ^ 2 \right ) \right ] . \label{radialhosol}\end{aligned}\ ] ] the functions @xmath120 and @xmath121 are two independent solutions of the differential equation @xmath122y(x ) \ = \ \left [ -\frac{\hbar^2}{2 m } \frac{d^2}{dx^2 } + v(x ) + e \right ] y(x ) \ = \ 0\ , , \ ] ] with the initial conditions @xmath123 and @xmath124 , and @xmath125 and @xmath126 . in addition , the wronskian @xmath127 is independent of @xmath0 , as can be proved by differentiation and by using the fact that @xmath128 and @xmath129 both satisfy eq . ( [ diffeqfg ] ) . by equating the values of @xmath130 at points @xmath131 and @xmath132 , and taking into account the initial conditions for @xmath128 and @xmath129 , we find a useful identity @xmath133 . now , the pi in ( [ diaglapl ] ) can be sliced at point @xmath33 , and expressed as @xmath134 the limits in eq . ( [ diaglapl ] ) are readily carried out with the help of the asymptotic formulas @xmath135 , and @xmath136 , valid for @xmath137 . subsequent integration over @xmath138 brings ( [ diaglapl ] ) to form @xmath139 where @xmath140 solves eq . ( [ diffeqfg ] ) with initial conditions @xmath141 and @xmath142 , and @xmath143 solves the same equation with @xmath144 and @xmath145 . the denominator in ( [ diaggreenexpl ] ) is the wronskian @xmath146 . the full derivation is given in appendix [ sec : app ] . although rather explicit , eq . ( [ diaggreenexpl ] ) is not well suited for the laplace transform inversion , since functions @xmath120 and @xmath121 contain @xmath77 in a non - trivial way , which , in addition , significantly hinges on the actual form of @xmath50 . for formal manipulations it is still better the employ the pi representation ( [ diaglapl ] ) . for instance , using eq . ( [ inverselap ] ) we can easily invert the laplace transform to return from @xmath77 back to the @xmath5-variable , namely @xmath147 } \ , . \label{boltzmann.ii}\end{aligned}\ ] ] note that we have utilized the asymptotic form @xmath110 which holds for @xmath148 . we shall see shortly that ( [ boltzmann.ii ] ) can be straightforwardly related to the local - time pi representation of the boltzmann density matrix . let us finally comment on the higher - angular - momentum terms which , as claimed , should not contribute to the expression ( [ diaglapl ] ) . for arbitrary angular momentum @xmath149 , we employ formula ( [ radialhosol ] ) with a slight modification @xmath150 . now , for example , in the limit @xmath151 , this goes like @xmath152 , which , multiplied by the prefactor @xmath153 , implies the behavior @xmath154 . that is , only the ( @xmath155)-term can give a non - vanishing contribution . consider again eq . ( [ lapdifeq ] ) and a finite interval @xmath156 $ ] . the corresponding green function of the operator @xmath157 can be easily constructed ( at least formally ) with the help of the sturm liouville theory @xcite . an immediate consequence of the latter is that for @xmath158 the green function has the form @xmath159 where the functions @xmath120 and @xmath121 satisfy eq . ( [ diffeqfg ] ) with the initial conditions @xmath160 and @xmath161 , and @xmath162 and @xmath163 , respectively . in addition , the above green function should be symmetric due to the hermitian nature of @xmath164 . the sturm liouville theory ensures that the solution to the second - order differential equation ( [ diffeqfg ] ) is unique , when specifying the values of @xmath165 and @xmath166 at some point @xmath167 . therefore , the functions @xmath128 and @xmath129 must coincide with @xmath168 and @xmath169 of eq . ( [ diaggreenexpl ] ) , and the diagonal part of ( [ sturmgreen ] ) , i.e. @xmath170 , reduces to expression ( [ diaggreenexpl ] ) . this is an important consistency check of our representation ( [ diaglapl ] ) . let us now generalize the pi representation ( [ diaglapl ] ) to the full bloch density matrix , i.e. , we wish to include also the off - diagonal matrix elements , @xmath67 . if we go back to the replica representation ( [ afterrepl ] ) , we realize that the requirement @xmath67 spoils rotational symmetry in the replica field space , and thus precludes straightforward reduction to a radial path integral . instead of refining the reduction procedure , we simply make a guess , which , as we prove in appendix [ sec : app ] , coincides with the well - established sturm liouville formula ( [ sturmgreen ] ) . our guess is based on mathematical results presented in @xcite . in particular , we claim that the extension of the representation ( [ diaglapl ] ) to off - diagonal matrix elements should read @xmath171 } \ , , \ ] ] with the action functional @xmath172 \ \equiv \ \mathcal{a}^e[\eta ] \ + \ \int_{x_-}^{x_+ } \!\ ! dx \ \ ! \eta^2(x)}\ , , \ ] ] where @xmath173 $ ] is defined in eq . ( [ actionho ] ) , and @xmath174 is a piecewise constant function @xmath175 \\ ( d-1)(d-3 ) & { \rm otherwise } . \end{cases}\ ] ] at this point we can invert the laplace transform with the help of eq . ( [ inverselap ] ) . as a result , we obtain the sought local - time pi representation of the bloch density matrix ( [ rhodef ] ) , namely @xmath176 } \ , .\ ] ] here , integrations over @xmath177 run from @xmath178 to @xmath179 , i.e. , the paths @xmath177 are non - negative . comparing this result with the anticipated heuristic form ( [ heurform ] ) , we can identify @xmath180 . representation ( [ offdiag ] ) allows us to identify the weight factor ( [ iii.7.a ] ) with @xmath181 \ = \ \frac{2}{d^2 } ( \eta_- \eta_+)^{\frac{1-d}{2 } } \eta(x_a ) \eta(x_b ) \exp \left\ { -\int_{x_-}^{x_+ } \!\!\ ! dx \left[\frac{\hbar^2}{2 m}{\eta'}^2 + \frac{m}{\hbar^2}\frac{\delta(x)}{8 \eta^2 } \right ] \right\ } .\ ] ] contrary to expectation , the right - hand - side of this expression does not depend on @xmath5 . sub - index @xmath89 in @xmath182 indicates that the weight factor must be regularized when we pull it out of the pi ( [ offdiag ] ) . by analogy with quantum mechanics one can represent ( [ offdiag ] ) in the discretized time - sliced form . in such a case the weight @xmath182 would be a product of terms involving the bessel functions @xmath113 , if @xmath183 , or @xmath184 , if @xmath185 ( see ref . @xcite ) . last but not least , expressions ( [ offdiag])-([weights ] ) indicate that the square root of @xmath3 is ( at least from physicist s point of view ) more convenient variable to describe local - time trajectories than @xmath3 alone . formula ( [ offdiag ] ) provides a way of rewriting the pi ( [ pix ] ) in terms of the local time . in this section , we consider more general scenario , in which the initial path integral is of the form @xmath186 \ \ ! \exp\left\{-\frac{1}{\hbar } \int_{0}^{\beta\hbar } \!\!\!d\tau \left[\frac{m}{2}\dot{x}^2 + v(x ) \right ] \right\ } , \ ] ] where @xmath128 is an arbitrary functional of the local time @xmath187 , which itself is ( as seen in section [ sec : heuristic ] ) a functional of the paths @xmath20 . relation ( [ pifunc ] ) represents a local - time analog of the feynman salam formula @xcite . to bring it into more manageable form , we may observe that for any @xmath2 , the action of @xmath3 in the pi ( [ pifunc ] ) can be taken over by the functional derivative @xmath188 , acting on the exponential . this becomes transparent after rewriting the potential part as @xmath189 . the entire functional @xmath190 $ ] can be therefore pulled out of the path integral , which then allows to write @xmath191 \rho(x_a , x_b,\beta ) \ , .\ ] ] when we employ the local - time representation of pi for @xmath12 ( cf . ( [ offdiag ] ) ) , each functional derivative @xmath192 will produce the term @xmath193 . in such a way @xmath194 can be written as @xmath195 e^ { - \mathcal{a}^{e=0}_\delta[\eta ] } \ , , \label{ltfunc}\end{aligned}\ ] ] where , strictly speaking , the functional @xmath196 $ ] is regularized in such a way that it depends on @xmath3 only for @xmath197 $ ] , and @xmath198 are sent to @xmath199 only at the end of the calculation . first , let us make the simple observation that the formula ( [ ltfunc ] ) reduces to ( [ offdiag ] ) for the choice @xmath190 = 1 $ ] . one of the most important mean values of a local - time functional , as evaluated with eq . ( [ ltfunc ] ) , is the mean of @xmath200 which gives the moment - generating functional . the local - time moment structure is particularly pertinent in various perturbative expansions , including low- and high - temperature expansions ( see section [ sec : asymp ] ) . another important example , namely the case of a one - point distribution function will be discussed in the following subsection . in passing we should note , that should we have started from ( [ replrepre ] ) and repeated the above procedure , an analog of eq . ( [ ltfunc ] ) for higher - dimensional spaces , @xmath99 , could be easily obtained . this would include the @xmath89-dimensional replica field @xmath107 in the @xmath201-dimensional euclidean configuration space . simple , though quite important consequence of eq . ( [ ltfunc ] ) is that it readily provides the @xmath116-point distribution functions of the local time . this is achieved when we set @xmath190 = \prod_{n=1}^n \delta ( l^{x_n } - l_n ) $ ] . in order to see what is involved let us now illustrate the calculation for @xmath202 ( with @xmath203 ) . our discussion will be greatly simplified by considering only a free particle ( i.e. , @xmath204 ) that starts and ends at the origin , i.e. , @xmath205 . this corresponds to a stochastic process known as brownian bridge . our goal is to derive the one - point distribution function , denoted @xmath206 , of the local time at @xmath207 . we define @xmath206 by eq . ( [ pifunc ] ) with @xmath190=\delta ( l^0 - l ) $ ] , and calculate it from the representation ( [ ltfunc ] ) as follows . in the laplace picture , @xmath208 , the path integral ( [ ltfunc ] ) can be sliced at @xmath209 so that @xmath210 the @xmath211-integration can be done easily by realizing that @xmath212 . furthermore , the limits in @xmath111 and @xmath89 can be carried out with the help of formulas ( [ etalimminus ] ) and ( [ etalimplus ] ) from appendix [ sec : app ] . consequently , we obtain @xmath213 , \ ] ] where , for the free - particle case , @xmath214/\sqrt{2me/\hbar^2}$ ] , and @xmath215/\sqrt{2me/\hbar^2}$ ] , as one can straightforwardly verify . in the limit @xmath216 , ( [ ex1 ] ) reduces to @xmath217 and its inverse - laplace transform yields @xmath218 we stress that @xmath206 thus obtained is , in fact , the ( unnormalized ) _ joint _ probability density for stochastic events @xmath219 " and @xmath220 " . by bayes theorem of the probability calculus , the desired conditional probability density @xmath221 $ ] is obtained from ( [ ltfree ] ) by dividing @xmath206 by the brownian - bridge probability density @xmath222 $ ] , which is ( omitting again normalization ) @xmath223 ( see , e.g. , ref . normalization factors mutually cancel in the fraction and we arrive at @xmath224 \ = \ \frac{\hbar^2 l \exp\left(- \frac{l^2 \hbar^2}{2\beta m } \right)}{{\beta m}}\ , , \end{aligned}\ ] ] which is clearly normalized to @xmath225 . one can proceed along the same lines also in more complicated higher - dimensional ( @xmath226 ) cases . our result agrees with the one found through other means in ref . a compelling feature of the local - time representation ( [ offdiag ] ) is that it naturally captures both small- and large-@xmath5 asymptotic regimes . this should be compared with the feynman kac pi representation ( [ pix ] ) , which is typically suitable only for the small-@xmath5 ( i.e. , large - temperature ) analysis . the latter is epitomized either by wkb approximation @xcite or wigner kirkwood expansion @xcite . in the large-@xmath5 ( small - temperature ) limit , the spectral representation of the gibbs operator , @xmath227 , reduces the bloch density matrix to the ground - state contribution @xmath228 that is not evident from the feynman kac pi representation @xcite . in connection with eq . ( [ spectralasymp ] ) it is useful to remind that in @xmath229 the discrete bound states can all be chosen to be real @xcite , so that the bloch density matrix is real and symmetric and can be written in the form @xmath230 let us first comment on the small-@xmath5 regime of the local - time representation , assuming @xmath66 for simplicity . this case was discussed in detail in our previous article @xcite . there one should first taylor - expand the potential @xmath50 around the point @xmath33 , and then expand the exponential part containing the structure @xmath231 , where @xmath232^m\ , .\end{aligned}\ ] ] after the term @xmath233 is factored out of the integral , the individual summands of the ensuing series are of the form ( [ ltfunc ] ) with the potential @xmath204 , and functional @xmath190 \propto \prod_n l^{x_n}/\hbar$ ] . the latter can be related to the power expansion in @xmath5 presented in @xcite through the equality of representations ( [ pifunc ] ) and ( [ ltfunc ] ) . the whole bloch density matrix ( containing also off - diagonal elements ) can be treated similarly in a full analogy with ref . @xcite . the large-@xmath5 expansion of eq . ( [ offdiag ] ) can be conveniently studied after rescaling @xmath234 , in which case we can write @xmath235 & \times & \ \delta \ ! \left ( \int_{x_-}^{x_+ } \!\!\ ! \eta^2 dx - 1 \right ) \exp \left\ { -\int_{x_-}^{x_+ } \!\!\ ! dx \left[\frac{\beta \hbar^2}{2 m}{\eta'}^2 + \beta v(x)\eta^2 + \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2 } \right ] \right\ } \nonumber \\[3 mm ] & = & \ \lim_{d \rightarrow 0 } \frac{2}{d^2 } \lim_{\eta_\pm \rightarrow 0 } ( \eta_- \eta_+)^{\frac{1-d}{2 } } \frac{\beta \delta^2}{\delta j(x_a ) \delta j(x_b ) } \int_{c - i\infty}^{c+i\infty } \frac{d \kappa } { 2\pi i } \int_{\eta(x_-)=\eta_-}^{\eta(x_+)=\eta_+ } \mathcal{d}\eta(x)\nonumber\\[1 mm ] & \times & \ \exp\left\ { - \beta \left[\int_{x_-}^{x_+ } \!\!\ ! dx \left(\frac{\hbar^2}{2 m}{\eta'}^2 + v(x)\eta^2 - \kappa \eta^2\right ) + \kappa \right ] \right\}\nonumber\\[1 mm ] & \times & \ \left . \exp\left\ { - \int_{x_-}^{x_+ } \!\!\ ! dx \left [ \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2 } + j\eta \right ] \right\}\right|_{j=0 } , \end{aligned}\ ] ] where @xmath236 is an arbitrary real number . with the _ method of images _ @xcite we can rewrite the radial pi involved as a superposition of two genuine one - dimensional pi s @xcite @xmath237 - \langle j|\eta \rangle \right\ } \exp\left [ - \int_{x_-}^{x_+ } dx\ \ ! \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2}\right]_{\!r}\nonumber \\[3 mm ] & & \mbox{\hspace{10mm}}= \ \int_{\eta(x_-)=\eta_-}^{\eta(x_+)=\eta_+ } \mathcal{d}\eta(x ) \exp\left\{-\beta\left[\langle \eta| \hat{h}| \eta \rangle -\kappa\left(\langle \eta|\eta \rangle -1 \right ) \right ] - \langle j||\eta| \rangle \right\}\nonumber \\[1 mm ] & & \mbox{\hspace{14 mm } } \times \ \exp\left [ - \int_{x_-}^{x_+ } dx\ \ ! \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2}\right]\nonumber \\[1 mm ] & & \mbox{\hspace{10 mm } } -\ \cos(\pi d/2 ) \int_{\eta(x_-)=-\eta_-}^{\eta(x_+)=\eta_+ } \mathcal{d}\eta(x ) \exp\left\{-\beta\left[\langle \eta| \hat{h}| \eta \rangle -\kappa\left(\langle \eta|\eta \rangle -1 \right ) \right ] - \langle j||\eta| \rangle \right\ } \nonumber \\[1 mm ] & & \mbox{\hspace{14 mm } } \times \ \exp\left [ - \int_{x_-}^{x_+ } dx\ \ ! \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2}\right ] , \label{method - images-1aa}\end{aligned}\ ] ] where dirac s notation was employed . a few comments are in order about the right - hand - side of the above relation . first , it should be noticed the presence of the parity - even terms @xmath238 in pi s . second , pi s differ by their respective dirichlet boundary conditions . third , the super - index `` @xmath239 '' was used to stress the restricted nature of the fluctuations in the radial pi measure while the measure without `` @xmath239 '' represents a usual one - dimensional pi measure , i.e. @xmath240 here `` @xmath241 '' denotes de witt s `` equivalence '' symbol @xcite . finally , the correct time - sliced form of the exponential with the centrifugal potential is ( cf . e.g. , refs . @xcite ) @xmath242_{\!r } \ \lim_{n\rightarrow \infty}\prod_{k=1}^n \sqrt{2\pi \frac{\beta \hbar^2}{m } \frac{\eta_k\eta_{k-1}}{\varepsilon } \tilde{\delta}_k } \ \ ! \exp\left(-\frac{\beta \hbar^2}{m } \frac{\eta_k\eta_{k-1}}{\varepsilon } \tilde{\delta}_k \right ) i_{\frac{d-2}{2}}\left ( \frac{\beta \hbar^2}{m } \frac{\eta_k\eta_{k-1}}{\varepsilon } \tilde{\delta}_k\right ) , \nonumber \\[2 mm ] & & \exp\left [ - \int_{x_-}^{x_+ } dx\ \ ! \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2}\right ] \ \lim_{n\rightarrow \infty}\prod_{k=1}^n \psi_{\frac{d-2}{2}}\left(-\frac{\beta \hbar^2}{m } \frac{|\eta_k\eta_{k-1}|}{\varepsilon } \tilde{\delta}_k\right ) , \label{centrifugal - slicing-1a}\end{aligned}\ ] ] with @xmath243 \\ 1 & { \rm otherwise } , \end{cases}\end{aligned}\ ] ] and ( see , e.g. , refs . @xcite ) @xmath244 \ = \ e^x \sqrt{\frac{2 x } { \pi } } k_p(x)\ , . \end{aligned}\ ] ] [ @xmath245 and @xmath246 are the modified bessel functions of the first and the second kind , respectively . ] in cases when @xmath247 meaning that @xmath248 ( e.g. , `` typical situation '' for very fine time slicings ) the asymptotic form of @xmath249 holds . with help of the preceding asymptotic behavior one obtains @xmath250\nonumber \\[1 mm ] & \sim & \ \prod_{k=1}^n \exp\left[- \frac{m}{\beta \hbar^2 } \frac{\delta_k}{8|\eta_k\eta_{k-1}| } \ \!\varepsilon + { \mathcal{o}}(\varepsilon^2 ) \right]\nonumber \\[1 mm ] & \sim & \ \exp\left [ - \int_{x_-}^{x_+ } dx\ \ ! \frac{m}{\beta \hbar^2}\frac{\delta(x)}{8 \eta^2}\right ] . \label{48.aa}\end{aligned}\ ] ] potential singularities of the integral at @xmath251 can be regularized , e.g. , by a principal value prescription . unfortunately , the formula ( [ 48.aa ] ) can not be directly used in our case because the boundary values @xmath252 and @xmath253 are arbitrarily close to zero , and hence the assumed asymptotic behavior for @xmath254 is not fulfilled . this situation can be rectified by factorizing out the problematic boundary points as @xmath255 & & \sim \ \frac{1}{\pi}\left(\frac{\beta \hbar^2}{2 m \varepsilon}\right)^{d-1 } \left(|\eta_-\eta_{1}||\eta_+ \eta_{_{n-1}}|\right)^{(d-1)/2}\ \ ! \prod_{k=2}^{n-1 } \exp\left[- \frac{m}{\beta \hbar^2 } \frac{\delta_k}{8|\eta_k\eta_{k-1}| } \ \!\varepsilon + { \mathcal{o}}(\varepsilon^2 ) \right ] . \ ] ] here we have utilized the asymptotic form @xmath256 valid for @xmath257 and @xmath258 ( @xmath259 ) . the passage from the radial pi ( [ offdiagrescaled ] ) to the ordinary ( 1-dimensional ) pi brings about an important advantage , namely , one can perform the wkb approximation . in particular , one can use the laplace s formula of the asymptotic calculus @xcite @xmath260 \ = \ \sqrt{\frac{2\pi}{\beta g''(t_0 ) } } \ \ ! f(t_0 ) \ \ ! \exp\left[-\beta g(t_0 ) \right ] \ + \ { \mathcal{o}}\left ( \frac{\exp\left[-\beta g(t_0 ) \right]}{\beta^{3/2 } } \right ) , \label{laplace}\end{aligned}\ ] ] with @xmath261 being a solution of @xmath262 ( provided @xmath263 has a smooth absolute minimum at the interior point @xmath264 ) . in case of need , the full asymptotic expansion can be found , e.g. , in ref . @xcite . in our case the customary pi substitution @xmath265^{-1/2}\ , , \label{vi.49a}\end{aligned}\ ] ] should be employed @xcite . here @xmath266 is the ground state energy and the difference operators ( lattice derivatives ) are defined as @xcite @xmath267\ , , \;\;\;\;\ ; \overline{\nabla } \eta(x ) \ = \ \frac{1}{\varepsilon}\left[\eta(x ) \ - \ \eta(x-\varepsilon ) \right]\ , , \end{aligned}\ ] ] with @xmath268 being the hermitian operator on the space of `` time - sliced '' functions with vanishing end points , i.e. , @xmath269 . the ensuing determinant @xmath270 in ( [ vi.49a ] ) can be conveniently computed by means of the gelfand yaglom method @xcite ( or shifting method @xcite ) . there the determinant @xmath271 where @xmath272 is finite in the @xmath273 limit and fulfills the differential equation @xmath274 d(x ) \ = \ 0\ , , \label{vi.51a}\end{aligned}\ ] ] with the cauchy conditions @xmath275 and @xmath276 . the solution @xmath272 can be written as a linear combination of two independent solutions of ( [ vi.51a ] ) . one is immediate , namely the wkb solution @xmath277 . a second solution can be constructed with the help of dalambert s formula directly from @xmath277 in the form ( cf . e.g. , ref . @xcite ) @xmath278 it is simple to verify that ( [ vi.52a ] ) is again a solution of ( [ vi.51a ] ) . the cauchy conditions in the gelfand yaglom method uniquely fix the constants in front of the two independent solutions so that the full solution @xmath272 acquires the form @xmath279 as a result we obtain in the large @xmath116 ( i.e. , small @xmath115 ) limit @xmath280 \ \stackrel{\varepsilon \rightarrow 0}{\longrightarrow } \ \frac{\eta_+ \eta_-}{\varepsilon } \int^{x_+}_{x_- } \frac{dx}{\eta_0 ^ 2(x)}\ , . \label{vi.55aa}\end{aligned}\ ] ] the great advantage of the gelfand yaglom method is that it does not require any detailed knowledge of the spectrum of the operator whose determinant is being computed . it also shows that the determinant in ( [ vi.49a ] ) diverges linearly with @xmath116 in the large @xmath116 limit . we now substitute for @xmath281 in ( [ laplace ] ) the functional expression @xmath282 \ , , \end{aligned}\ ] ] where , @xmath277 comes from the path that minimizes the functional @xmath283 . according to the rayleigh ritz variation principle ( see , e.g. , refs . @xcite ) , such a function @xmath177 is the ground - state wavefunction of the hamiltonian @xmath164 , i.e. , @xmath284 with @xmath285 . with a hindsight , we have used the latter value of @xmath286 already in eqs . ( [ vi.49a ] ) , ( [ vi.51a ] ) and ( [ vi.55aa ] ) . notice that the stationary point in @xmath286 is real but the integration contour in @xmath286 is parallel to the imaginary axis . both reality and positivity of @xmath177 pose no restriction in the rayleigh ritz principle , because the ground state can always be chosen real and positive @xcite . putting it all together , we get the leading large-@xmath5 behavior in the form @xmath287 as it should be ( cf . ( [ spectralasymp2 ] ) ) . we conclude the discussion of low - temperature expansion by noting that the rayleigh ritz variation principle states that _ all _ eigenvalues and ( normalized ) eigenvectors of @xmath164 come from stationary solutions of @xmath288 , and conversely @xcite . in the spirit of the wkb approximation one should sum over all path integrals evaluated about _ all _ stationary solutions . it is , however , only the ground state configuration @xmath289 that acquires the global minimum and which gives the larges contribution to the wkb approximation . this fact was implicitly used in our preceding reasonings . should we have included also other stationary solutions we would recover higher order terms in the spectral expansion of the bloch density matrix ( [ spectralasymp2 ] ) . in this paper , we have derived the local - time pi representation of the bloch density matrix . we have shown that the result obtained , apart from being of interest in pure mathematics ( stochastic theory , sturm liouville theory , etc ) , can serve as a useful alternative to the traditional feynman kac pi representation of green functions of fokker planck equations . furthermore , by analytically continuing the result back to the real time via the inverse wick rotation , @xmath290 , one obtains the local - time pi representation of quantum - mechanical transition amplitudes , i.e. , matrix elements of the evolution operator @xmath291 . from the physicists point of view , perhaps the most important application of local - time pi s lies in statistical physics , and namely in the low and high temperature treatments of the bloch density matrix . this is because in conventional pi s only a very tiny subset of paths gives a relevant contribution in these asymptotic regimes . in particular , the high - temperature regime of the boltzmann density function @xmath292 is dominated by paths that spend a sizable amount of time in the vicinity of the point @xmath0 . similarly , the low - temperature regime is controlled by paths with a large local time near the global minimum of the potential . here we have exemplified the conceptual convenience of the local - time formulation by providing a generic analysis of the low - temperature behavior of the bloch density matrix . our formulation proved to be particularly instrumental in obtaining the correct asymptotic behavior ( known from the spectral theory ) which is otherwise notoriously difficult to obtain within the feynman kac pi framework @xcite . as a byproduct we have uncovered an interesting connection between a low - temperature pi expansion , the gelfand yaglom formula and the rayleigh ritz variational principle . in order to further reinforce our analysis , we formulated a local - time analog of the feynman matthews salam formula which is ( similarly as its qft counterpart ) expedient in number of statistical - physics contexts . the prescription obtained was substantiated by an explicit calculation of a one - point distribution function of the local time . it appears worthwhile to stress that our local - time representation ( with its build - in replica field trick ) is in its present form applicable only to one - dimensional quantum mechanical systems . with a hindsight we reflected this fact already in our choice of the incipient pi ( [ pix ] ) where we assumed @xmath293 and @xmath294 . though one may easily proceed up to eq . ( [ replrepre ] ) without any restriction on the value of @xmath201 ( in fact , eq . ( [ replrepre ] ) is valid for any @xmath99 with @xmath295 ) , a further progress in this direction is hindered by the fact that the replica fields depend on a @xmath201-dimensional argument @xmath0 , and thus the pi in ( [ replrepre ] ) can no longer be regarded as a quantum mechanical pi ( i.e. , pi over fluctuating paths ) . in effect , we can not use existing mathematical techniques of the pi calculus ( e.g. , transformation of pi s to polar coordinates ) , that we have employed to get the radial pi ( [ diaglapl ] ) . the issue of the extension of our local - time pi representation to higher - dimensional configuration space is currently under active investigation . the authors acknowledge a. alastuey , h. kleinert , z. haba and d.s . grebenkov for fruitful discussions . the work has been supported by the gar grant no . ga14 - 07983s . was also partially supported by the ctu in prague grant no . sgs13/217/ohk4/3t/14 and by the dfg grant : kl 256/54 - 1 . in this appendix we show that the representation ( [ laploffdiag ] ) reduces to the well established result ( [ sturmgreen ] ) of the sturm liouville theory . since @xmath83 is symmetric in @xmath33 and @xmath34 , we will assume , without loss of generality , that @xmath296 . the path integral in ( [ laploffdiag ] ) can be expressed via eq . ( [ radialhosol ] ) as @xmath297 the limits in eq . ( [ laploffdiag ] ) can be carried with the help of the asymptotic formulas @xmath135 , and @xmath136 , valid for @xmath298 . we obtain @xmath299 & & \xrightarrow{d \rightarrow 0 } \ \frac{1}{\sqrt{\eta_a } } \exp \left ( -\frac{\hbar^2}{2 m } \frac{f'_1(x_a)}{f_1(x_a ) } \eta_a^2 \right)\ ! , \end{aligned}\ ] ] where @xmath140 satisfies eq . ( [ diffeqfg ] ) with initial conditions @xmath300 and @xmath301 , and similarly , we find @xmath302 where @xmath303 satisfies eq . ( [ diffeqfg ] ) with initial conditions @xmath304 and @xmath305 . formula ( [ laploffdiag ] ) then reduces to @xmath306 & & \times \ \ ! \exp \left [ - \frac{\hbar^2}{2 m } \left ( \frac{w(g_2,f_1 ) \eta_a^2}{f_1(x_a ) g_2(x_a ) } + \frac{w(g_3,f_2 ) \eta_b^2}{f_2(x_b ) g_3(x_b ) } \right ) \right]\ , , \end{aligned}\ ] ] where @xmath307 and @xmath143 satisfy eq . ( [ diffeqfg ] ) with initial conditions @xmath308 and @xmath309 , and @xmath310 and @xmath311 , respectively . the wronskian @xmath127 is independent of @xmath0 , as discussed in section [ sec : radial ] , and antisymmetric , i.e. , @xmath312 . wronskians @xmath313 and @xmath314 assume a particularly simple form when evaluated at points @xmath34 and @xmath33 , respectively , due to the initial conditions satisfied by @xmath169 and @xmath315 . we find @xmath316 and @xmath317 . rescaling @xmath318 , @xmath319 , and using the relation @xmath320 we obtain @xmath321 the integrations are readily performed using the formula @xcite @xmath322 yielding @xmath323 to prove equality with ( [ sturmgreen ] ) , we only have to show that @xmath324 this is done by realizing that @xmath143 , being a solution of the second - order linear differential equation ( [ diffeqfg ] ) , can be uniquely composed as a linear combination of two other solutions @xmath140 and @xmath303 , @xmath325 indeed , thus defined @xmath169 satisfies the initial conditions @xmath310 and @xmath311 . the low temperature expansion is typically obtained indirectly via duality approaches . among these , a particularly powerful non - perturbative approximation scheme is the so - called _ variational _ or _ optimized _ perturbation theory @xcite . f. steiner , _ path integrals in polar coordinates from ev to gev _ , in _ path integrals from mev to mev , bielefeld , 1985 _ , edited by m.c . gutzwiller et al . ( world scientific , singapore , 1986 ) pp . 335 - 359 .

when a commensurate one - dimensional charge density wave ( cdw ) is doped , the resulting charges may be viewed as defects in the cdw . if the interactions between them are sufficiently strong compared to their effective kinetic energy , then they will form an ordered lattice ( which can be stabilized by a crossover to 3d order if there are many @xmath2 chains coupled together ) of defects and the cdw will become incommensurate with the lattice . this can be more energetically favorable than simply doping holes into the rigid band formed in the presence of cdw order at a fixed wavevector . the reason is that , in the latter case , we must pay the single - particle energy gap of the cdw . in a two - dimensional @xmath0-density wave ( ddw ) state , however , there are nodes in the order parameter , so it is relatively painless for doped holes to simply go to the nodes , which expand into fermi pockets ( in quasi-@xmath2 systems , however , this might not be possible , so doping must lead to incommensurability @xcite ) . however , this can not continue indefinitely if the ddw state is stabilized by approximate nesting , since the fermi surface ( fs ) would eventually move away from the nesting wavevector . thus , incommensurate ddw order is a strong possibility , at least over some range of hole dopings . in this paper , we explore this possibility . near half - filling , the hubbard model , the @xmath3 model , and generalizations of these appear to have many phases which are nearby in energy @xcite . thus , small perturbations can strongly influence the competition between them . consequently , it should not be too surprising that experiments on the cuprate superconductors have also uncovered evidence for a cornucopia of phases , particularly on the underdoped side of the phase diagram , which appear in particular materials and for certain values of the doping level , temperature , magnetic field , etc . @xcite . depending on the interfacial energies between these phases , one way in which their competition can be resolved is through the formation of stripes @xcite or other inhomogeneous patterns of microscale phase - separation . neutron scattering and scanning tunneling microscopy experiments point towards this possibility@xcite . furthermore , as the cuprates are doped , their fermi surfaces evolve away from the commensurate nesting wavevector @xmath4 , so we would expect translational symmetry - breaking order parameters to occur at incommensurate wavevectors . hence , it would seem important for any theory of the pseudogap to incorporate the tendency towards incommenuration , which we take as further impetus to study incommensurate ( ic ) order parameters related to ddw order . in this paper we consider various ways in which @xmath0-density wave ( ddw ) order can go incommensurate . before turning to this discussion , we review the reasons for expecting ddw order to occur in the pseudogap regime of the cuprates . it has been suggested@xcite that competition between ddw order and @xmath5-wave superconductivity ( dsc ) can explain many of the anomalous properties of the pseudogap@xcite . the ddw order parameter is given by @xmath6 with @xmath7 and a _ commensurate _ ( c ) ordering wavevector @xmath8 . the particle - hole condensate in breaks translation by one lattice spacing , parity , and time - reversal but is invariant under the combination of any two of the above symmetries . in real space it is represented by currents which alternate from one plaquette to the next . one motivation for considering ddw order is to explain @xcite the abrupt depletion of the superfluid density below the critical @xmath9 in bi2212 and similar dopings in other materials @xcite . because the order has a @xmath5-wave momentum dependence , its onset would explain arpes measurements above @xmath10 in the pseudogap @xcite . furthermore , since ddw is a spin - singlet condensate , it would also explain the suppression of the spin susceptibility in nmr measurements and opening of a spin gap in inelastic neutron scattering @xcite . also , the depletion of the c - axis conductivity @xcite , whose contribution mostly comes from the antinodal regions , would also be naturally explained by the development of ddw order , which gaps these regions . direct attempts to measure ddw order through neutron scattering have neither ruled it out @xcite nor unambiguously verified its presence , but there are intriguing suggestions that it may be present@xcite . thus , it is natural to ask whether ddw order is compatible with incommensurability . ( for a discussion of the experimental signatures of ddw order , the reader is referred to the literature @xcite . the case of ddw in @xmath11 is discussed in@xcite . for numerical evidence of the existence of ddw in other systems , one is referred to@xcite . ) the plan of our paper is as follows . in section [ susc ] we develop some preliminary intuition about the energetics of ic density wave order by analyzing the ddw susceptibility of a fermi liquid ( both the nested and non - nested cases ) . in section [ c - ic ] we review some of the theory of commensurate ( c ) ddw order and show various ways to make it incommensurate . we also suggest the kind of microscopic physics that could lead to c and ic ddw , and in addition discuss possible experimental signatures of ic ddw . finally , in section [ num ] , we compute the mean - field phase diagrams of an extended hubbard model . our main results are that ( a ) ic ddw develops for a wide range of parameters , but ( b ) the ic wavevectors can be very small . their relevance to the cuprates may , at most , be for a rather narrow set of dopings . we also offer a physical explanation of our results , which sheds some light on the nature of c - ic ddw transition in extended hubbard models . we note that similar incommensurate order has been found in refs . the easiest way to see that ic ddw phases could arise in effective models ( which , in this paper , will , in turn , emerge from microscopic extended hubbard models with correlated hopping terms@xcite ; see section [ micro ] for details ) is to consider the bare susceptibility as a function of the incommensurability wavevector @xmath12 , where @xmath13 : @xmath14 where @xmath15 is defined after eq . [ cop ] . here , for simplicity , ddw is considered without the presence of any other order . in eq . [ suscep ] , @xmath16 is the fermi occupation number , and the energy dispersion contains nearest , @xmath17 , and next - nearest neighbor , @xmath18 , hopping parameters , @xmath19 . the ddw equivalent of the stoner criterion is satisfied when the right - hand - side of this equation approaches the inverse of the pertinent coupling @xmath20 . this observation allows us to interpret the positions of the peaks of @xmath21 as the wavevectors at which ddw order is likely to occur as we lower the temperature . a similar analysis was done by schulz@xcite for the case of af and ic sdw order in the hubbard model close to half - filling , where perfect nesting at half - filling was asumed . in fact , is at nonzero @xmath12 . here , @xmath22 is measured along the zone diagonal in units of @xmath23 , where @xmath24 is the lattice spacing.,height=326 ] his results can be recovered by letting @xmath25 in if we set @xmath26 . the support of @xmath27 in the integrand in is at @xmath28 and symmetry - related antinodal points of the fs since this is where the ddw order parameter is large . however , since these points are also the location of van hove singularities , the regions around them also dominate the af susceptibility . thus , at least in the @xmath26 case , one expects that whenever an ic sdw instability occurs , an ic instability for ddw should also occur . indeed , for both orders , the peak of @xmath21 occurs at @xmath29 for low enough temperatures and sufficiently away from half - filling , e.g. if @xmath30 , as shown in fig . [ suscfig ] . note that the same analysis repeated with any other density wave or spin - density wave order would yield an ic instability as long as the dominant channels of the modulation coincide with fermi surface points which give large contributions to the susceptibility . for instance , a @xmath31-density wave@xcite with wavevector @xmath32 has a peak in @xmath33 for @xmath34 away from half filling at @xmath35 , in the case @xmath26 . a natural question to ask is : to what extent does the ic ddw instability mimic the ic sdw one ? to answer this , one needs to consider specific models as we do in [ micro ] . however , two differences can be anticipated even beforehand . first , the ddw order parameter has nodes , while the af order parameter does not . second , in a more realistic model , one must consider competition with dsc , which also could change features of the transition to ic state . a better understanding of ic ddw order parameter is necessary to better understand the the situation away from half - filling . given that the susceptibility points towards an ic instability , the next issue is what ic ddw order would look like . before addressing this question , let us first summarize some features of the c - case.@xcite in the c case , a general singlet particle - hole condensate can be written as ( neglecting the @xmath36 contribution ) : @xmath37 here , @xmath38 is the magnitude of the order parameter . the wavevector @xmath39 is a reciprocal lattice vector . the spin index @xmath40 included above will be omitted in the future for simplicity unless clarity requires otherwise . in , the form factor @xmath15 , which can transform non - trivially under the point group of the 2-d lattice , is related to the angular momentum of the particle - hole condensate . for ddw order , @xmath41 which corresponds to @xmath42 and @xmath0 symmetry . a @xmath31-density - wave with @xmath43 has @xmath44 , which corresponds to @xmath45 . in real space , becomes @xmath46 where the vertices @xmath47 of the orders discussed above are given by @xmath48 note that the factors of @xmath49 in @xmath50 and @xmath51 signify that the corresponding phases break time - reversal , e.g. current flows along the bonds . positive / negative signs in represent current going in / out of the vertex . translation invariance , by default , is broken by any density wave order since there is a preferred vector @xmath52 , but in the c case , invariance under a _ subset _ of the lattice group is retained . ddw order , for example , is invariant under translation by linear combinations of even reciprocal lattice vectors along each bond direction . the current along the bonds is given by : @xmath53 here @xmath54 is along either the @xmath55- or @xmath56-direction . because of the equivalence of @xmath52 and @xmath57 in the c case , the current is simply @xmath58 . in the ddw case this results in an alternating plaquette current / checkerboard magnetic field patterns.@xcite to obtain an ic version of ddw order , one first notes that ddw is chiral and hence breaks the ising symmetry which reverses all the clockwise plaquette currents into counter - clockwise and vice versa . therefore one can construct anti - phase domain walls as in fig.[fig : domains ] . a simple caricature of a bond - oriented domain wall is given in fig.[fig : domains]a . its real space representation is @xmath59 where @xmath60 is the step function , so @xmath61 with @xmath62 . at the midpoint of the domain wall , the currents in the @xmath63-direction vanish while the currents in the @xmath64-direction are halved . the diagonal domain wall depicted in figure [ fig : domains]b has a similar representation . note that because of the halving of the @xmath64-direction currents at the midpoint of the bond - domain wall in , we need to include the term : @xmath65 where the new vertices @xmath66 are given by : @xmath67 such vertices seem necessary for ic transitions of current density waves . one could imagine situations in which the transition to a disordered state is driven by thermal proliferation of such vertices , similar to earlier analysis of the 6-vertex model@xcite . although the analysis of such a transition is an interesting topic on its own , we will not pursue it in this paper . we are more interested in the energetics of the ic phase , so a long - wavelength mean - field treatment is sufficient . to analyze the c - ic transition , one then considers an array of alternating domain walls with average separation @xmath68 . then , in the single harmonic approximation one retains only the @xmath69 term in and identifies @xmath70 with the incommensurability wavevector . the new short - range vertex contributions are also neglected in the lowest harmonic approximation . the neglect of higher harmonics should not change the transition temperature@xcite or the value of the critical doping for the onset of ic order since @xmath38 is small and the corrections are of higher order in @xmath38 . an incommensurate instability in a clean system occurs when @xmath68 becomes finite . ( in a disordered system , it occurs when when @xmath71 , where @xmath72 is a characteristic length scale for the disorder . ) within the single harmonic approximation , an array of anti - phase domain walls with separation @xmath73 , looks simpler in momentum space : @xmath74 note that in the limit @xmath75 , the above reduces to . one could have guessed the result by writing @xmath76 in the commensurate case , and then let @xmath52 go incommensurate . this suggests other ways to get ic order . in particular the oddness of ddw order with respect to rotations by @xmath77 and transpositions about , say , the @xmath4 direction lead to the following example of a ddw _ checkerboard pattern _ : @xmath78 where the operation @xmath79 can be either a transposition , or a rotaion by @xmath77 . the checkerboard pattern in corresponds to simply superimposing two domain walls rotated by @xmath77 with respect to each other . in a similar spirit , one could also investigate the possibility of _ non - topological _ domain walls@xcite , e.g. ic domains as the above , superimposed with an uniform order , so the overall order parameter does not change chirality . however , as we will see in sec . [ num ] , much of the physics of the c - ic transition will already be transparent from the simpler choices and of order parameter . finally , note that within the single harmonic approximation , and are current - conserving only to lowest order in @xmath22 . for the purpose of clarifying the energetics of the ic order close to the c - ic transition , or in the cases where the magnitude of the incommensurability is small , this lowest order approximation is good enough . our numerical results , sec . [ num ] , _ a posteriori _ show that we are in such regime . we will consider the above order parameters in the context of the following hamiltonian @xcite : @xmath80 in this formula , @xmath81 is hopping with @xmath82 for nearest neighbors , @xmath83 for next nearest neighbors and @xmath84 otherwise . @xmath85 is a correlated hopping term which simultaneously hops an electron from site @xmath86 to site @xmath49 and hops an electron from @xmath87 into the vacated site @xmath86 . the on - site and nearest - neighbor repulsions are , respectively , @xmath88 and @xmath89 . the indices @xmath90 signify lattice sites and @xmath91 the spin . we will consider this hamiltonian at the mean - field level . at this level , the energetics of ddw order is the same as for a ` bcs - reduced ' hamiltonian for ddw order of the form : @xmath92 with @xmath93 . a similar reduced hamiltonian for @xmath5-wave superconductivity has @xmath94 . note that the processes responsible for both dsc and ddw are essentially kinetic . a hamiltonian of the form ( [ micro - ham ] ) has one important virtue : it has a @xmath5-wave superconducting ground state over a range of dopings ( and an antiferromagnetic ground state at half - filling ) . since this is the _ sine qua non _ for any description of the cuprates , we believe that this is a good starting point . one may object that correlated hopping terms have to be put in by hand in a microscopic hamiltonian . however , they naturally arise even from the one - band hubbard model away from half - filling , where their contribution can be computed in @xmath95 perturbation theory : @xmath96 at dopings @xmath97 , for @xmath98 , we would have @xmath99 . thus , not far away from half - filling , correlated hopping terms are certainly not negligible . let us generalize the mean field description of c ddw to the ic case . the path we will take is similar to that in sec . [ icop ] , where we wrote down the _ commensurate _ order parameter without making use of the @xmath100 symmetry , and as a result the generalization to incommensurate @xmath52 was straightforward . we thus obtain a functional @xmath101 , equation , whose minimization yields self - consistently both the order parameters @xmath102 and the deviation from commensurability @xmath103 . the only subtlety , as we discuss at the end of this section , is that although @xmath104 reduces to the mean - field free energy in the c phase , in the ic phase it is _ not _ the free energy even within the single harmonic approximation . instead , minimization of @xmath104 is equivalent to a variational search of the ground state . our derivations generalize the treatment of nayak and pivovarov @xcite of competition of c ddw with dsc and possibly af . the c ddw mean field hamiltonian we start with is : @xmath105 where @xmath106 . note that in the @xmath107 symmetry for _ incommensurate _ @xmath52 is not equivalent to the hermiticity of the hamiltonian due to the @xmath108-dependance of @xmath109 ( e.g. @xmath110 for ic * q * ) . this is unlike the case of af order , where the interaction is simply @xmath111 and we have @xmath112 . in order to check with earlier results on the c - ic transition of af order in the _ weak coupling _ hubbard model@xcite ( see sec . [ susc ] ) , we will also include a term like in our full reduced hamiltonian . of course , in the strong coupling , @xmath113 , limit , one is not justified in treating af as a fermi surface instability , especially close to half filling . a @xmath3 description is more appropriate in this case . however , as mentioned in sec . [ micro ] , both ddw and dsc are essentially kinetic in origin within the extended hubbard model , so even in the strong onsite repulsion limit , there is a good justification to use mean field theory to describe their competition away from half filling . in the c case , the standard dsc term in our reduced hamiltonian is : @xmath114 where @xmath115 . using a nambu basis @xmath116 , the full hamiltonian @xmath117 can be rewritten as @xmath118 where : @xmath119 with @xmath120 . the action derived from the above hamiltonian yields , upon integrating the fermions out , the free energy in the ordered state : @xmath121 with @xmath122 the energy eigenvalues of , and @xmath123 if we assume @xmath100 symmetry , @xmath124 would be given by : @xmath125 ^ 2}\ ] ] where @xmath126 and @xmath127 . however , in the ic case , @xmath100 need not hold , in which case the above would generalize to : @xmath128\delta^2_\text{dsc}+ f^4_-({\bf k})\delta^4_\text{dsc}. \end{split}\ ] ] the effect of a nonzero incommensurability @xmath12 can be seen not only in the appearance of the terms @xmath129 and @xmath130 , but also in the dsc - and doping - dependent function @xmath131 . our final step before letting @xmath52 be incommensurate is to pass the integration in from rbz to bz : @xmath132 once we convert the rbz integrals to bz integrals as in , we numerically minimize ( see sec . [ num ] ) the resulting ` free energy ' with respect to both the order parameters @xmath102 and the deviation from commensurability @xmath12 . ev and @xmath133ev are large ( compare to figs . [ fig : ddwpd3 ] and [ fig : ddwpd4 ] ) @xmath134ev . the next nearest neighbor hopping , @xmath135ev , is small compared to typical values fitted to arpes@xcite , but ic order was nevertheless robust away from half - filling . inset : ic wavevector @xmath22 at @xmath136ev as a function of doping . wavevectors are in units of @xmath137 , where @xmath24 is the lattice spacing.,height=288 ] it must be emphasized that the expression to be minimized in sec . [ num ] is _ not _ the true free energy of the system if @xmath52 is incommensurate . one way to see this is to note that the eigenvalues are not invariant under @xmath138 if @xmath139 and therefore they can not be interpreted as band energies . in a c - ic transition , the excitations split into a generally uncountably many bands even within the single harmonic approximation@xcite . therefore , the problem of finding the mean - field free energy is notoriously difficult even in the single harmonic limit . the calculation we are doing , on the other hand , is essentially a mean - field variational search for the ground state , and not for the excitations . a similar calculation was done by schulz@xcite for the case of the c - ic transition of the weak - coupling hubbard model close to half - filling . by continuity of the true free energy@xcite , one expects that the resulting variational estimates for @xmath140 would be accurate only close to the c - ic transition , where @xmath22 is small . ev , @xmath141ev are ` large ' ( note temperature scale ) similar to fig . [ fig : ddwpd1 ] , with @xmath142ev . however this time we taka a more realistic @xmath143ev . again ic order is robust at dopings relevant for the pseudogap . inset : ic wavevector @xmath144 at @xmath136ev as a function of doping . [ fig : qdw4 - 2tp012 ] for a more detailed view of @xmath144 ( and also fig . [ fig : qdw4 - 2tp006 ] for the case of @xmath145ev ) .,height=288 ] ev , @xmath146ev , @xmath147ev . the next nearest neighbor hopping , @xmath148ev . note the appearance of dsc proximate to ` magic ' dopings . inset : the ic wavevector @xmath149 at @xmath150ev . see fig.[fig : qdw25 - 1tp012 ] for comparison between @xmath149 and @xmath151 . , height=288 ] to obtain the phase diagrams for the models in , for various fixed dopings @xmath63 and temperatures @xmath152 , we minimize @xmath153 with respect to both the order parameters @xmath102 ( as was done previously in the c - case@xcite ) and the ic wavevector magnitude @xmath22 . each choice of ic order parameter ( e.g. bond or diagonal ) corresponds to a separate minimization procedure . to distinguish which ic order dominates at a particular point in @xmath154 space , we compared the free energies of the orders considered , and chose the ic phase of minimal energy . note that , as given in , the free energy is a function of @xmath155 , so to obtain the dependance on the doping @xmath63 , we first perform a legendre transformation and minimize instead @xmath156 where the chemical potential is obtained from @xmath157 with @xmath158 . our approach is slightly different than that previously used in the c - case@xcite . there , the competition between af , ddw and dsc was analyzed , and some sensitivity on the details of the microscopics was found . phase diagrams resembling that of the cuprates were obtained . the general trend was that any phase diagram which included @xmath5-wave superconductivity also had ddw order competing with it . by changing the bare coupling constants , rather different phase diagrams could be obtained and these generally did not have superconductivity . in our treatment here , we do not consider the competition with af order , which is not a fermi surface instability anyway in the context of the cuprate superconductors . in , we still include af , but only in order to check with previous results@xcite of ic af order away from half filling in the _ weak coupling _ @xmath159 limit . on the other hand , the af order appearing at half filling in the cuprates is a strong coupling phenomenon . a reduced hamiltonian is the wrong starting point at half filling in the limit @xmath113 . therefore , we only consider competition between ddw and dsc and look for ic order in this case . all the phase diagrams constructed this way are robust in the sense that ddw and dsc are found to coexist at mean field level for almost all reasonable values of the coupling constants . sample phase diagrams are shown in figs . [ fig : ddwpd1]-[fig : ddwpd4 ] . in all of them , ic order is present at dopings @xmath160 where ddw and dsc compete . in the first two of these figures , we vary @xmath18 at relatively high @xmath161 and @xmath162 , while in the last two we explore the phase diagram for more realistic ddw and dsc couplings ( compare the temperature scales to those in the cuprates ) . note the complexity of figs . [ fig : ddwpd3 ] , figs . [ fig : ddwpd4 ] compared to figs . [ fig : ddwpd1 ] , figs . [ fig : ddwpd2 ] . at realistic couplings , the ic wavevector profile is complicated , and furthermore dsc order appears in more than one doping region within the ddw dome . in fact , the appearance of incommensurability seems correlated with the appearance of dsc . a more systematic analysis to be discussed below reveals that generally the types of c - ic transitions that were produced in our phase diagrams fall into two categories : ( a ) those which correspond to motion of the bare fermi surface away from commensurate nesting ( shaded regions ) ; except @xmath163ev . there is a distinct ic response ( inset ) at @xmath164 and @xmath165 . [ fig : qdw2 - 1tp012 ] for comparison between @xmath151 and @xmath149.,height=288 ] where ddw and dsc compete . note that nonzero @xmath22 is found above as well as below the dsc critical temperature.,height=288 ] \(b ) those which appear near certain ` magic ' filling fractions . case ( a ) is similar to the familiar c - ic mechanism discussed by schulz@xcite , while case ( b ) appears to be a consequence of negative effective stiffness at special fractions , the exact nature of which at present is not clear . in both cases @xmath22 is independent of the onset of dsc , fig . [ fig : qoft ] , but because incommensurability relieves the ddw - dsc competition , the onset of dsc correlates to the presence of ic ddw . let us first discuss ( a ) . it is well known@xcite that in the weak - coupling hubbard model at @xmath26 , the af instability at half - filling is driven ic for any non - zero doping ( see sec . [ suscep ] ) . the source of the c - ic transition can be traced to the approximate nesting of the bare fermi surface ( fs ) at ic wavevectors away from half - filling . close to half - filling , this picture would produce a linear deviation from commensurability , @xmath166 , fig . [ fig : afpd1 ] . the incommensurability ( deviation from @xmath4 ) is horizontal or vertical , not diagonal , as we would expect because the bare susceptibility is more divergent due to a greater overlap between nested portions of the fermi surface@xcite . a similar scenario happens in the case of ddw order . however , two notable differences from the af case are present . first , ddw is less sensitive to @xmath167 , since it is driven mainly by antinodal regions of high density of states ( d.o.s . ) . therefore , while there is no af instability without nesting ( at half - filling af fermi surface instability is only possible if @xmath168 ) , ddw is still possible in this regime . this suggests that whenever the antinodal regions of the bare fs evolve away from the anti - nodal regons ( which happen to also be regions with a high density of states as a result of the presence of van hove singularities ) , a c - ic transition is likely to occur . . the results of schulz@xcite were recovered . here , we used @xmath169ev with @xmath134ev . bond - oriented ic order was found to be energetically favorable , as a direct computation@xcite in the weak repulsion regime would yield . _ inset _ : as expected , the ic wavevector at @xmath170 is linear in @xmath63 . , height=288 ] the second difference from the af case is that ddw has nodes . this implies that the chemical potential in the ordered state need not be higher than the maximal ddw gap . as a result , ddw order can stay commensurate for wider ranges of doping . this is indeed what we find . for example , for all @xmath171 , we obtain @xmath172 near zero temperature . in general , we find that the doping values @xmath173 for which the above type of c - ic ddw is possible are unrealistically high in the context of the cuprates , if one considers more realistic bare fs , @xmath174 . furthermore , in the ic ddw case , there need not be a simple relation @xmath166 between the doping and doping and deviation from commensurability . let us now discuss the c - ic transitions that are present in our phase diagrams proximate to ` magic ' filling fractions @xmath175 . although our understanding as to why they arise is incomplete , we would like to point out several interesting features about our results . first , to our knowledge , there is nothing special about the bare fs close to these filling fractions , likewise for the fs of the commensurate ordered state . the c - ic transition is not driven by nesting , since upon variation of @xmath18 we see similar signatures proximate to the ` magic ' dopings . second , the magnitude of the ic wavevector is generally small ( corresponding to periodicity of more than @xmath176 lattice spacings ) and shows no clear dependance on @xmath177 , figs . [ fig : qdw4 - 2tp006 ] -[fig : qdw2 - 1tp012 ] . third , we checked that similar behavior is present deep into the af ordered state in the total free energy . [ fig : qaf004t0tp004 ] . unlike the ddw case , however , the ic lattice spacing in the af case depends on @xmath177 : @xmath22 follows an increasing staircase pattern at ` magic ' dopings . note that our results for af order seem to contradict schulz s results@xcite which predict a linear dependence of @xmath149 . the naive contradiction is easily resolved by realizing that schulz s computation@xcite of @xmath149 relies on the stoner criterion , which is strictly valid at @xmath10 . indeed , while deep into the ordered state the ic wavevector is stepwise increasing at ` magic ' fractions , at higher temperatures we regain the linearity of @xmath178 , fig . [ fig : qaf004t035tp004 ] . moreover , the fact that both the bare and the gapped fs in the c case are generally not nested at the the ` magic ' fractions suggests that the transition is not driven by a divergent susceptibility , but is first order . indeed , at low temperatures , the chemical potential near the special dopings is non - monotonic for both the af and the ddw+dsc diagrams . similar behavior of @xmath151 was found by nayak and pivovarov in the context of c ddw competing with dsc@xcite . a simple view of the energetics is given by the following landau - ginsburg free energy of two nodeless competing orders ( the generalization to order parameters with nodes is straightforward ) : @xmath179 in this free energy , @xmath180 plays the role of the ddw order parameter and @xmath181 the role of superconductivity . when @xmath182 becomes negative , @xmath180 orders at non - zero wavevector @xmath183 . this occurs at the dopings indicated in figures [ fig : ddwpd3 ] and [ fig : ddwpd4 ] . when @xmath181 orders , this further favors incommensurability if @xmath184 since the competition between the two orders is partially alleviated . this alleviation is a possible explanation of why the onset of dsc at low @xmath152 is correlated to the presence of ic - ddw . and @xmath151 ( inset ) at @xmath136 for @xmath185ev , @xmath186ev . note that @xmath187ev which is smaller than the value taken in fig . [ fig : ddwpd2].,height=288 ] except this time @xmath143ev as in fig . [ fig : ddwpd2].,height=288 ] and @xmath149 ( inset ) at @xmath150ev corresponding to the parameters in fig . [ fig : ddwpd3 ] . although here the ic response is not at ` magic ' fractions , note that in fig . [ fig : ddwpd3 ] there is dsc both at @xmath164 and @xmath165 . , height=288 ] ( and at the same temperature ) except this time there is a clear c - ic transition at ` magic ' fractions ( inset ) . at the same time @xmath151 is non - monotonic at these dopings . the coupling constants @xmath188ev and @xmath189ev are the same as in fig . [ fig : ddwpd4].,height=288 ] ev ) staircase ic spectrum of af around ` magic ' filling fractions . the parameters we have taken here are @xmath190 ev , @xmath142 ev . inset : @xmath151 is non - monotonic at both jumps of @xmath149 , but is smoothly decreasing whenever @xmath22 is pinned.,height=288 ] the same phenomenon can be viewed from the opposite perspective : when ddw order becomes incommensurate , there are regions in which it is suppressed ( as in the domain walls of fig . 2 ) . superconductivity can occur more easily there . hence , incommensurability alleviates the competition between dsc and ddw order at the mean - field level , thereby enhancing superconductivity . of course , ic order appears to suppress superconductivity in the cuprates , but this could be due to effects beyond a simple mean - field theory . when the superconducting order parameter becomes strong enough , it may become energetically favorable for phase separation to occur . we can guess when this occurs from the usual maxwell construction whenever the mean field chemical potential @xmath151 does not vary monotonically . this appears to occur at a few isolated ` magic ' doping values . however , in a real system , proximity to a second order critical point@xcite , coulomb repulsion@xcite , or disorder could drive the system in a state of mesoscopic phase separation . moreover , the lattice pins any ic density wave order , so once the system is ic in order to minimize the free energy , the deviation from commensurability @xmath22 would be pinned at a finite value . finally , it is important to remember that disorder could have important effects . it would not only disorder the array of domain walls , but could lead to droplets of one phase in regions of another . regains its linearity around @xmath191 as predicted by schulz@xcite . here @xmath192ev and all other parameters are as in fig . [ fig : qaf004t0tp004 ] . inset : @xmath151 is monotonic throughout , suggesting that at higher temperatures , the c - ic transition is driven purely by ic nesting of the fs away from half - filling.,height=288 ] we stress that while the general trend towards ic order may be clear , many of the quantitative aspects are not clear , in particular the precise copings at which ddw order becomes incommensurate . it is not clear to us why such c - ic transitions happen near ` magic ' filling fractions . the fact that similar behavior occurs in the case of af order suggests that the tendency of phase separation at ` magic ' filling fractions is a general trend for a variety of density wave order . because the af only case was explored upon taking the @xmath193 limit in the full free energy , it is not clear yet whether the tendency of phase separation at ` magic ' filling fractions is purely a feature of density waves competing with other order parameters , or of the lattice . it might be worth reconsidering the af case in the presence of other order parameters ( such as ddw , dsc ) . what makes our results surprising is that unlike in the frustrated phase separation picture for stripe formation , there is no coulomb repulsion present . hence , there is no obvious connection between phase separation and incommensurability . ( please note that while ic ddw order can induce charge density wave order at twice the wavevector , we have not included this effect , which we expect to be small , in our calculation . ) therefore , while coulomb repulsion may play a role in determining the cdw spacing at @xmath164 in @xmath194 , in our calculation , the phase separation at ` magic ' fractions can only arise as interplay between the lattice , competing order physics and energetics . what exactly is the combination of the three deserves some future research as the analog of the fulde - ferrel - larkin - ovchinnikov ( fflo ) state of dsc in a magnetic field parallel to the plane@xcite . in the absence of next - nearest neighbor hopping , the two mean fields are related via a particle - hole transformation in the same way _ commensurate _ ddw is related to dsc at half - filling . however , while in the latter case one can rotate c ddw to dsc@xcite , in the case of finite @xmath155 , the particle - hole transformation can not interchange fflo and ic ddw . while the mean - field hamiltonians of ic ddw and the fflo are related by a particle - hole transformation , the spectra are not in the ic ddw state there is a splitting to a generally uncountably many bands ( if @xmath195 is irrational as we discussed in sec . [ num ] ) since the order parameter connects @xmath108 and @xmath196 while in the case of dsc there is no such splitting . we are indebted to the authors of @xcite for pointing out their work to us . ] . according to the discussion following it seems likely that within the dsc state a competing ddw order parameter is ic as it develops . hence , we briefly describe the experimental signatures that would distinguish between c and ic ddw order . originally , c ddw was also named ` hidden order ' in the context of the pseudogap , because most of the experimental signatures associated with it are indirect . the magnetic field , created by the alternating currents around neighboring plaquettes , is too small to be detected by neutron scattering ( @xmath197 g for @xmath198 mev ) . at the same time , because of the symmetry of the c ddw magnetic field , there is no signature in nmr . the charge of the c ddw state is uniform so stm would not show signatures , unless they are associated with interference of nodal quasiparticles due to scattering off impurities . it has been argued recently@xcite that such interference disperses weakly enough to account for the pseudogap d.o.s . incommensurate modulations observed by stm . for a further discussion of the signatures of c ddw , the reader is referred to the literature @xcite . an ic ddw is more easily detectable directly . the local electronic density of states would be modulated due to the ic order parameter , and hence could be observable by stm . secondly , the ic modulations of the staggered currents would produce nmr line splitting of both the cu and o atoms when the applied magnetic fields are _ perpendicular _ to the a - b planes . this is in sharp contrast with ic sdw , where there is an nmr line - splitting due to spatially modulated local magnetization , which is directed _ along _ the a - b plane . let us estimate the modulated magnetic field strength arising from ic ddw ( the reader is referred to@xcite where similar considerations are applied to the case of orbital antiferromagnetism proposed to exist in @xmath199 ) . suppose we have an array of domain walls , fig . [ fig : domains ] , along the y - axis . deep between the domains the field is zero as in the case of c ddw . along a domain at @xmath200 , where the field is maximal , again most of the fields cancel except those created by the vertical alternating currents of strength @xmath201 at @xmath202 and those of strength @xmath203 at @xmath204 . equivalently this would be the field of a single vertical array of alternating currents of strength @xmath203 at @xmath202 : @xmath205 where @xmath206/[\sqrt{((i-1)a + y_0)^2 + x^2_0}]$ ] and @xmath207 for o , cu atoms respectively . clearly at @xmath208 , cu , atoms the field is zero by symmetry . for o atoms the sum can be performed yielding @xmath209 . in the dilute domain limit , this field is independent of the distance between the domains . for a sinusoidal current modulation , @xmath210 would depend on the modulation length @xmath211 . if @xmath212 then a good estimate of @xmath210 is given by with @xmath201 replaced by @xmath213 where @xmath177 is the periodicity of the modulation ( in units of the lattice spacing ) . therefore , for large @xmath177 we obtain @xmath214 . incommensurate order is an important possibility when competing orders are considered . at a first - order transition the system is phase separated and hence necessarily inhomogeneous . if the competition between the orders is weak , coexistence is possible , but modulations of the competing orders are likely to be induced . the effects of the lattice , disorder , lomg - range coulomb repulsion@xcite and proximity to criticality@xcite is to generally stabilize incommensurability on mesoscopic length scales . this is unlike the case of , say , the liquid to vapor phase transition where the competition results in phase separation on macroscopic length scales . in the case of ddw and dsc , the competition is certainly weak within the framework of the extended hubbard models we have considered . because the processes stabilizing both orders are of similar nature and are at the same time smaller than all of onsite repulsion , exchange and hopping , it does not cost much energy to convert ddw order into dsc and vice versa . a simple resolution of the competition is for ddw order to become incommensurate . we find that this occurs for some doping levels , but the incommensuration tends to be relatively small . such ic ddw could , in principle , be observable at low temperatures by nmr since it creates an inhomogeneous magnetic field at o atoms _ perpendicular _ to the ab - plane . most of nmr experiments test for much stronger magnetic fields _ parallel _ to the ab - plane on the cu atoms , and at higher temperatures such ic ddw phase would be smeared by thermal fluctuations , which may exlpain why such ic order , if it exists , it hasnt been seen by experiment . we would like to thank s. brown , s. chakravarty , r. jamei , s. kivelson , and e. pivovarov for helpful discussions . special thanks to e. pivovarov for letting i. d. generalize some of his computer code . this work has been supported by the nsf under grant no . dmr-0411800 . 99 u. schollwck , s. chakravarty , j. o. fjaerestad , j. b. marston , and m. troyer , phys . lett . * 90 * , 186401 ( 2003 ) ; e. dagotto , rev . phys . * 66 * , 763 ( 1994 ) ; c. s. hellberg and e. manousakis , phys . lett . * 83 * , 132 ( 1999 ) ; s. sorella , g.b . martins , f. becca , c. gazza , l. capriotti , a. parola , e. dagotto , phys . lett . * 88 * , 117002 ( 2002 ) . s. r. white and d. j. scalapino , phys . b * 61 * , 6320 ( 2000 ) ; d. f. b. ten haaf , p. w. brouwer , p. j. h. denteneer , and j. m. j. w. o. putikka , m. u. luchini , and r. r. p. singh , phys . lett . * 81 * , 2966 ( 1998 ) ; l. p. pryadko , s. a. kivelson , o. zachar , phys . rev . lett . * 92 * , 067002 ( 2004 ) ; e. dagotto and t. m. rice , science 271 , 618 - 623 ( 1996 ) ; g. roux , s. r. white , s. capponi , a. laeuchli , d. poilblanc , phys . b 72 , 014523 ( 2005 ) . experiments in the cuprates have been interpreted in terms of competing orders discussed in refs . ; and also in c. m. varma , phys . b 55 , 14554 ( 1997 ) ; s. c. zhang , science 275 , 1089 ( 1997 ) . e. w. carlson , v. j. emery , s. a. kivelson , and d. orgad , in _ the physics of conventional and unconventional superconductors _ , k. h. bennemann and j. b. ketterson ( springer - verlag 2003 ) , cond - mat/0206217 ; a. h. castro neto and c. morais smith , in _ strong interactions in low dimensions _ , ed . d. baeriswyl and l. degiorgi ( kluwer acad . pub . 2004 ) , cond - mat/0304094 ; v. j. emery , s. a. kivelson , and j. m. tranquada , proc . usa * 96 * , 8814 ( 1999 ) ; m. vojta and s. sachdev , phys . lett . * 83 * , 3916 - 3919 ( 1999 ) ; j. zaanen , nature 404 , 714 ( 2000 ) ; and references therein . j. m. tranquada , physica b 241 - 243 , 745 ( 1998 ) ; in _ neutron scattering in layered copper - oxide superconductors _ , ed.a . furrer ( kluwer , dordrecht , the netherlands ) , p. 225 . j. e. hoffman , k. mcelroy , d .- h . lee , k. m lang , h. eisaki , s. uchida , and j. c. davis , science * 297 * , 1148 ( 2002 ) ; k. mcelroy , r. w. simmonds , j. e. hoffman , d .- h . lee , j. orenstein , h. eisaki , s. uchida , j. c. davis , nature * 422 * , 592 ( 2003 ) ; c. howald , h. eisaki , n. kaneko , m. greven , and a. kapitulnik , phys . b * 67 * , 014533 ( 2003 ) . s. chakravarty , r. b. laughlin , d. k. morr , and c. nayak , phys . rev . * b 63 * , 094503 ( 2001 ) . c. nayak , phys b 62 * , 4880 ( 2000 ) . t. timusk and b. statt , rep . progr . phys . * 62 * , 61 - 122 ( 1999 ) . s. tewari _ et al . _ , , 224516 ( 2001 ) . j.l . tallon and j.w . loram , physica c * 349 * , 53 ( 2001 ) ; j.w.loram _ et al . _ , j. phys . solids @xmath215,2091 ( 1998 ) ; j.l.tallon _ et al . b @xmath216 , 531 ( 1999 ) . y. j. uemura _ et al . _ , lett , @xmath217 , 2317 ( 1989 ) ; c. panagopoulos _ et al . b * 60 * , 14617 ( 1999 ) ; c. panagopoulos _ et al . _ , physica c , * 341 - 348 * , 843 ( 2000 ) . s. chakravarty , c. nayak , s. tewari , , 100504/1 - 4 ( 2003 ) . homes _ et al . _ , , 1645 ( 1993 ) . c. stock _ et al _ , , 024505 ( 2002 ) . h. a. mook _ et al . _ , , 144513 ( 2002 ) . s. chakravarty , h. y. kee , and c. nayak , _ intl . j. phys . _ * 15 * , 2901 ( 2001 ) . s. chakravarty , c. nayak , s. tewari , x. yang , , 277003/1 - 4 ( 2002 ) . et al . _ , nature * 424 * , 912 ( 2003 ) . s. tewari _ et al . _ , , 014514 ( 2004 ) . _ , , 137002 ( 2004 ) . c. wu , j. zaanen , s - c zhang , , 247007 ( 2005 ) . u. schollwoeck _ et al . _ , 186401 ; j. o. fjaerestad , j. b. marston , u. schollwoeck ann . ( n.y . ) * 321 * , 894 ( 2006 ) . ; s. capponi , c. wu , s - c . zhang , 220505(r ) ( 2004 ) . e. cappelluti and r. zeyher , phys . rev . * b 59 * , 6475 - 6486 ( 1999 ) . kee and y. b. kim , phys . rev . * b 66 * , 012505 ( 2002 ) . c. nayak , e. pivovarov , , 064508 ( 2002 ) . h. j. schulz , j. de physique * 50 * , 2833 ( 1989 ) . s. chakravarty , phys . rev.*b 66 * , 224505 ( 2002 ) . a. kotani , j. phys . jpn , * 42 * , 408 ( 1977 ) . l. pryadko , _ et . al . _ , phys . rev . * b 60 * , 7541 ( 1999 ) . a. damascelli , z. hussain , z - x shen , rev . phys . * 75 * , 473 ( 2003 ) . s. kivelson , _ et . al . _ , rev . phys . * 75 * , 1201 ( 2003 ) . r. jamei , s. kivelson , b. spivak , cond - mat/0408066 . c. bena _ et . _ , phys . rev . * b 69 * , 134517 ( 2004 ) ; c. bena _ et . _ cond - mat/0311299 ; a. ghosal , a. kopp , s. chakravarty , cond - mat/0412241 . v. tripathi , p. chandra , p. coleman , j. phys . : condens . matter 17 , 5285 ( 2005 ) .

successful development of the van der waals ( vdw ) heterostructures @xcite based on stacking of two - dimensional ( 2d ) crystals and graphene layers ( gls ) promises a significant progress in infrared ( ir ) and terahertz ( thz ) optoelectronics @xcite . the unique properties of gls provide an opportunity to create new effective photodetectors operating in a wide range of photon energies @xcite . supplementing their design by the band structure engineering using 2d - materials for the barrier layers opens up new prospects in further enhancement of the detector performance . recently , new ir and thz detectors using gl - vdw heterostructures were proposed and evaluated : + ( i ) thz photodetectors based on a double - gl structure , in which the _ intraband _ inter - gl resonant transitions are assisted by ir or thz photons @xcite . the resonant nature of such transitions , in which the electron momentum is conserved @xcite , promotes their elevated probability and , hence , the elevated quantum efficiency . these photodetectors require inclined radiation incidence or a radiation coupler ( grating structure ) ; + ( ii ) thz and ir photodetectors based on the gl - vdw heterostructures using the photon - assisted _ interband _ transitions between the neighboring gls @xcite ; + ( iii ) thz detectors based on the gl - vdw heterostructures operating as hot - electron bolometers using the effect of the emission from gls of the electrons , heated by the thz radiation , due to the _ intraband _ ( drude ) absorptions in gls @xcite . in this paper , we propose and evaluate the ir photodetectors based on the gl - vdw heterostructures with different numbers of the active undoped gls placed between the emitter and collector and separated by the inter - gl barriers ( glips ) . these photodetectors are using the _ interband _ photoexcitation of the electrons in gls from the valence band states to the exited states slightly below or higher than the barrier edge . the photoexcited electrons transfer ( by tunneling through the barrier top or directly ) to the continuum states above the barrier and propagate through the barrier layers . the gl - vdw double - gl photoconductive device with the emitter and collector gls separated by the barrier and using a similar mechanism was recently proposed @xcite fabricated and measured @xcite . however , as we demonstrate below , the insertion of the inner gls , which provides the extra electron photoemission and capture , can lead to a dramatic increase in the glip responsivity . this is due to the photocurrent gain associated with the redistribution of the electric field in the heterostructure and leading to amplification of the electron injection from the emitter . thus the glips in some sense resemble quantum - well infrared photodetectors ( qwips ) typically using iii - v materials and extensively studied , fabricated , and used in applications during two decades @xcite . such a similarity is due to the propagation of the photoexcited electrons over the barriers with their fraction being captured back to the bound states ( in gls and qws , respectively ) . a relatively low capture probability results in the elevated values of the detector responsivity . apart from this , in both glips and qwips the detector detectivity increases with an increasing number of gls and qws ( see , for example , @xcite ) . the main difference between the glips and qwips is in the use of the interband ( in glips ) and intraband ( in qwips ) transitions . this allows eliminating the grating structures required in the n - qwips for radiation coupling . in realistic glips with a relatively large conduction band off - set , the proper electron photoexcitation requires higher photon energies @xmath0 than in qwips . this implies that the glips might generally operate at shorter radiation wavelengths than the qwips . however , modifying the shape of the barriers by doping of the inter - gl layers , one can achieve an effective electron photoescape at fairly low photon energies ( in far - ir range ) . as shown in @xcite , the probability of the electron capture into the gls can be much smaller than that for the qws ( see , for example , @xcite ) since doping of the qws in the qwips leads to the inclusion of the electron - electron scattering mechanism of the capture . smaller capture efficiency is beneficial for a higher detector responsivity @xcite ( see also , @xcite ) . the absence of the gl doping promotes lower electron emission of the thermalized electrons than that in the doped qws and , hence , leads to a smaller dark current . finally , the probability of the electron photoexcitation from the gls is substantially higher than from the qws . these glip features ( the sensitivity to the normally incident radiation and higher photoexcitation and lower capture probabilities ) provide potential advantages of the glips over the qwips and some other photodetectors . the photodetectors based on gl - vdw heterostructures with the tunneling transparent inter - gl barrier layers considered recently @xcite and the glips with relatively thick layers studied in the present paper are analogous . however their operation principles are quite different : the photo - stimulated cascade tunneling between gls in @xcite and the photoexcitation of electrons to the continuum states and propagation in glips . in the latter device , only a small fraction of the propagating electrons can be captured to the gls . such a distinction results in different spectral properties and other detector characteristics . in particular , in the glips under consideration , the photoexcitation probability can be much larger than the probability of the inter - gl photo - stimulated transfer . this is due to the diminished wave function overlap in the neighboring gls @xcite . apart from this , the glips considered here and the hot - electron glips using the bolometric mechanism @xcite are characterized by a rather low capture probability of electrons in the continuum states @xcite . this promotes high values of the detector responsivity @xcite this is in contrast with the detection mechanism associated with the cascade inter - gl photo - stimulated tunneling . the spectral ranges of the photodetectors using the inter - gl photo - stimulated cascade transfer 20 and the electron heating @xcite from one side and the spectral range of the glip operation can be essentially different . one important feature of the glips is that their operation do not require the mutual alignment of the gls . various device models used for studies of the vertical electron transport in qws ( and gls ) introduce the concept of an `` ideal '' emitter ( which provides as many injected electrons as the bulk of the structure `` requires '' ) . in the framework of this concept , the electric - field distributions are assumed to be uniform . a more accurate consideration of the transport phenomena in such structures ( including glips ) in the dark conditions and under irradiation accounts for the self - consistent electric field distributions and the nonideality of the emitter @xcite . this paper uses this more accurate approach . the paper is organized as follows . in section 2 , we describe the possible glip device structures and the glip operation principle . section 3 deals with the device mathematical model . the latter includes the equation for the self - consistent electric potential and the equations governing the electron balance in the gls . in the latter equation , the electron capture into the gls is described phenomenologically invoking the concept of the capture parameter @xcite ( see , also , @xcite . in section 4 , we derive the dark current in the glips as a function of the structural parameters and the applied voltage using the model described in sec . 3 . in section 5 , we calculate the photocurrent ( the variation of the current under the incident radiation ) . the expressions obtained in sections 4 and 5 are then used in sec . 6 for the derivation of the glip detector responsivity , photoelectric gain , and detectivity . in section 6 , we compare the glips with some other ir photodetectors . in the conclusion section , we draw the main results of the paper . in the appendix , we discuss some simplifications of the main model . we consider the glips based on the gl heterostructure which consists of the @xmath1 inner gls clad between the @xmath2 barrier layers with the top and bottom contact gls . the conduction band offsets between the emitter gl and the barrier layer , @xmath3 , as well as between the inner gls and the barrier layers between them , @xmath4 , are smaller than the pertinent offsets in the valence band . generally , the materials of the emitter barrier layer and of other barrier layers can be different , so that @xmath5 . the top and bottom gls are assumed to be doped to provide their sufficiently high lateral conductivity . these extreme gls serve as the tunneling emitter ( injecting electrons into the gl - heterostructure bulk ) and the collector , respectively . the top and bottom gls are supplied with contacts between which the bias voltage @xmath6 is applied . we consider the gl heterostructures with either the n - type or p - type doping of both the emitter and collector gls and undoped inner gls and the barrier layers , i.e. , the heterostructures with different types of the emitter and collector can also be treated in the framework of our model . the applied voltage in such heterostructures , stimulates the electron tunneling through the barriers with triangular tops . the thermalized electrons tunneling from the gls and the electrons excited by the incident ir radiation propagate over the barriers . figures 1(a ) and 1(b ) show schematically the glips based on gl - vdw heterostructures considered in @xcite and their band diagrams . figure 1(c ) corresponds to the glips with the bound - to - continuum tunneling and photon - assisted transitions considered in the present work . figure 2 shows the band diagrams of the glip ( with the structure of figure 1(c ) ) using the interband intra - gl photoexcitation with the sequential tunneling of the photoexcited electrons ( a ) in dark conditions and ( b ) under the strong ir irradiation at the sufficiently high bias voltage ( @xmath7 , where @xmath8 is the electron charge and @xmath9 is the temperature in the energy units ) . the applied voltage induces the extra carriers of the opposite signs and changes the electron fermi energies in the emitter and collector gls . the electron fermi energies , @xmath10 and @xmath11 , in these gls ( at their equal doping ) are different : @xmath12 and @xmath13 here @xmath14 is the fermi energy in the extreme gls at their donor ( upper sign for the n - type contact gls ) or acceptor ( for the p - type contact gls ) density equal to @xmath15 , @xmath16 is the planck constant , @xmath17 cm / s is the electron and hole velocity in gls , and @xmath18 and @xmath19 . hence , the bias voltage can to some extent affect the emitter injection characteristics decreasing the electron activation energy in the emitter @xmath20 ( see the appendix ) . to enhance the electron collection by the collector , the bottom gl can be replaced by a doped multiple - gl or by a thick wider band - gap doped collector region . in principle , the glips with the n - type emitters can be supplied with p - type collectors and vice versa . the operation of the glips with the device structure and the band diagrams corresponding to figures 1(c ) and 2 considered below is associated with the following . the incoming ir photons generate the electron - hole pairs ( due to the photon absorption and the vertical interband transitions ) in each gl . depending on the ratio @xmath21 $ ] , the photogenerated electrons appear in the continuum states above the barriers directly ( if @xmath22 ) or via the intermediate excited state with the energy @xmath23 with respect to the dirac point following the electron tunneling through the triangular barrier top @xcite [ see figure 2(b ) ] . upon the photoexcitation to the states over the barriers , the electrons propagate toward the collector . during the electron transport over the barrier , a fraction of the propagating electrons is captured into the gls . since we consider devices with the barriers , in which the conduction band off - set @xmath4 and the quantity @xmath24 are smaller than the valence band off - set , the escape of the photogenerated holes from the gls can be neglected . the glip barriers in the devices under consideration can be made of the van der waals structures with the inse , ws@xmath25 , and other barrier layer materials @xcite . the poisson equation governing the electric potential distribution @xmath26 in the direction perpendicular to the gl plane ( the @xmath27direction ) and the equations governing the electron balance in each gl at the steady - state irradiation are presented in the following form : @xmath28 @xmath29 here @xmath30 and @xmath31 are the 2d densities of electrons and holes in the inner gls with the index @xmath32 ( @xmath33 and @xmath34 corresponding to the emitter and collector gls , respectively ) , @xmath35 is the density of the electron current across gls , @xmath36 is the capture efficiency for the electrons crossing the @xmath37gl @xcite , @xmath38 and @xmath39 are the rates of the tunneling emission of the thermalized electrons and the tunneling emission of the electrons photoexcited ( photogenerated ) from the gl valence band in the @xmath37th gl , @xmath40 is the probability of the interband photon absorption in gls , ( @xmath41 is the fine structure constant , where @xmath42 is the speed of light in vacuum and @xmath43 is the barrier material refraction index ) , @xmath44 is the probability of the photoexcited electrons tunneling escape from the @xmath37th gl to the continuum states above the inter - gl barriers , @xmath45 is the ir radiation intensity ( photon flux inside the heterostructure ) , @xmath46 and @xmath47 are the thickness of the inter - gl barrier layer and its dielectric constant , respectively , and @xmath48 is the dirac delta function ( which reflects narrow localization of the electron and hole charges in the gls ) , in contrast to qwips using the intraband transitions , the photoexcitation rate in glips is virtually independent of @xmath49 . both thermalized electron tunneling rate @xmath38 and the escape probability of the photoexcited electrons @xmath44 from the @xmath50-th gl ( and , hence , the photoemission rate ) can markedly , namely , exponentially increase with the electric field in the barrier under this gl @xmath51 . the escape probability of the photoexcited electrons is also determined by the effective energy of the photoexcited electron @xmath52 . although the parameter @xmath36 characterizing the capture into the @xmath50-th gl depends , to some extent , on the local electric field , this dependence is much weaker than the field - dependence of the tunneling from the gls . in reality , @xmath36 is most likely to be a functional of the electric - field spatial distribution @xcite . this is particularly true if the energy relaxation length @xmath53 is larger or comparable with the heterostructure thickness @xmath54 . the dependence @xmath55 versus @xmath6 is smoothly decreasing with @xmath6 ( see , the previous calculations @xcite ) . hence , in the simplest approach can be presented as a function of the average field ( in fact , of the bias voltage ) as assumed in the following . in line with this , we set @xmath55 , where @xmath56 is a function of @xmath6 . the tunneling rates of the thermalized and photoexcited electrons ( neglecting the intraband drude absorption in the emitter gl due to relatively high energies of ir photons ) can be presented as @xmath57 @xmath58 here @xmath59 is the field characterizing the tunneling through the triangular barrier top @xcite ( see also @xcite ) , @xmath60 is the electron mass in the barrier material , @xmath61 is the maximum current density which can be extracted from an _ undoped _ gl ( @xmath62 ) at a given temperature , @xmath63 is the try - to - escape time , @xmath64 is the characteristic time of the photoexcited electrons energy relaxation , and @xmath65 $ ] if @xmath66 and @xmath67 if @xmath68 . the factor @xmath69 accounts for the difference in the tunneling transparency of the thermalized and photoexcited ( with the energy @xmath24 ) electrons . assuming @xmath70 ev and @xmath71 ( @xmath72 is the free electron mass ) , we obtain @xmath73 v/@xmath74 m . at @xmath75 , the right - hand side of equation ( 3 ) corresponds to @xmath76 ( i.e. , the majority of the photogenerated electrons leaves the gls . in the opposite case @xmath77 , @xmath78 . the energy relaxation of photoexcited electrons in the gl - structures under consideration can be associated with ( i ) the emission of optical phonons with the transition of an electron to a low energy state in the gl conduction band ( at @xmath79 , where @xmath80 ev is the optical phonon energy ) , ( ii ) the electron recombination assisted by the emission of optical phonon ( when @xmath81 , and ( iii ) the collisions of the photoexcited electrons with the thermal electrons and holes ( in a wide range of ir photon energies @xmath0 . according to the above , one can set @xmath82 here @xmath83 is the time of spontaneous optical phonon emission in gls and @xmath84 is the time of the photoexcited electrons relaxation on the thermalized electrons ( and holes ) in the gls . the latter is much shorter in both undoped and doped gls ( see , for example , @xcite ) . in the following we assume that @xmath85 considering only the mechanism ( iii ) . since @xmath84 depends on the electron density , we distinguish @xmath86 in the emitter gl and @xmath64 in the inner gls . as demonstrated in the appendix , @xmath87 . although the ratio @xmath88 can be estimated for the heterostructures with different barrier layers , we will mainly considered is as a phenomenological parameter . in equation ( 3 ) , we have neglected the thermionic emission from the gls . the pertinent contribution is proportional to the factor @xmath89 . the latter is assumed to be small compared with the factor @xmath90 in equation ( 3 ) at the actual values of the electric field in the barriers and @xmath91 k and lower . equation ( 1 ) is supplemented by the following boundary conditions : @xmath92 the current density @xmath35 is equal to the density of the current injected from the emitter gl through a pertinent barrier . considering both the injection of the thermalized electrons from the doped emitter gl and the escape of the electrons photoexcited in this gl , one can use the following formula relating this current density and the electric field @xmath93 : @xmath94 here @xmath95 $ ] is the work function from the emitter to the barrier layer normalized by the conduction band offset . the factor @xmath96 is associated with an increase of the thermalized electron density by the electric field in the emitter barrier layer @xmath93 in the glips with the -type emitter layer ( see the appendix ) . one needs to stress that the value of @xmath97 in the glips with the emitter gl of p - type is larger whereas @xmath98 is smaller ( @xmath99 ) then those for the glips with the n - type emitter gl ( due to different signs of the fermi energy counted with respect to the dirac point ) . versus @xmath100 of glips with ( a)different @xmath1 for @xmath70 ev and @xmath101 , ( b)different @xmath1 for @xmath70 ev and @xmath102 , , and ( c ) different @xmath4 at @xmath103 and @xmath102 . , width=264 ] in the absence of irradiation ( @xmath104 ) , the current across the glip constitutes the dark current , so that @xmath105 . as follows from equations ( 2 ) , ( 3 ) , and ( 7),the electric fields in the barriers are the following functions of the dark current density : @xmath106 in the near emitter barrier ( @xmath107 ) , and @xmath108 in other barriers ( @xmath109 ) . taking into account conditions ( 6 ) and using equations ( 8) and ( 9 ) , we obtain @xmath110 equation ( 10 ) yields the following expression for the dark current - voltage characteristics @xmath111 \biggr\},\end{aligned}\ ] ] [ eq11 ] where @xmath112 . from equation ( 11 ) we obtain @xmath113.\ ] ] [ eq12 ] here the quantity @xmath114\ ] ] changes from @xmath115 if @xmath116 to @xmath117 at @xmath118 . hence @xmath119 is rather smooth function of @xmath6 . consequently , at moderate voltages @xmath116 , corresponding to the most practically interesting conditions when the dark current is small , @xmath120\nonumber\\ \simeq\frac{\gamma^{3/2}v}{(\gamma^{3/2 } + n)d } , \end{aligned}\ ] ] [ eq14 ] @xmath121\nonumber\\ \simeq \frac{v}{(\gamma^{3/2 } + n)d}.\end{aligned}\ ] ] [ eq15 ] it is instructive that , according to equation ( 12 ) , @xmath122 being proportional to @xmath123 strongly increases with decreasing capture efficiency @xmath56 , particularly if @xmath124 ( i.e. , at @xmath125 . this implies that at sufficiently small @xmath56 when @xmath126 , the net dark current can exceed the current of the tunneling thermalized electrons ( which is approximately proportional to the number of inner gls @xmath1 ) , i.e. , the main component of the dark current is associated with the current injected from the emitter controlled by the tunneling from the internal gls being , certainly , smaller than @xmath127 . another feature of the glip current - voltage characteristics and the electric - field spatial distributions given by equations ( 12 ) - ( 14 ) is their weak dependence on the parameters of the emitters @xmath97 and @xmath98 ( at @xmath128 and particularly when @xmath129 ) . the latter says that if @xmath124 , the tunneling emitter under consideration can be considered as an ideal . one needs to stress that due to a decreasing voltage dependence of the capture efficiency @xmath56 , this dependence provides an increasing pre - exponential extra factor in the dark current - voltage relation given by equation ( 12 ) and enhances its steepness . considering the equality of the electric fields in all layers except the emitter barrier layer [ see equations ( 8) and ( 9 ) ] and using equations ( 13 ) , the charge density of the gl adjacent to the emitter gl @xmath130 is given by @xmath131 depending on the emitter gl doping ( n- or p - type ) ) , which determines @xmath97 ( @xmath132 and and @xmath133 in the n- and p - type emitter gl , respectively ) , the charge density @xmath134 can be both negative ( i.e. , the electron density exceeds the hole density ) and positive ( the hole density is larger than the electron density ) , whereas the charge densities of other inner gls @xmath135 ( @xmath136 ) are equal to zero . the case @xmath132 when @xmath137 and @xmath138 , corresponds to the potential in the glip heterostructure profile schematically shown in figure 2(a ) . figure 3 shows examples of the normalized dark current density @xmath139 versus the normalized voltage @xmath100 calculated using equation ( 11 ) for the glips with @xmath140 ev and different number of the inner gls @xmath1 for different parameters @xmath4 and @xmath56 . the comparison of the plots in figures 3(a ) and 3(b ) confirms a weak dependence of the dark current on the number of gls @xmath1 and a pronounced rise of the dark current with decreasing capture efficiency @xmath56 ( particularly at lower normalized voltages @xmath141 ) . higher dark currents at smaller band - offsets @xmath4 are naturally attributed to tunneling emission strengthening from all the gls with decreasing of the tunneling barrier height @xmath4 . as follows from equation ( 12 ) and figure 3 , @xmath122 drastically decreases with decreasing voltage . however , at small voltages , @xmath142 can be so small that the tunneling exponent @xmath143 becomes smaller than the thermionic exponent @xmath144 @xcite . this implies that at such voltages , the tunneling injection from the emitter gl gives way to the thermionic emission . this determines the minimal value of @xmath6 , at which equations ( 11 ) - ( 13 ) are still valid : @xmath145 . at @xmath146 , @xmath147 . this implies that the dependences shown in figure 3 are correct up to small values of @xmath141 ( at @xmath148 for @xmath149 ev ) we limit our consideration of the glip operation primarily by the situations when the incident ir radiation is relatively weak . in the case of irradiation , equations ( 9 ) and ( 10 ) should be replaced by more general equations accounting for the tunneling emission of both the thermalized and photoexcited electrons . since the photoexcitation leads to the variations of the electric fields in the barrier layers , equation ( 10 ) in this situation should be replaced by the following equation which governs the photocurrent density @xmath150 : @xmath151}\nonumber\\ + \sum_{n = 1}^{n}\frac{1}{\ln [ j_{0}/(j_{dark}p + j_{photo}p - e\beta\,i\theta_n)]},\end{aligned}\ ] ] where @xmath44 is defined by equation ( 4 ) and @xmath122 is governed by equation ( 10 ) leading to equation ( 11 ) and its consequences . at relatively low ir radiation intensities , such as @xmath152 , the electron photoexcitation can be treated as a perturbation . considering this , into the quantities @xmath153 and @xmath44 in equation ( 17 ) on can insert the unperturbed values of the electric fields , namely , @xmath154 and @xmath155 . as a result , for the photocurrent density @xmath156 from equation ( 17 ) we obtain @xmath157}{p\beta}\biggr ] } { \biggl ( \displaystyle\frac{\theta_{dark}^2}{\gamma^{3/2 } } + n\biggr)}\nonumber\\ \times \frac{1}{1 + \displaystyle\frac{\tau_{esc}}{\tau_{relax } } \exp\biggl ( \frac{\eta^{3/2}\,e_{tunn}}{e_b^{(dark)}}\biggr)}.\end{aligned}\ ] ] [ eq18 ] here @xmath158 characterizes the glip in dark . according to equations ( 8) and ( 9 ) , @xmath159 is expressed via the dark current density and , invoking equation ( 12 ) , via the voltage : @xmath160 ^ 2\nonumber\\ = \gamma^3\biggl[1 + \frac{v\,\ln(1/kp)}{(\gamma^{3/2 } + n)de_{tunn}}\biggr]^2.\end{aligned}\ ] ] if @xmath161 or at moderated voltages @xmath162 , one can put in equation ( 18 ) @xmath163 , so that one obtains @xmath164}{p\beta}\biggr]}{1 + \displaystyle\frac{\tau_{esc}}{\tau_{relax}}\exp\biggl [ \frac{\eta^{3/2}(\gamma^{3/2 } + n ) \,de_{tunn}}{v}\biggr]}\end{aligned}\ ] ] [ eq20 ] the appearance of @xmath165 in equation ( 20 ) is associated with the contribution of the photoexcitation from the emitter gl . since normally @xmath166 , for the glips with not too large gls ( @xmath167 ) , equation ( 20 ) can be simplified and presented in the following form : @xmath168 } .\ ] ] [ eq21 ] it is worth noting that the photocurrent density given by equations ( 18 ) , ( 20 ) , and ( 21 ) exceeds the current created by the electrons photoexcited from the emitter gl by a factor @xmath169 , which can be very large . this is because the photoemission from the inner gls stimulates a strong injection of the thermalized electrons from the emitter gls resulting in a substantial amplification of the photocurrent . taking into account equation ( 20 ) , one can find small variations of the electric fields caused by the irradiation in all barrier layers @xmath170 as functions of the bias voltage @xmath6 and the radiation intensity @xmath45 . in particular , the electric field in the emitter barrier layer @xmath171 ( as under the dark conditions ) . however , due to the attenuation of the ir radiation as its propagates across the glip structure , the electric - fields in the consequent gls slightly increase : @xmath172 ( such a change is proportional to the radiation intensity @xmath45 , which is weak in the case under consideration ) . the photoexcitation leads to the variations of the electric fields in the barrier layers : an increase in @xmath173 and in a decrease in @xmath174 . at sufficiently strong irradiation , the electric field in the emitter barrier @xmath173 can become larger than the electric field in the bulk @xmath174 . in this case , the gl adjacent to the emitter gl is charged positively , and the potential profile becomes similar to that shown schematically in figure 2(b ) . an high radiation intensities , the electric field in the heterostructure bulk drops , so that the potential drop across the emitter layer tends to the applied voltage @xmath6 . this implies that @xmath173 tends to @xmath175 . consequently , the net current density tend to a value which roughly can be estimated as @xmath176^{-1}$ ] . however at very high radiation intensities , the glip operation might , in particular , be complicated by the following effects : the saturation of the interband absorption @xcite , the influence of elevated density of the photoexcited electrons on their escape @xcite , and the influence of the space - charge of the electrons propagating across the barrier layers on the electric - field distributions . these effects are beyond of our model framework . and different @xmath4 and ( b ) @xmath177 ev and different @xmath1 at @xmath178 ( @xmath179 , and @xmath180),width=264 ] at different normalized voltages @xmath141 and ( b ) for different ratios @xmath88 at @xmath178 ( @xmath70 ev , @xmath181 , and @xmath102).,width=264 ] the detector responsivity @xmath182 is usually defined as @xmath183 here @xmath184 is the device area , @xmath185 is the incident ir radiation power , and @xmath186 is the incident ir radiation flux ( hence , @xmath187 , where @xmath188 is the coefficient of the ir radiation reflection from the glip top . considering this and using equation ( 21 ) , we arrive at the following formula for the glip responsivity : @xmath189 } .\ ] ] here @xmath190 , where @xmath191 is the characteristic responsivity and the factor @xmath192 is determined by the conditions of reflection of the incident ir radiation from the glip top interface ( the presence or absence of an anti - reflection coating ) and its propagation across the heterostructure . the reflection from the substrate can also affect the value of @xmath193 . in the simplest case , the reflection factor @xmath193 can be expressed via @xmath43 : @xmath194 . the deviation of the nominator in equation ( 23 ) from unity reflects the dependence of the responsivity on the emitter parameter @xmath165 . this deviation practically vanishes at large @xmath1 and @xmath128 . in the latter limit , the emitter can be considered as an ideal ( i.e. , providing the necessary injection to compensate the photoexcitation from the inner gls @xcite ) , and the glip responsivity is practically independent of @xmath1 if the average electric field @xmath195 is fixed . since in glips the coefficient @xmath196 is relatively large and the capture efficiency @xmath56 is relatively small ( @xmath197 @xcite ) compered to the standard heterostructures , the characteristic responsivity @xmath198 can be fairly large . indeed , setting @xmath199 ev and @xmath101 , we obtain @xmath200 a / w figures 4 and 5 show the responsivity , @xmath201 ( normalized by the reflection factor @xmath193 ) spectral characteristics of the glips with different band offsets @xmath4 , number of the barrier layers @xmath1 , normalized voltages @xmath100 , capture efficiency @xmath56 , and @xmath88 . in these and the next figure it is assumed that the electron fermi energy in the emitter gl @xmath140 ev . figure 6 shows the responsivity @xmath201 as a function of the normalized voltage calculated for the glips with different @xmath4 . among the main features of the responsivity spectral dependence , the following should be pointed out : ( a ) weak dependence of the spectral characteristics on thee number of the gls @xmath1 ; ( b ) an increase in the responsivity with decreasing capture efficiency @xmath56 ; ( c ) a marked shift of the responsivity maximum toward larger photon energies with increasing the band - offset @xmath4 , normalized voltage @xmath141 , and relative escape time @xmath88 ; ( d ) the responsivity maxima correspond to the photon energies markedly smaller than @xmath202 , so that the glips can exhibit relatively high response at @xmath203 . a fairly pronounced dependence of @xmath201 on @xmath141 ( i.e. , the applied voltage ) implies the possibility of an effective voltage control it is important to stress that at small values of the capture parameter @xmath56 , the photocurrent can markedly exceed the current created solely by the electrons photoemitted from the gls due to the effect of photoelectric gain , which implies that the electron photoemission stimulates much stronger tunneling emission from the emitter . taking into account that the net photoemission rate from all gls is equal to @xmath204 } , \ ] ] the photocurrent density ( 21 ) ) can be presented as @xmath205 where @xmath206 is the photoelectric gain . due to small values of @xmath56 in glips , @xmath207 even at fairly large ( but realistic ) @xmath1 . the dark current limited detector detectivity , @xmath208 , and its normalized version @xmath209 are expressed via the responsivity and the dark current density as ( for example , @xcite ) : @xmath210 here for @xmath122 , @xmath201 , and @xmath211 one can use equations ( 12 ) , ( 23 ) , and ( 25 ) , respectively . considering this , we find ( at @xmath212 ) @xmath213}{1 + \displaystyle\frac{\tau_{esc}}{\tau_{relax}}\exp\biggl ( \frac{n\eta^{3/2}de_{tunn}}{v}\biggr ) } \label{eq29}\end{aligned}\ ] ] here @xmath214[eq30 ] equation ( 29 ) immediately shows that the glip detectivity increases with the increasing number of the inner gls @xmath1 , while the responsivity ( see equation ( 23 ) ) is a weak funcion of @xmath1 . indeed , at large @xmath1 , @xmath215 , but @xmath201 is virtually independent of @xmath1 . some deviation from the latter is associated with the emitter effect ( characterized by @xmath165 ) . the spectral dependence of @xmath216 repeats that of @xmath201 ( see figures 4 and 5 ) . the detectivity @xmath216 markedly increases with the decreasing voltage @xmath6 . this is owing to the pertinent drop of the dark current . however , since at small values of @xmath6 , the tunneling current from the emitter gl can become smaller than the thermionic current , substituting @xmath217 ( see above ) to equation ( 30 ) , we obtain the following estimate for the maximum of @xmath216 : @xmath218}.\ ] ] [ eq31 ] at @xmath219 , the parameter @xmath69 can be small . in this case , equation ( 31 ) yields @xmath220 [ eq32 ] and differerent @xmath4 for @xmath221ev , @xmath179 , and @xmath180 . , width=264 ] the comparison of glips and qwips ( based , in particular , on ingaas / algaas and other heterostructures with the n - type qws ) intended for the operation in mid- and near ir ranges , shows that glips can exhibit the following advantages : ( i ) sensitivity to normally incident radiation ( no radiations coupling structures are needed ; ( ii ) higher probability of the photo - assisted escape from the gl bound states to the continuum states above the barriers due to a relatively large universal constant @xmath222 . ( compare with @xmath223 for a gaas qw with the typical electron density @xmath224 @xmath225 @xcite ) ; smaller capture efficiency of the electrons propagating over the barriers the into the gls ( up to the orders of magnitude - compare the data from @xcite and @xcite ) . these advantages imply that the glip responsivity can exceed that of qwips . since the modulation bandwidths of both glip and qwips is determined by the vertical electron transport across the barrier layers , one might assume that these bandwidths are can close to each other . as follows from equation ( 21 ) and equation ( 4 ) from @xcite , the responsivity , @xmath201 , of a glip and the responsivity , @xmath226 of a lateral gl - photodiode ( glpd ) with a lateral p - i - n junction , are ( at @xmath75 ) approximately equal to @xmath227 respectively . hence , @xmath228 since , the capture efficiency @xmath56 can be very small , the above ratio can exceed unity . since the electron transit time in the glips ( which are vertical transport devices ) ) can be shorter than in the glpds ( with the lateral transport ) , the glips can surpass the glpds in speed . in the double - gl glips @xcite and the utc - pds based on inas / inp heterostructures @xcite , the electrons photoexcited in the emitter transfer ( via the tunneling through the triangular barrier ) to the undoped drift region . in contrast to the double - gl glips and utc - pds , in which the responsivity is determined by the escape of the electrons photogenerated in the emitter layer , the glips can exhibit the effect of amplification of the current produced the electrons photoexcited from the inner gls . this effect in the glips works at a weak electron capture into the inner gls , i.e. , if @xmath229 , but at moderate modulation frequencies . the glip , double - gl glip , and utc - pd response at relatively high modulation frequency is affected by the electron transit across this region . one can assume that the electron transit time in utc - pds can be somewhat shorter than in the glips because of better transport properties of the drift region materials in the former devices . however , the delay in the escape of the photoexcited electrons from the emitter limited by their diffusion ( or in the best case , by the drift in the built - in electric field ) across the heavily doped emitter layer , can be the main factor limiting the utc - pd modulation bandwidths . this implies that the glips ( without such a delay in the emitter gl ) can exhibit a high - speed performance close to that of the utc - pds and , hence , compete with the latter . similar ir photodetectors can be based on the hgcdte heterostructures with multiple zero - gap qws , although the specifics of these heterostructures requires a separate study . we proposed the glips - infrared photodetectors based on vdw hetereostructures with gls serving as the emitter , collector and absorbing layers . the glips exploit the interband electron transitions between the valence band in the gls and the continuum states in the conduction band of the barrier materials . the glips should be able to operate in different ranges of the ir spectrum depending on the values of the conduction band offset between the gl and the barrier layer material . we showed that due to small capture probability of the excited electrons into the gls together with a relatively strong interband photon absorption , the glips can exhibit elevated photoelectric gain and responsivity . an increase in the number of gls in the glip heterostructure can provide relatively high values of the glip detectivity . among other glip advantages are highly conducting transparent contact gl layer , sensitivity to normally incident radiation and high speed operation . the glips , which add the diversity to the ir and thz detector technology , can surpass other already realized and newly proposed photodetectors for the imaging and optical communication systems and thz photomixers . the tunneling rates of the thermalized and photoexcited electrons depends on the electric field , @xmath230 , at the gl . in equation ( 4 ) we set the escape frequency equal to @xmath231 and @xmath232 for the thermalized and photoexcited electrons respectively . thus , it was assumed that @xmath233 is an exponential function of @xmath234 . for more accurate calculations of the escape frequency @xmath233 as a function of the quantity @xmath69 we found the schrodinger equation solutions with the complex energy for gls embedded in a dielectric in the transverse electric field . the gl potential was modeled by a one - dimensional delta - well @xcite : @xmath235 , where @xmath236 is the dirac delta - function and @xmath237 is the electron wave function localization length . this quantity depends on the in - plane electron momentum due to the mismatch of dispersion laws in the gl and the barrier layer @xcite . figure 7 shows the @xmath233 vs @xmath69 dependences calculated for different values of electric field @xmath230 using the above approach . the related dependences calculated using the above simplified formula are also shown by dashed lines for the material parameters corresponding to ws@xmath25 barrier layers ( @xmath70 ev , @xmath71 , and @xmath238 v/@xmath74 m ) . at small @xmath69 the fit is not so ideal but can be considered as reasonable somewhat affecting the shape of the glip spectral characteristics only at @xmath219 . the inverse relaxation time assumed to be primarily determined by the electron - electron scattering , so @xmath239 as functions of @xmath69 in the undoped gls ( @xmath62 ) and doped emitter gl ( @xmath240 mev ) at @xmath241 and @xmath91 k are gls shown in figure 7 by the dotted lines . these curves were obtained by calculating the imaginary part of the electron rpa self - energy @xcite . as follows from figure 7 , in the most interesting actual range of @xmath242 the order - of magnitude estimates yields @xmath243 s and @xmath244 . the latter characteristic times are markedly smaller that the characteristic time of the spontaneous emission of optical phonons @xmath83 for both the intraband and interband electron transitions . thus , for glips with ws@xmath25 barriers one can arrive at the following rough estimate : @xmath245 . as a function of the normalized photon energy @xmath69 for different electric fields @xmath230 calculated using a rigorous model ( solid lines ) and a simplified model with constant pre - exponential factor ( dashed lines ) . the dotted lines correspond to the inverse relaxation time @xmath246 associated with the scattering on the thermalized electrons ( and holes ) for the undoped and doped gls ( @xmath62 mev and ( @xmath240 mev , respectively ) at @xmath91 k. , width=283 ] neglecting above the local dependence of the capture parameter , we found that the electric field distribution across the device structure includes a low - field domain ( in the near emitter barrier ) with @xmath247 and a high = field domain occupying the rest of the structure ( bulk ) where @xmath248 ( see equations ( 8) and ( 9 ) . however , if @xmath36 depends on @xmath249 , the near emitter domain might be extended to a few barriers @xcite . it is worth mention that if the capture efficiency @xmath36 is a sufficiently strong function of @xmath249 , the monotonic electric - field spatial distributions can become unstable against the perturbation with the length @xmath250 @xcite . such an instability can lead to quasi - chaotic spatio - temporal electric field variations , which eventually result in the formation of stable quasi - periodic electric - field distributions @xcite . the electric field in the emitter barrier induces extra electrons in the emitter gl and , hence , leads to an enhancement in the electron fermi level or the hole fermi level if this gl is donor or acceptor doped , respectively . as a result , the parameter @xmath165 can markedly depend on @xmath173 ( decreasing with increasing @xmath173 ) somewhat affecting the current - voltage characteristics . considering this effect , we find @xmath251 @xmath252 . hence @xmath253^{3/2}.\ ] ] setting @xmath254 ev and @xmath241 , one obtains @xmath255 v/@xmath74 m . if @xmath256 , @xmath257 . in this case , @xmath258 , where @xmath259 > 1.\end{aligned}\ ] ] [ eq37 ] at @xmath260ev , @xmath261 ev , @xmath262 , and @xmath241 one obtains @xmath263 . thus , at not too large voltages and emitter electric fields , the injection ( from the contact ) of the extra electrons in the emitter gl does not affect the shape of the current - voltage characteristics , but leads to a replacement of the quantity @xmath61 by @xmath264 . if the emitter barrier layer has the thickness @xmath265 different from @xmath46 , equation ( 6 ) should be modified . the ir irradiation can somewhat reinforce the thermionic electron emission from the emitter gl due to heating high - density 2d electron plasma in the doped emitter gl ( see @xcite ; this effect was disregarded in section 5 ) . the electron photoexcitation can lead to a heating of the carriers in the emitter gl slightly modifying the injection properties of the emitter . however , in the regimes considered above , the emitter properties weakly affect the glip characteristics , particularly , for the multiple gl devices ( @xmath266 ) . this mechanism as well as the photoemission from the emitter gl can be crucial for the double - gl glips without the inner gls ( i.e. , with n=0 ) @xcite . the operation principle of the double - gl glips is similar to that used in the uni - travelling - carrier photodiodes ( utc - pds ) @xcite . but such glips , being potentially interesting , do not exhibit the effect of photoelectric gain , which can substantially increase the responsivity . the bolometric mechanism associated with the electron heating due to the drude absorption can be efficient at the relatively low radiation frequencies in the gl devices with all of the gls doped @xcite ( not considered here ) . the evaluation of the glip dynamic response to transient ir radiation requires a substantial generalization of our model . it will be done elsewhere . here we refrain to a qualitative reasoning . the dynamic operation of the glips under consideration is determined by following characteristic times : the electron transit time across the glip @xmath267 , the time of the gls charging by the injected current due to the electron capture @xmath268 and the time of emptying of the gls by the electron photoexcitation from the gls @xmath269 . the transit time can be estimated as @xmath270 , where @xmath271 is the average electron velocity across the barrier layers . the latter can be set to be equal to the electron saturation velocity @xmath272 . the characteristic times @xmath268 and @xmath269 decrease with an increasing dc component of the incident ir radiation intensity . at small capture parameters , even at fairly strong irradiation , @xmath273 . at the modulation frequencies of the incident ir radiation @xmath274 , the glip responsivity is well described by the above formulas [ e.g. , by equation ( 28 ) ] . in this modulation frequency range , the responsivity is about @xmath275 and virtually independent of the number of the gls @xmath1 . however , in the range @xmath276 , the electron capture does not manage to provide the gl recharging . as a result , the electric field @xmath173 and the current injected from the emitter gl do not follow the temporal variations of the ir radiation . this implies that the mechanism of the photoelectric gain is effectively switch - off . this leads to smaller values of the responsivity , which dependence on both @xmath56 and @xmath1 is rather complex . at higher modulation frequencies @xmath277 , the responsivity markedly drops with increasing frequency @xmath278 . the electron transit time @xmath267 limits the glip modulation bandwidths . setting @xmath279 cm / s and @xmath280 nm , we obtain @xmath281 thz . the effect of the structural parameters on the dependences of the responsivity of single- and multiple - qwips on the radiation modulation frequency was studied previously @xcite . however , the results of these studies can not be directly applied to glips . recent experimental studies of the high - speed operation of the photodetectors based on double - gl vdw heterostructures with the interband photoexcitation @xcite appears to be fairly promising . the authors grateful to v aleshkin , a dubinov and a satou for useful discussion and to n ryabova and s boubanga - tombet for assistance . the work at riec and uoa was supported by the japan society for promotion of science ( kakenhi grants no.23000008 and no.16h06361 ) . v r , d s and v l acknowledge the support by the russian scientific foundation ( grants no.14 - 29 - 00277 and no.16 - 19 - 10557 ) and the ministry of education and science of the russian federation ( grant no.16.19.2014/k ) . the work at rpi was supported by the us army research laboratory cooperative research agreement . xia , f. et al . 2009 photocurrent imaging and efficient photon detection in a graphene transistor . _ nano lett . _ * 9 * 1039 mueller t , xia f n a and avouris p 2010 graphene photodetectors for high speed optical communications _ nat . _ 4297 301 song j c w , rudner m s , marcus c m and levitov l s 2011 hot carrier transport and photocurrent response in graphene _ nano lett . _ * 11 * 4688 ryzhii v , ryzhii m , mitin v and otsuji t 2010 terahertz and infrared photodetection using p - i - n multiple - graphene - layer structures _ j.appl.phys_*107 * 054512 liu c h , chang y - c , norris t b and zhong z 2014 graphene photodetectors with ultra broadband and highresponsivity at room temperature _ nat.nanotechnol . _ 9273 kim c o , kim s , shin d h , kang s s , kim j m , jang c w , joo s s , lee j s , kim j h , choi s .- h and hwang e 2014 high photoresponsivity in an all - graphene p - n vertical junction photodetector _ nat.commun . _ * 5 * 53249 ryzhii v , ryabova n , ryzhii m , baryshnikov n v , karasik v e , mitin v and otsuji t 2012 terahertz and infrared photodetectors based on multiple graphene layer and nanoribbon structures _ _ * 20 * 15 tredicucci a and vitielo m s 20014 device concepts for graphene - based terahertz photonics _ ieee j. sel quantum electron . _ * 20 * 8500109 ryzhii v , otsuji t , aleshkin v ya , dubinov a a , ryzhii m , mitin v and shur m s , 2014 voltage - tunable terahertz and infrared photodetectors based on double - graphene - layer structures _ appl . * 104 * 163505 britnell l et al . 2012 electron tunneling through ultrathin boron nitride crystalline barriers _ nano lett . _ * 12 * 1707 georgiou t et al 2013 vertical field - effect transistor based on graphene - ws@xmath25 heterostructures for flexible and transparent electronics _ nat . nanotechnol . _ * 7 * 100 britnell l , gorbachev r v , geim a k , ponomarenko l a , mishchenko a , greenaway m t , fromhold t m , novoselov k s and eaves l 2013 resonant tunneling and negative differential conductance in graphene transistors _ nat . commun . _ * 4 * 1794 feenstra r m , jena d and gu g 2012 single - particle tunneling in doped graphene - insulator - graphene junctions _ j. appl _ * 111 * 043711 vasko f t 2013 resonant and nondissipative tunneling in independently contacted graphene structures _ phys . rev.b _ * 87 * 075424 ryzhii v , otsuji t , ryzhii m , aleshkin v ya , dubinov a a , svintsov d , mitin v and shur m s 2015 graphene vertical cascade interband terahertz and infrared photodetector _ 2d materials _ * 2 * 025002 ryzhii v , satou a , otsuji t , ryzhii m , aleshkin v ya , dubinov a a , v. mitin v and shur m s 2014 graphene vertical hot electron terahertz detectors _ j appl . * 116 * 114504 ryzhii v , ryzhii m , v. mitin v , shur m s , satou a and t. otsuji t 2013 terahertz photomixing using plasma resonances in double - graphene layer structures _ phys . _ * 113 * 174505 ma q , andersen t i , l. nair n l , gabor n m , massicotte m , lui c h , young a f , fang w , watanabe k , taniguchi t , kong j , gedik n , koppens f h l and jarillo - herrero p 2016 tuning ultrafast electron thermalization pathways in a van der waals heterostructure _ nat . * 12 * 455 ershov m , ryzhii v and hamaguchi c 1995 contact and distributed effects in quantum well infrared photodetectors , appl . * 67 * 3147 thibaudeau l , bois p and duboz j y 1996 a self - consistent model for quantum well infrared photodetectors _ j. appl . phys . _ * 79 * 446 ryzhii v 1997 characteristics of quantum - well infrared photodetectors _ j. appl . phys . _ * 81 * 6442 ryzhii v , khmyrova i , ryzhii m , suris r and hamaguchi c 2000 phenomenological theory of electric - field domains induced by infrared radiation in multiple quantum well structures _ b _ * 62 * 7268 gong c , zhang h , wang w , colombo l , wallace r m and cho k 2013 appl . phys . lett . * 103 * 053513 band alignment of two - dimensional transition metal dichalcogenides : application in tunnel field effect transistors gierz i , calegari f , aeschlimann s , chavez cervantes m , cacho c , chapman r t , springate e , link s , starke u , ast c r and cavalleri a 2015 tracking primary thermaliza- tion events in graphene with photoemission at extreme time scales _ phys . lett . _ * 115 * 086803 ishibashi t , furuta t , fushimi h , kodama s , ito h , nagatsuma t , shimizu n and miyamoto y 2000 inp / ingaas uni - traveling - carrier photodiodes ieice trans . * 938 dyakonov m and shur m s 1998 consequences of space dependence of effective mass in heterostructures _ j. appl . phys . _ * 84 * 3726 massicotte m , schmidt p , vialla f , schadler k g , reserbat - plantey a , watanabe k , taniguchi t , tielrooij k j and koppens f h l 2016 picosecond photoresponse in van der waals heterostructures _ nat . nanotechnol . _ * 11 * 42

fourier s law describes the transport of heat in a macroscopic object . the calculation of the thermal conductivity that appears in the fourier s law is fundamental and important in understanding the properties of materials . such a calculation must be grounded upon quantum mechanics with phonons as basic quasi - particles . this is a non - trivial task and was successfully carried out many years ago by peierls @xcite using the boltzmann equation for phonons . molecular dynamic ( md ) simulation is another approach . however , md results are not correct at low temperatures as it is purely a classical approach . the early works are mostly concerned with bulk materials with many degrees of freedom and periodic lattices @xcite . in recent years , more attentions are paid to heat transport in small or low - dimensional systems such as carbon nanotubes and molecular junctions @xcite . in nano- to meso - scales , new features come in . in most of these conditions , the discreteness of the atom is important . the concept of phonons developed for the periodic lattices is somewhat difficult to apply if a system does not possess translational invariance , such as a nano - junction . in such situation , the boltzmann - peierls approach is not applicable . there are several alternative methods that have overcome the above problems . first , the landauer formula @xcite is a simple and clear description of the heat transport in purely ballistic regime at low temperatures . there are also a few other approaches @xcite with different approximations such as the description based on density matrices . one of the features of the existing theories is that they work in either the ballistic regime or diffusive regime , but not both . it is rather difficult to have a complete theory that can encompass both regimes , apart from phenomenological treatments @xcite . the nonequilibrium green s function formalism described in this article is a serious attempt to be such a complete theory . this theory is certainly `` first principles '' given the atomic potentials . several similar formulations have already been done at the elastic level without nonlinear interactions @xcite . in ref.@xcite , we presented a new formulation with the nonlinear interaction treated systematically . mingo also gave a similar result @xcite for the nonlinear interactions . our approach is essentially a generalization of the nonequilibrium green s function method in electronic transport @xcite , with fermions replaced by bosons as basic entities . although the techniques have been extensively described in the literature , most of them are centered around fermions . in this paper , we give a description of the method , emphasizing on both the theoretical formulation and computational implementation . we begin with an elementary introduction of the green s functions , including the contour - ordered green s functions . we then discuss equations of motion of the green s functions , feynman diagrammatic expansion , and dyson equations . we discuss the heat current and derive a formula for effective transmission when there are nonlinear interactions . after the introduction of the method , we present results of one - dimensional ( 1d ) chains , benzene ring , and carbon nanotubes , and discuss some of the interesting features in such systems . the nonequilibrium green s function methods are discussed in refs . @xcite , and equilibrium green s functions are explained in many textbooks such as refs . @xcite . for completeness , we introduce our notations and give a quick review of the definitions and properties of various green s functions in this subsection . we define the retarded green s function as @xmath0 \rangle,\ ] ] where the column vector @xmath1 with component @xmath2 is not the usual displacement operator from equilibrium but a renormalized one , i.e. , @xmath3 , such that the kinetic energy is of the form @xmath4 . the superscript @xmath5 stands for matrix transpose . the square brackets are the commutators . @xmath6 is a square matrix with elements @xmath7\rangle$ ] . the physical dimension of @xmath6 is time . by definition , @xmath8 equals zero when @xmath9 . in equilibrium or nonequilibrium steady states , the green s function depends only on the difference in time , @xmath10 . the fourier transform of @xmath11 is defined as @xmath12 = \int_{-\infty}^{+\infty}\!\!\ ! g^r(t ) e^{i\omega t}dt.\ ] ] note that we use square brackets to delimit the argument for the green s functions in frequency domain . the dimension of the retarded green s function in frequency domain is time squared . the inverse transform is given by @xmath13 e^{-i\omega t}d\omega.\ ] ] for notational simplicity , we ll set @xmath14 . we can always get the final required expressions by a simple dimension analysis . the rest of the definitions are the advanced green s function @xmath15 \rangle,\ ] ] the `` greater than '' green s function @xmath16 the `` less than '' green s function @xmath17 the time - ordered green s function @xmath18 and anti - time - ordered green s function @xmath19 the following linear relations hold both in frequency and time domains from the basic definitions : @xmath20 also , the relations @xmath21 and @xmath22 are useful . out of the six green s functions , only three of them are linearly independent . however , in systems with time translational invariance , the functions @xmath6 and @xmath23 are hermitian conjugate of one other : @xmath24 = ( g^r[\omega])^\dagger.\ ] ] so in general nonequilibrium steady - state situations , only two of them are independent . in this paper , we consider them to be @xmath6 and @xmath25 . there are other relations in the frequency domain as well : @xmath26^\dagger & \!=\ ! & - g^{<}[\omega],\\ g^r[-\omega ] & \!=\ ! & g^r[\omega]^*,\\ g^{<}[-\omega ] & \!=\ ! & g^{>}[\omega]^t \!=\ ! - g^{<}[\omega]^ * \!\!+\ ! g^r[\omega]^t \!\!-\ ! g^r[\omega]^*.\end{aligned}\ ] ] the last two equations show that we only need to compute the positive frequency part of the functions . in thermal equilibrium , there is an additional equation relating @xmath6 and @xmath25 : @xmath27 = f(\omega ) \bigl ( g^r[\omega ] - g^a[\omega ] \bigr ) , \label{grtogl}\ ] ] where @xmath28 is the bose - einstein distribution function at temperature @xmath29 . equation ( [ grtogl ] ) is obtained by writing the green s functions as a sum of energy eigenstates ( known as the lehmann representation ) . thus in equilibrium , there is only one independent green s function ; we take it to be @xmath6 . the contour - ordered green s function is a convenient book - keeping to treat the different green s functions in a concise notation . we can consider a contour - ordered green s function as a function @xmath30 with arguments @xmath31 and @xmath32 defined on the complex plane . the contour runs from @xmath33 slightly above the real axis to @xmath34 and loops back from @xmath34 slightly below the real axis to @xmath33 . the contour - ordered green s function can be mapped onto four different normal green s functions by @xmath35 , where @xmath36 , and @xmath37 is the time - ordered green s function , @xmath38 is the anti - time - ordered green s function , @xmath39 , and @xmath40 . in a noninteracting harmonic system with a hamiltonian @xmath41 the retarded green s function in the frequency domain is given by @xmath12 = \left ( ( \omega + i \eta)^2 i - k \right)^{-1 } , \label{gr - free}\ ] ] where @xmath42 is an identity matrix , and @xmath43 . adding a small @xmath44 helps to choose correctly the inverse fourier transform integration path in the complex @xmath45-plane , so that the retarded green s function has the required causal property , @xmath46 for @xmath47 . equation ( [ gr - free ] ) can be derived by an equation of motion method . we note that the retarded green s function is a symmetric matrix since @xmath48 is symmetric , but this feature is not preserved with nonlinear interactions . our system is a junction with a central region and two leads which serve as heat reservoirs . we treat the leads explicitly as quasi - one - dimensional periodic lattices . this setup is experimentally relevant and conceptually useful for computation . we consider non - conducting solid and treat only the vibrational degrees of freedom for heat transport . let the ( mass - normalized ) displacement from some equilibrium position for the @xmath49th degree of freedom in the region @xmath50 be @xmath51 ; @xmath52 , for the left , center , and right regions , respectively . the quantum hamiltonian is given by @xmath53 where @xmath54 , @xmath55 is a column vector consisting of all the displacement variables in region @xmath50 , and @xmath56 is the corresponding conjugate momentum . @xmath57 is the spring constant matrix and @xmath58 is the coupling matrix of the left lead to the central region ; similarly for @xmath59 . we note that the dynamic matrix of the full linear system is @xmath60 the nonlinear part of the interaction will take the form @xmath61 the quartic interaction is important in stablizing the system , as a purely cubic nonlinear interaction makes the energy unbounded from below . a schematic to illustrate the two adiabatic switch - ons . ] we need to answer an important question : what is the distribution of the system ? the distribution enters the definition of the green s functions as a density matrix @xmath62 for the average @xmath63 . for an equilibrium problem , this is just the boltzmann factor , but in a nonequilibrium situation , it is not known and must be computed in some way . the adiabatic switch - on gives us a clear conceptual framework how this problem can be solved , at least formally . we imagine that at @xmath64 the system has three decoupled regions , each at separate temperatures , @xmath65 , @xmath66 , and @xmath67 . the nonlinear interactions are turned off . the green s functions @xmath68 are known and take the form of eq . ( [ gr - free ] ) . the couplings @xmath69 and @xmath59 are then turned on slowly , and a steady state of the linear system is established at some time @xmath70 . the green s functions of the linear nonequilibrium system will be denoted by @xmath71 . for this linear problem , the result does not depend on @xmath66 . finally , the nonlinear interaction @xmath72 is turned on , and at time @xmath73 , a nonequilibrium steady state is established , see fig . [ fig - adiabatic ] for illustration . the full nonlinear green s functions will be denoted by @xmath74 . the density matrices at time @xmath75 , @xmath5 , and @xmath73 are related in the following way in the interaction picture : @xmath76 where @xmath77 is the time - order operator . in this subsection , we derive the equations of motion for green s functions . the equations of motion can be used to perform systematic perturbation expansion for the green s functions , or as a starting point for mean - field approximations . there are at least two ways to derive the equations of motion for nonequilibrium green s functions . the first is to derive the equations of motion for the time - ordered green s function and then generalize it to the contour - ordered version by evoking the structure isomorphism of the two sets of green s functions @xcite . another possibility is to consider directly the contour - ordered green s function @xcite . we ll take the latter approach . we define a general @xmath78-point contour - ordered green s function as @xmath79 the index @xmath80 labels the region , @xmath49 labels the degrees of freedom in that region , and @xmath31 is the contour variable . the function is symmetric with respect to simultaneous permutations of the triplet @xmath81 . we view the function as an analytic function ( at least along the path ) in variable @xmath31 which varies on the keldysh contour @xcite . we also introduce a generalized @xmath82 function which is defined to be 1 if @xmath31 is later than @xmath32 along the contour and 0 otherwise . its derivative is a generalized @xmath83-function , @xmath84 . the @xmath85 function can be used to represent the contour - order operator @xmath86 . for convenience , we allow for @xmath87 , which is just a constant @xmath88 . our definition is slightly different from that of the usual quantum field theory where our prefactor is always @xmath88 ( instead of the usual @xmath89 ) . to start with , we consider one of the important correlation functions for heat transport , @xmath90 . the first derivative with respect to the second argument always leads to a direct differentiation inside the contour - order operator since the equal - time coordinates @xmath1 commute , @xmath91 the second derivative is similar in this case as @xmath92 and @xmath93 also commute at equal time . thus , substituting the equation of motion for @xmath94 ( having identical form in quantum and classical cases ) , we get @xmath95 in matrix notation , it is @xmath96 to solve this differential equation on the contour , we define the ( contour - ordered ) green s function of the left lead to satisfy @xmath97 we can think of @xmath98 as the inverse of the operator @xmath99 . by multiplying @xmath100 to eq . ( [ eqg_cl ] ) from the right , @xmath101 to eq . ( [ eqg_l ] ) from the left , and then subtracting the two equations and integrating over the variable @xmath32 along the contour , we get ( after integration by part ) @xmath102 the reason that the terms in the second line are identically zero involves some considerations . first , since the central part and left lead are decoupled at the time @xmath103 , we must have @xmath104 and @xmath105 . this takes care of the lower bound . next , we require that the free left lead green s function satisfies the boundary condition @xmath106 , @xmath107 , at the upper limit @xmath108 . this requirement is equivalent to @xmath109 , and @xmath110 , or @xmath111 ( as well as that their first derivatives are zero ) . the condition @xmath112 is consistent with the causality requirement of the retarded green s functions , but the conditions for @xmath113 or @xmath114 are insufficient , as the information about temperature of the system is unspecified . we fix the ambiguity by requiring that @xmath115 be the equilibrium value at inverse temperature @xmath116 . in other words , @xmath117 should exactly be the contour - ordered green s function of the free lead as defined in subsec . [ sec - green ] . such a choice is consistent with the adiabatic switch - on . a full justification of the final result , @xmath118 can also be given from a perturbation expansion , as has been done in ref . @xcite . equation ( [ gcl ] ) can be generalized to the case of @xmath119 with a similar result . we now consider the green s functions of the form @xmath120 involving only the central part . the aim is to derive a recursion relation for the central part green s functions by eliminating references to full green s functions involving the leads . this type of green s functions involves two new features , ( 1 ) we must take care of the derivatives of the @xmath85 functions which produce commutators @xmath121 $ ] , ( 2 ) we generate higher order green s functions due to the nonlinear interactions . the contour order leads to @xmath122 terms , each of which is a permutation of the displacement variable @xmath1 from the original order @xmath123 , multiplied by a term of permutation of the arguments @xmath31 from the standard order @xmath124 . differentiation of these @xmath85 functions leads to a sum of contractions between the variable that we are differentiating and all the other variables . this reduces the order of the green s function by two . we note that @xmath125 = i \delta_{jl}.\ ] ] for the moment , we ll consider only the cubic nonlinearity . four terms are produced when @xmath126 is substituted inside the order operator , due to linear coupling @xmath127 , @xmath128 , @xmath59 , and nonlinear @xmath129 . the first three terms are similar in structure to that in eq . ( [ eqg_cl - component ] ) . the cubic nonlinear interaction increases the value of @xmath78 by 1 . we note that the equation for @xmath126 is evaluated at the same time @xmath31 . this would be complicated if we have to keep track of which @xmath31 s are equal in the green s functions . we transfer this information to the coupling @xmath129 to define the three - body interaction in contour variable form ( for the force ) @xmath130 since the contour integral is @xmath131 , we must define @xmath132 this is because our @xmath83-function has the standard property that @xmath133 , but when written in the ordinary time @xmath134 and index @xmath135 , we must have the extra @xmath135 factor to get the required value of @xmath136 . recursive expansion rule for green s functions . a vertex with @xmath78 double lines denotes an @xmath78-point green s function . the single line denotes @xmath71 . a single - line three - terminal vertex is associated with @xmath137 . ] the two terms involving the leads produce green s functions of the form @xmath138 and @xmath139 , which can be replaced by @xmath140 using the result eq . ( [ gcl ] ) derived earlier . we can now solve for the center - only green s function using the green s function of the linear system : @xmath141 we then obtain the following simple recursive rules for green s functions ( as shown in fig . [ fig - exprule ] ) : \(1 ) replace leg 1 by inserting a nonlinear coupling @xmath142 such that the outer leg is immediately connected with @xmath71 while the other two terminals increase the order of the green s function by 1 . quartic interaction is similar , but the process will increase the order by 2 . \(2 ) add imaginary unit @xmath143 times a sum of the @xmath144 graphs formed by pairing each leg with leg 1 and connecting with the propagator @xmath71 , multiplied by a @xmath145 order remaining green s function . \(3 ) symmetrize the graphs , if desired , i.e. , do steps ( 1 ) and ( 2 ) for every leg , 1 , 2 , @xmath146 , @xmath78 , add them up , and then divide by @xmath78 . the above rules may be conveniently implemented in a symbolic computer language , such as mathematica . the contour - ordered green s function can also be obtained by a perturbation expansion of the interaction picture evolution operator ( scattering matrix operator ) and is expressed as : @xmath147 where the displacements refer to the central region ; we have dropped the superscript @xmath148 for brevity . the operators in the top line are in the heisenberg picture , while that in the second line are in the interaction picture . the green s function @xmath71 of the linear system ( when @xmath149 ) can be computed from that of the free subsystems ( an integral equation form of eq . ( [ eq_g0 ] ) ) : @xmath150 where @xmath151 , @xmath152 similarly for @xmath153 . this dyson equation can be derived by considering the first step of the adiabatic switching - on process . since it is a linear system , the self - energy @xmath154 is known exactly . by expanding the exponential in eq . ( [ expand ] ) , a series in the nonlinear interaction strength is obtained . the reduction in terms of the unperturbed green s function @xmath71 relies on the fact that wick s theorem @xcite is applicable here . although a formal proof of wick s theorem is difficult , we can see that the density matrix @xmath62 must be quadratic in exponential . this is because the @xmath155 is quadratic and the system is linear . the fact that the equation of motion method and the perturbation expansion give identical results is a confirmation of the validity of wick s theorem . the expansion contains connected as well as disconnected feynman diagrams . the disconnected diagrams are constant in time , and give rise to a thermal expansion effect . we can show that these diagrams do not contribute to the thermal transport , as they are proportional to @xmath156 in the frequency domain . the thermal current formula has a factor of @xmath45 which makes it zero . finally , the connected part of the green s function satisfies a similar contour - ordered dyson equation relating @xmath157 to @xmath71 through a nonlinear self - energy @xmath158 : @xmath159 in ordinary green s functions and in frequency domain ( @xmath45 argument suppressed ) , the dyson equations have solutions @xcite : @xmath160 the connected part of the green s function @xmath157 will be used for @xmath74 in eq . ( [ heat - current ] ) below , and the extra subscript @xmath161 will be dropped from this point . the feynman diagrams for nonlinear self - energy @xmath158 with the prefactors for graphs of cubic and quartic interactions to second order in @xmath162 . ] in fig . [ fig - feynmant4 ] we give the feynman diagrams to `` second order '' of the nonlinear interactions for cubic and quartic terms , where the graphs have explicitly @xmath162 and @xmath163 in the coefficients ; the neglected graphs have higher powers of @xmath162 . these diagrams are the same as those in ref . @xcite , but the interpretation is different . it is for the contour - ordered green s functions here and for the retarded green s functions of the creation / annihilation operators in ref . @xcite . to find the actual expressions for each of the diagrams , we have the following simple rules : ( 1 ) label each interaction vetex line by the @xmath31 variables and site indices , with the interaction function @xmath164 ( or @xmath165 for the quartic interaction ) . ( 2 ) each line connecting the labels is associated with a propagator @xmath71 . ( 3 ) integrate over all the internal variables along the contour ; sum over the internal site indices . most of the integrations are easy , as @xmath166 contains two @xmath83-functions in @xmath31 . for example , the second graph is @xmath167 where @xmath168 and @xmath169 are external variables and the rest are internal . after integrations , we get @xmath170 now we can rewrite in real time @xmath134 and the branch index @xmath135 using the mapping @xmath171 , @xmath172 , and @xmath173 . noting that @xmath174 by time translational invariance , we have @xmath175 the result can now be transformed to frequency domain if desired . we define the current flow from the left lead to the central region as @xmath176 the interpretation of the current is somewhat subtle as the leads are semi - infinite in extent . the definition is meaningful . by the heisenberg equation of motion , we obtain , at @xmath73 , @xmath177 . the expectation value can be expressed in terms of a green s function @xmath178 . using the fact that operators @xmath1 and @xmath179 are related in fourier space as @xmath180 = ( -i\omega ) u[\omega]$ ] , we can eliminate the derivative and get , @xmath181\right ) \omega\ , d\omega.\ ] ] using the result derived earlier relating the mixed lead - center green s function to the center - only green s function and applying the langreth theorem @xcite to write contour - ordered green s function in ordinary green s functions in frequency domain , we have @xmath182 = g^r_{cc}[\omega ] v^{cl } g^{<}_l[\omega ] + g^{<}_{cc}[\omega]v^{cl}g^{a}_l[\omega]$ ] . the final expression for the energy current is @xmath183 \sigma^{<}_l[\omega ] + g^{<}[\omega ] \sigma^a_l[\omega ] \bigr ) , \label{heat - current}\ ] ] where the self - energy due to the interaction with the lead is @xmath184 . for simplicity , we have dropped the subscript @xmath148 on the green s functions denoting the central region . this result is the phonon analog of the electronic current formula @xcite . since the energy current must be conserved , we have @xmath185 . in addition , @xmath186 is real . we can obtain a symmetrized expression for the current @xmath187 where @xmath188 . we define the thermal conductance as @xmath189 where @xmath190 is the difference of the temperatures in leads , such that @xmath191 and @xmath192 . using eq . ( [ current-2 ] ) , we can actually take the limit @xmath193 by introducing variational derivatives of the green s functions , self - energies , or current as , e.g. , @xmath194 the difference in the numerator is the value of a function when the leads are at two different temperatures @xmath65 and @xmath67 , respectively , minus the value when the system is in equilibrium at @xmath195 . with this notation , we can express the conductance in a form similar to the landauer formula @xcite for the ballistic transport as @xmath196 \frac{\partial f(\omega)}{\partial t}\ ] ] with an effective transmission coefficient , @xmath197 & = & \frac{1}{2 } { \rm tr } \bigl\ { g^r ( \gamma_l + \frac{1}{2}\gamma_n - s ) g^a \gamma_r \bigr\ } + \nonumber \\ \label{eff - trans } & & \frac{1}{2 } { \rm tr } \bigl\ { g^a \gamma_l g^r ( \gamma_r + \frac{1}{2}\gamma_n + s ) \bigr\},\end{aligned}\ ] ] where the nonlinear effect is reflected in the extra terms , @xmath198 , and @xmath199 \left({\partial f \over\partial t}\right)^{-1}.\ ] ] equation ( [ eff - trans ] ) is a generalization of the caroli formula @xcite for ballistic transport . we ll discuss in sec . [ sec3b ] how the variational derivatives of the self - energy can be computed . mean - field approximation to the self - energy . the double line denotes the full green s function @xmath74 ; the single line is for @xmath71 . ] we have seen in ref . @xcite that high - order feynman diagrams are important at high temperatures . however , it is computationally very expensive to include these high - order diagrams , although we can do calculations on very small systems such as a three - site model . the standard technique is to use a mean - field approximation which in some sense partially takes into account these high - order graphs . we propose an approximation to the self - energy shown in fig . [ fig - mf ] , where the double line represents the nonlinear full green s function @xmath74 , while a single line is the usual bare green s function @xmath71 . with such a choice , we find that to next order , all the cubic interaction graphs are reproduced exactly ( including the prefactor ) , except that two of the diagrams are missing . with the self - consistent mean - field approximation , the solution has to be found iteratively . the convergence of the iterative process becomes an issue . with only cubic interactions , the system of equations will not converge at high temperatures . we understand that this is simply a consequence of the unstable nature when only cubic terms are used . we thus add the fourth order potential which is essential for stability . however , the convergence is still difficult at high temperatures , mainly due to the singular nature of the bose distribution . at high temperatures , the classical molecular dynamics ( md ) is a correct and numerically exact method for the heat conduction problem . however , the standard heat baths used in md such as nos - hoover do not correspond to exactly the action of the leads in our models . a correct heat bath should follow a langevin dynamics with colored noises . dhar @xcite has given a derivation for the junction problem . the idea is to solve the lead degrees of freedom @xmath200 and @xmath201 formally first , which is easy since it is a linear system . the result is then substituted into the equation for the center region . this gives the following equation : @xmath202 where @xmath203 is the nonlinear force , @xmath204 is the same retarded self - energy of the lead , @xmath205 , used in the nonequilibrium green s function calculation , but in the time domain . by integration by part , we could also cast eq . ( [ langevin ] ) in a standard generalized langevin form @xcite involving a damping force , but @xmath127 has to be shifted to @xmath206 $ ] . an extra contribution from the left lead is @xmath207 where @xmath208 is the retarded green s function of the left lead in the time domain . the expression for the right lead @xmath209 is similar . a time @xmath5 in the lower limit in the integral and eq . ( [ random - noise ] ) is some time in the remote past where the system is decoupled and the leads are in respective thermal equilibrium . we ll take the limit @xmath210 . we turn eq . ( [ langevin ] ) into a stochastic differential equation by requiring that @xmath211 and @xmath212 are random variables satisfying the boltzmann - gibbs distribution @xmath213 thus , @xmath214 is colored noise with the properties @xmath215 the last line of the above equation is a kind of `` fluctuation - dissipation '' theorem . we note that the dot in @xmath216 denotes the derivative with respect to @xmath5 . for a sensible heat bath , the noise should not depend on @xmath5 , but only on the difference @xmath10 . indeed , this can be done if we rewrite the result in vibrational eigen - modes . we can write the fourier transform of the noise correlation as @xmath217 where the square root of a matrix is defined by its eigenvalue decomposition taking only the positive eigen frequencies . this is well defined since @xmath218 is positive definite . the properties of the right lead noise @xmath209 are analogous . such a set of stochastic differential equations can be simulated on a computer . the green s functions of the free leads @xmath219 are required in calculation for the lead self energies @xmath154 . although the lead green s function is a semi - infinite matrix , due to a localized coupling @xmath69 or @xmath59 , only a corner set of elements is required . these submatrix elements are termed surface green s functions . the surface green s function can be computed rather quickly by recursive iterations @xcite or decimation . for completeness , we give an algorithm in pseudo - code form . the reader is referred to the literature for the derivation of such algorithm . the program takes three square matrices of equal size for @xmath220 , @xmath221 , and @xmath222 . the matrix @xmath220 is the block immediately adjacent to the center , while @xmath221 , and @xmath223 are repeated for the semi - infinite chain of the lead . they form the whole dynamic matrix of the lead in the following way : @xmath224 the left arrow ` @xmath225 ' below denotes assignment ; @xmath226 is an error tolerance . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ @xmath227 + @xmath228 + @xmath229 + @xmath230 + @xmath231^{-1}$ ] + do + @xmath232 + if @xmath233 exit + @xmath234 + @xmath235 + @xmath236 + @xmath237 + @xmath238 + @xmath239 + @xmath240 + @xmath231^{-1}$ ] + end do + @xmath241^{-1}$ ] . + _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ the diagrams in the mean - field approximation can be computed in the time domain easily and efficiently ; the convolutions in frequency domain become simple multiplications in time domain . thus we have adapted a fast fourier transform method for the calculation . this reduces the computational complexity , particularly for the quartic interaction terms . typically @xmath242 fourier grid points are needed for good convergence . using the message passing interface ( mpi ) , a parallel implementation with data partition good linear speed - up is obtained upto 64 processes . in the mean - field treatment of the heat transport , there is a need to find the variational derivatives of the green s functions and self - energy for the conductance calculation . this is done by solving iteratively the set of equations @xmath243 the expressions for @xmath244 and @xmath245 can be obtained by differentiating the self - energy results directly . for convergence control , linear mixing of the new and old values is used . one important question to ask about the nonequilibrium formulation is that whether the theory gives ballistic and diffusive heat transport in a single framework . clearly , if the nonlinear self - energy @xmath158 can be computed accurately , there is no doubt that the answer is yes however , this is not clear if only the leading order perturbation theory is used . in ref . @xcite , spohn gives a derivation of a boltzmann equation for the phonons in a simple 1d chain with onsite cubic nonlinearity . it is clear that in the boltzmann equation framework , we obtain fourier s law and diffusive heat transport . we demonstrate in this subsection that diffusive heat transport is not observed if only perturbative result is used . this subsection also serves as an example of illustration of the concepts discussed earlier by a simple and concrete model . we consider a 1d chain with inter - particle spring constant @xmath48 and onsite spring constant @xmath246 . we take a simple cubic onsite potential for the interaction , @xmath247 , for @xmath49 belonging to the central region . this means that @xmath248 . apart from the nonlinear interaction , the left lead , central region , and the right lead are identical . the classical equation of motion is @xmath249 where @xmath250 for @xmath251 and @xmath252 for @xmath253 or @xmath254 . the retarded green s function @xmath255 can be obtained by solving the linear equation @xmath256 g_0^r = i$ ] , where ( infinite in both directions ) matrix @xmath257 is @xmath258 on the diagonal and @xmath259 on the first off - diagonals . the explicit solution is @xmath260 = \frac { \lambda_1^{|j - l| } } { ( \lambda_1 - \lambda_2 ) k},\ ] ] where @xmath261 and @xmath262 are the roots of the equation @xmath263 note that @xmath264 . we define the two roots such that @xmath265 and @xmath266 . the surface green s function @xmath219 satisfies a similar equation as @xmath255 except that it is semi - infinite in extent . we consider the left lead ( @xmath267 ) . the result for the right lead is identical . since the matrix @xmath69 is nonzero only for one corner element , we need the @xmath268 component of the matrix @xmath219 . consider only the @xmath269 column , the equations for @xmath219 in component form are @xmath270 @xmath271 substituting the trial solution @xmath272 , we find that @xmath273 the `` less than '' green s function for the left lead is then ( c.f . ( [ grtogl ] ) ) @xmath274 where @xmath275 is the bose - einstein distribution function at the temperature of the left lead . the @xmath276 green s function can be obtained with the equation @xmath277 , @xmath278 ; @xmath279 , similarly for @xmath153 . the matrix elements of @xmath280 are all zero except that @xmath281 and @xmath282 . using these results , after some calculation , we can write @xmath283 = \frac { f_l \lambda_1^{j - l } + f_r \lambda_1^{l - j } } % { ( \lambda_1 - \lambda_2 ) k } \theta(\omega).\ ] ] the @xmath284 function is defined to be 1 inside the phonon band and zero outside the band ; more precisely , @xmath285 if @xmath286 , and is 0 otherwise . due to the onsite cubic nonlinearity , the leading order nonlinear self - energy calculation is simple . the first graph contribution is @xmath287 ^ 2.\ ] ] the contribution from the second graph is a constant in @xmath45 : @xmath288 = 2 i\ , t^2 \sigma \delta_{\sigma , \sigma ' } \delta_{jl } \sum_{\sigma'',s } \sigma '' { g_0}^{\sigma\sigma''}_{js}[0 ] { g_0}^{\sigma''\sigma''}_{ss}(0).\ ] ] both the summation over @xmath289 and over index @xmath290 can be performed analytically . we find @xmath288 = 2 i\ , t^2 \sigma \delta_{\sigma , \sigma ' } \delta_{jl } \bar{g}_j^r[0 ] \bar{g}^t(0),\ ] ] where @xmath291 & = & \frac { 1 + \lambda_1 - \lambda_1^j - \lambda_1^{n - j+1 } } % { ( 1-\lambda_1)(\lambda_1 - \lambda_2 ) k},\\ \bar{g}^t[\omega ] & = & \frac { 1 + ( f_l + f_r ) \theta(\omega ) } % { ( \lambda_1 - \lambda_2 ) k } , \\ \bar{g}^t(0 ) & = & \frac{1}{2\pi } \int_{-\infty}^{+\infty } \!\!\!\ ! \bar{g}^t[\omega ] d\omega.\end{aligned}\ ] ] for numerical accuracy , we compute the conductance directly instead of using finite differences . let the temperature on the left lead be @xmath292 and right lead be @xmath192 . we define the thermal conductance as that in eq . ( [ conductance - def ] ) and the effective transmission by eq . ( [ conductance-2 ] ) . for the expression for effective transmission @xmath293 $ ] , we also need to take the limit @xmath193 . the advantage of doing this is that the leading terms corresponding to equilibrium green s functions cancel exactly , since there should be no heat current when @xmath294 . we can also equivalently define the temperature variation through the expansion : @xmath295 and similarly for @xmath296 . we can simplify the effective transmission as @xmath197 & = & \frac{1}{2 } { \rm tr}\bigl\ { ( g^r - g^a)(\sigma_r^ { < } - \sigma_l^ { < } ) + \nonumber\\ & & \quad\quad\quad\quad i g^ { < } ( \gamma_r - \gamma_l ) \bigr\}/(f_l - f_r ) \nonumber \\ & = & i k ( { \rm im}\ , \lambda_1 ) ( f_{11}^l - f_{nn}^r ) , \end{aligned}\ ] ] where the @xmath297 function is defined as @xmath298 \left(\frac{\partial f}{\partial t}\right)^{-1}.\ ] ] @xmath299 is the same except that the first term is negative . in evaluating the @xmath297 functions , we should set @xmath195 as we have already taken the @xmath193 limit . the variations required in eq . ( [ f - vari ] ) can be computed by the variation in nonlinear self - energy , which in turn can be computed by the variation in the linear green s function @xmath71 . due to the dyson equation , the variation of the retarded green s function has a simple result : @xmath300 the variation in @xmath25 is rather complicated : @xmath301 \bigr\ } - { \rm h.c . } \nonumber \\ & & + \ , \delta g_0^ { < } + g^r ( \delta \sigma_n^ { < } + \sigma_n^r \delta g_0^ { < } \sigma_n^a ) g^a,\end{aligned}\ ] ] where @xmath302 , h.c . stands for the hermitian conjugate of the preceding term , and @xmath303 , @xmath304 & = & \frac{4i\ , t^2}{2\pi } \!\ ! \int_{-\infty}^{+\infty } \!\!\!\!\ ! d\omega'\ , { g_0}^{r,<}_{jl}[\omega-\omega ' ] \delta { g_0}^{<}_{jl}[\omega ' ] . \quad\end{aligned}\ ] ] thermal conductance of the cubic onsite model as a function of temperature for several values of the cubic coupling strength @xmath73 , 0.2 , 0.5 , 0.7 , 1.0 , and 2.0 ev/(@xmath305u@xmath306 ) . the length of center chain is @xmath307 . ] thermal conductance as a function of temperature for several system lengths @xmath308 with @xmath309 ev/(@xmath305u@xmath306 ) . ] we have made analytic calculations as far as possible . the rest of integrations and matrix inversions are calculated numerically . we use physical units with @xmath310ev/(@xmath311u ) , @xmath312 , ( u for unified atomic mass units ) . figure [ kappa ] shows the temperature dependence of the thermal conductance @xmath313 of a short nonlinear chain of length @xmath307 . the @xmath73 curve is the ballistic landauer formula result , the rest of the curves show the effect of nonlinearity . the prominent feature here is that the conductance decreases quickly as the nonlinear coupling @xmath134 increases from 0 to 2 . in particular , the system behaves like a thermal insulator if the on - site nonlinearity is large enough . the thermal conductance is small at low temperatures and at high temperatures , as was found experimentally . at the extremely high temperatures , the conductance decreases with temperature according to @xmath314 ( data not shown ) . this agrees with several other theories of thermal conductance at high temperatures . can we see diffusive transport in this model ( under the perturbative approximation ) ? the answer is no . in fig . [ size - datb ] , we show the size dependence for a fixed @xmath134 . at temperatures below 200@xmath315k , the results are essentially independent of the lengths , showing ballistic heat transport . at around 2000@xmath315k , from sizes 2 to 10 , we see decrease of the thermal conductance roughly as @xmath316 , but this trend does not continue , and the conductance saturates to some value for systems larger than 20 . this feature is generic ; the fpu model with first - order perturbation treatment also has a similar behavior . the curves also display some oscillatory structure . the number of oscillations appears to be roughly equal to half of the chain length @xmath308 . the feature is damped out as the chain becomes longer . it would be nice if a molecular dynamics simulation could confirm these features . the failure to see diffusive behavior may be understandable as our result takes into account only the first - order perturbation in the nonlinear interaction . other features seem qualitatively correct in comparison with real systems ; this is very encouraging . the fermi - pasta - ulam ( fpu ) model has been studied as a prototype model in nonlinear dynamics @xcite . in recent years , a lot of works have been done on the thermal transport in the fpu model @xcite . by molecular dynamics and mode - coupling theory , lepri _ et al . _ @xcite found that the thermal transport in such a system does not have an ordinary diffusive behavior in the sense that the thermal conductivity diverges with the system size as some power , @xmath317 . the actual value of the exponent @xmath50 is somewhat controversial . the earlier studies gave @xmath318 , but the latest analysis suggests @xmath319 @xcite . quantum thermal transport in fpu model has seldom been studied . in particular , the change from ballistic to diffusive transport has not been addressed . this question can not be correctly answered within molecular dynamics approach as the ballistic transport is quantum - mechanical in nature . we use realistic model parameters derived from expanding the morse potential , @xmath320 with @xmath321ev , @xmath322@xmath323 , up to the fourth order in @xmath324 , where @xmath325 is mass - renormalized displacement for the @xmath143th particle . it is a general fpu model with both cubic and quartic interactions . the leads do not have nonlinear interactions . we use the mass of carbon , @xmath326u , in the calculation . thus we can consider the results correspond to that of a carbon chain . thermal conductance of a fermi - pasta - ulam chain with chain lengths 4 , 8 , 16 , 32 , 64 , 128 , 256 , from top to bottom . semi - infinite one - dimensional harmonic chains serve as heat baths with spring constants @xmath327 and @xmath328 ev/(u@xmath311 ) for the left and right leads , respectively . small onsite quadratic potentials are applied with @xmath329 and @xmath330 ev/(u@xmath311 ) . ] in fig . [ fig - dexp ] we present the mean - field results of conductance @xmath313 as a function of temperature for different lengths of the chain . as we can see , below about @xmath331k , the conductance is independent of chain lengths ; this is the ballistic region of heat conduction . at high temperatures , the conductance decreases with the system size . for a completely diffusive heat transport , we should have @xmath332 . we are far from this . since in fpu model real diffusive behavior is impossible , the reduction in conductance in the classical regime should be @xmath333 . the observed exponent for sizes up to 256 is much smaller . in fact , this result is expected because the mean free path in a typical carbon systems ( such as carbon nanotubes @xcite or diamond ) is of order @xmath334 m . since the conductivity decreases with temperature typically as @xmath314 , even at few thousand kelvin , the mean free path is expected to be 100 to 1000 lattice spacings . thus we see only tendency towards diffusive behavior . it is also clear that ballistic transport and diffusive transport are at quite different time and length scales . the present calculation is already very large on an ibm 500-processor parallel computer . it is out of the possibility to see diffusive behavior for large sizes within the present approach . but this example does give us an indication that the mean - field theory could give diffusive behavior for appropriate systems . molecular dynamics results of the thermal conductance for the model as in fig . [ fig - dexp ] . ( a ) `` generalized langevin '' thermal heat baths from sizes 4 , 8 , 16 , , 2048 , from top to bottom . ( b ) nos - hoover thermostat with size 8 , 16 , 32 , , 1024 . ] it is instructive to compare the mean - field results with molecular dynamics results . in fig . [ fig - md ] , we present the md results with the generalized langevin - like heat - bath and the nos - hoover heat bath . the colored noise is implemented using a standard spectrum method with fast fourier transform @xcite . it turns out that the results for the thermal conductance is very sensitive to the bath used and detail couplings between the bath and system . it is clear that quantum nature shows up already as high as @xmath335k for this model chain . this can be understood by a very high debye temperature @xmath336 , where @xmath337 . the landauer formula , eq . ( [ conductance-2 ] ) , provides a rigorous upper bound for the thermal conductance ( when @xmath338 = 1 $ ] , @xmath339 ) , @xmath340 nw / k . interestingly , this is satisfied at low temperatures by the generalized langevin data . to what extent the mean - field theory is `` correct '' ? we can see that it is qualitatively correct for the accessible temperature and size ranges . however , if we trust the langevin heat - bath result , then , it appears that the mean - field approximation somehow over estimates the effect of nonlinearity . effective transmission of the benzene ring junction at a temperature of @xmath331k . the solid line is for a full model with force constants calculated from gaussian , while the dotted line is from a simplified 2d model . the vertical bars indicate the vibrational frequencies of isolated benzene rings of the full model ( topmost ) and simplified model ( dashed ) , respectively . ] molecular devices have attracted a lot of attentions in recent years , due to their potential for future generation electronics . experiments have been conducted on such systems ; theoretical analyses of their electronic and thermal transport properties have been performed @xcite . in this subsection , we examine the thermal transport properties of a benzene molecule connected to two polyethylene leads . two opposite sites of a benzene ring , which are normally the hydrogen atoms , are now bound covalently to carbon atoms , forming a 1,4-dibutyl - benzene molecule representing the whole system . the system has 14 carbon atoms and 22 hydrogen atoms . the interactions are calculated using gaussian 03 @xcite . we use the hartree - fock method with the 6 - 31 g basis set . we have also used a convenient new feature of gaussian , which can calculate anharmonic interactions up to the 4th order . [ fig : benzenetr ] ( solid line ) shows the effective transmission function @xmath293 $ ] of the benzene junction model at a temperature of @xmath331k calculated using the mean - field theory . the peak features are complicated , resulting from the vibrational modes of the molecule , the interaction with the leads , and the vibrational spectra of the leads , as well as the temperature effect . to understand the transmission function better , we correlate the peak features with the vibrational spectrum of an isolated benzene molecule calculated from gaussian , shown in fig . [ fig : benzenetr ] as solid bars . the transmission peaks are roughly located at where the vibrational frequencies appear , but somewhat shifted to low frequencies . it is difficult to identify the peaks with eigenmodes unambiguously . we notice that the hydrogen atoms do not play much role in the thermal transport . in particular , there are frequencies around @xmath341 associated with the vibrations of h atoms , but there is no transmission in these high frequencies . vibrations of h atoms have only local effects . we confirm that the h atoms are less important by a much simplified model retaining only the degrees of freedom of the carbon atoms in a 2d model . the leads are pseudo polyethylene chains ; we use an united atom approximation , so the ch@xmath342 group is treated as a single atom . the force constant between each atom is obtained from the dreiding model @xcite . there are only bond interactions in the leads , and we have used the morse function to calculate the bond potential , which has the form @xmath343 , where @xmath344(kcal / mol ) , @xmath345 , and @xmath346 is the displacement ( with the equilibrium bond length @xmath347 ) . for the benzene ring , we also include an angular potential , which has the form @xmath348 , where @xmath349 ( kcal / mol ) . we make a taylor expansion of the potential and calculate the force constants up to the 4th order . the transmission for the simplified model ( dotted line in fig . [ fig : benzenetr ] ) is smoother due to much reduced degrees of freedom . it is qualitatively in agreement with the full model . in this case , the peaks appear to be in good agreement with the vibrational modes of a hexagonal ring . thermal conductance @xmath313 as a function of temperature @xmath5 for a twenty - atom ( 5,0 ) carbon nanotube junction . ] thermal transport of carbon nanotubes has been studied by molecular dynamics @xcite , landauer formula ballistic transport @xcite , and a phenomenological theory taking into account the nonlinear effect @xcite . there are also experimental works on carbon nanotubes @xcite . we have also studied the heat transport properties of the @xmath350 carbon nanotube using a tight - binding ( tb ) method @xcite , of which the tb parameters for the carbon atom have been checked recently @xcite . the left lead , the central junction , and the right lead consist of forty , twenty , and forty atoms , respectively . a unit cell of the @xmath350 nanotube has twenty atoms . the tube s axial dimension is long enough ( about @xmath351 ) to allow a single @xmath352 point to be used for the @xmath353-point sampling . we have written a force subroutine in order to perform atomic relaxation of the nanotube as well as to calculate force constants by numerical finite differences . the spring constants @xmath354 , the nonlinear force constants @xmath129 and @xmath355 are systematically obtained by displacing one , two , and three atoms , respectively . for each displacement , the forces acting on the atoms are used in the central - difference scheme to calculate the force constants . since the tb method we have adopted is free from numerical noises , we have used a reasonably small maximum displacement of @xmath356 of @xmath357 ( i.e. , the atoms are displaced not more than @xmath356 from their equilibrium positions ) . we note that this value of @xmath356 is two orders of magnitude smaller than that commonly used in a supercell force - constant method @xcite . in fig . [ fig - ntb ] we present the results of thermal conductance calculations . we find that for such a complicated system , the mean - field iterations are hardly converging , thus we only give the ballistic ( where the nonlinear interactions are turned off ) , and a first - order perturbative result using the ( single - line ) diagrams in fig . [ fig - mf ] . it is clear that the nonlinear interactions have a large effect even at moderately high temperatures . the thermal conductance is slightly peaked around @xmath358k for the nonlinear result . we expect that the result should be reliable at least up to room temperatures . we presented a nonequilibrium green s function method for thermal transport in nano - junctions . we gave a unique derivation of the equations of motion of green s functions and rules for generating and computing feynman diagrams in a contour - ordered green s function setting . the nonlinear interaction was treated with a systematic perturbative expansion for self - energies . for practical calculations , the mean - field approximation has to be adopted . we are still left with a technical difficulty that the mean - field equations sometimes do not converge . several versions of computer programs have been written for consistency checks ( a mathematica notebook for an initial fast prototyping , a fortran program with mpi for very general models , and a specialized program for the 1d cubic onsite model ) . we have applied the programs to a number of model systems . several sophisticated ` first - principles ' methods were used in the force - constant calculations for realistic models . for the fpu model , we can accurately predict the ballistic thermal transport and crossover to diffusive behavior at sufficiently high temperatures . we proposed to use generalized langevin heat baths , which are more appropriate for nanojunctions . for the benzene ring model , we observed interesting behaviors in the transmission function related to the atomic vibrational modes of the molecule . for carbon nanotubes , it is clear that nonlinear effect is quite strong even at low temperatures . the green s function discussed here is more fundamental than the phonon distribution function in the boltzmann equation . the connection of the present formulation with the quantum boltzmann equation or the phonon boltzmann equation is something worth pursuing . formally , they are related by the wigner transformation . if we can reduce the matrix of the green s functions to vector of distribution functions ( in momentum space ) , potentially much larger systems can be treated if we are willing to forsake some atomic details . clearly , the nonequilibrium green s function method is at its best for small nanoscale systems . a combination of the present method with electronic transport and electron - phonon interactions is worthy of further explorations . we thank lin yi for introducing us the nonequilibrium green s function method . we also thank hong guo , qinfeng sun , sai kong chin , and jingtao l for discussions . c. k. g. thanks d. a. papaconstantopoulos , l. shi , and m. j. mehl for a copy of the tight - binding program _ static _ and various initial help with the use of the program . this work was supported in part by a national university of singapore academic research fund . the computations were carried out on the institute of high performance computing ibm clusters and on the singapore - mit alliance opteron cluster . y. meir and n. s. wingreen , phys . lett . * 68 * , 2512 ( 1992 ) ; a .- jauho , n. s. wingreen , and y. meir , phys . b * 50 * , 5528 ( 1994 ) . h. haug and a .- jauho , _ quantum kinetics in transport and optics of semiconductors _ ( springer , berlin , 1996 ) . r. e. cohen , m. j. mehl , and d. a. papaconstantopoulos , phys . b * 50 * , 14694 ( 1994 ) ; m. j. mehl and d. a. papaconstantopoulos , phys . rev . b * 54 * , 4519 ( 1996 ) ; d. a. papaconstantopoulos , m. j. mehl , s. c. erwin , and m. r. pederson , mater . * 491 * , 221 ( 1998 ) .

the large aperture of the _ james webb space telescope _ ( _ jwst _ ) in combination with its sensitive near - infrared ( near - ir ) and mid - infrared ( mid - ir ) instruments will provide new opportunities to study comet dust , gas , and nuclei at moderate and large heliocentric distances ( defined here as beyond 3 au ) , and at excellent spatial resolutions for closer objects . moreover , _ jwst _ s operational strategies , scheduling , and instrument capabilities allow us to follow comets over a wide range of times and heliocentric distances . nominal performance expectations for _ jwst _ hardware and flight software will enable the telescope to follow moving targets with apparent rates up to @xmath1 s@xmath2 with small pointing errors ( currently specified to be 17 mas rms at 0030 s@xmath2 ) . _ jwst _ will use the jpl horizons system ephemerides for pointing and tracking of comets with known orbital elements , although an observer may provide ephemerides for targets not in the jpl database . accurate ephemerides and/or peak - up operations are essential to place comets and other solar system small bodies in the narrow ( @xmath3 ) spectrometer slits . observing visits are limited to the dwell time of the guide star on the fine - guidance detectors , which have a field - of - view of @xmath4 arcmin . non - sidereal guiding strategies are beginning to be developed and refined by the _ jwst _ team and will be verified during the initial observatory commissioning period . release of moving target capabilities is anticipated to be at the start of routine science operations . see @xcite and @xcite for summaries of solar system science with _ the study of comets with _ jwst _ will address current key decadal questions in planetary science , as well as astronomy and astrophysics , that are challenging to address with ground - based observations alone . these questions pertain to the initial stages of solar system formation and its subsequent evolution , and include : `` what were the initial stages , conditions , and processes of solar system formation and the nature of the interstellar matter that was incorporated ? '' `` what were the primordial sources of organic matter , and where does organic synthesis continue today ? '' `` how have the myriad chemical and physical processes that shaped the solar system operated , interacted , and evolved over time ? '' @xcite . here , we highlight some of the observational challenges , experimental techniques , and representative science opportunities that _ jwst _ offers as a tool to improve our knowledge of comets . such a space - based platform has easy access to key cometary emission bands from h@xmath0o , co@xmath0 , co , and ch@xmath5 that would otherwise be hindered by severe telluric absorption . except for co@xmath0 , studies of these latter molecules are possible from ground - based telescopes , e.g. , through photodissociation products ( e.g. , oh and o from h@xmath0o , o from co@xmath0 ) , through non - resonance emissions ( e.g. , water hot - bands ) that are not blocked by the earth s atmosphere , or individual rovibrational lines doppler - shifted from their terrestrial counterparts ( for a review , see * ? ? ? . however , having direct access to the fundamental bands , and without doppler shift restrictions , can enable new and exciting results . above the atmosphere , background emission is greatly reduced , allowing for greater sensitivity to the 3- region and beyond , enabling survey studies of dust emission , water ice absorption , or ( refractory ) organic features that might otherwise take years to achieve from ground - based telescopes . time - domain _ jwst _ observational campaigns covering a wide range of heliocentric distances become possible for many more targets , rather than just the occasional hale - bopp - class comet . in section [ sec : brightness ] , we present our methods to estimate the continuum arising from dust ( reflected and thermal emission ) , and to estimate the strengths of emission bands from molecules . these methods are used to determine integration times and signal - to - noise ratios throughout our paper . we then propose four science examples that _ jwst _ will be especially suited to address . in section [ sec : gas - coma ] , we discuss the telescope s ability to measure the primary drivers of comet activity : h@xmath0o , co@xmath0 , and co. in section [ sec : dust ] , we summarize simultaneous observations of water gas and dust comae and how they can be used to assess nucleus heterogeneity . in section [ sec : ice ] , we propose observations that can address the nature of ice in cometary comae and nuclei . finally , in section [ sec : faint ] , we demonstrate how _ jwst _ can be used to detect gas in main belt and other faint comets . the effects of the observatory s non - sidereal tracking limit on observations of comets , and specific comet observing opportunities are summarized in section [ sec : targets ] . at _ jwst _ wavelengths ( @xmath6 ) , comet spectra are generally dominated by sunlight scattered by and thermal emission from coma dust . it is a challenge for observers to accurately predict the surface brightness of a given target , due to each comet s inherent variability and coma physical properties ( dust and gas ejection speeds , grain parameters , presence of ice , etc . ) . for comets in the inner solar system , the observed energy from dust thermal emission and that from light scattered by dust are approximately equal in the 24 range , thus a single planck function or stellar template are poor approximations . whether the goal is to observe the continuum from dust , absorption or emissivity features , or gas emission , having a simple model to estimate continuum brightness is beneficial to observers interested in this wavelength regime . light scattered by comet dust can be estimated using the @xmath7 parameter of @xcite . @xmath7 is the product of the albedo , @xmath8 , the dust filling factor , @xmath9 , and the radius of the circular aperture in consideration , @xmath10 . under certain assumptions and conditions , it is proportional to the dust production rate @xcite . it carries the units of @xmath10 , and is typically expressed in cm , @xmath11 where @xmath12 is the observer - comet distance ( cm ) , @xmath13 is the heliocentric distance of the comet ( au ) , @xmath14 is the observed flux density of the comet within a circular aperture ( w m@xmath15 @xmath2 ) , @xmath10 is the radius of the aperture projected to the distance of the comet ( cm ) , and @xmath16 is the flux density of sunlight ( w m@xmath15 @xmath2au@xmath15 ) . we have written the albedo as @xmath17 to emphasize that the observed albedo is a function of phase angle ( sun - comet - observer angle , @xmath18 ) . for observations corrected to a phase angle of 0 , we will simply write @xmath7 . the albedo of @xcite is a factor of 4 larger than the geometric albedo , @xmath19 , of @xcite : @xmath20 . throughout the paper , we use the combined halley - marcus phase function from d. schleicher to scale @xmath21 to @xmath7 @xcite . for quantifying the thermal emission , @xcite introduced the parameter @xmath22 , which is the thermal emission equivalent to @xmath7 : @xmath23 where @xmath24 is the apparent emissivity , @xmath14 is now the observed thermal emission from dust ( w m@xmath15 @xmath2 ) , and @xmath25 is the planck function evaluated at temperature @xmath26 ( w m@xmath15 @xmath2 , and k , respectively ) . similar to @xmath7 , this quantity carries units of length , determined by the units of @xmath10 . @xcite recommend using the effective ( color ) temperature of the continuum , with a default temperature of @xmath27 , 10% warmer than an isothermal blackbody sphere in local thermodynamic equilibrium with insolation ( @xmath28 ) . for exposure time calculations , we propose using a combination of the two empirical quantities above : @xmath29 where @xmath30 is the phase function of the dust evaluated at the phase angle @xmath18 . the filling factors for @xmath7 and @xmath22 are not necessarily the same , as different populations of dust could dominate the measured scattered and thermal emission . in table [ tab : ef2af ] we present seven observations of five comets with contemporaneous @xmath7 and @xmath22 estimates . the ratios @xmath31 aspan the range 2.3 to 4.2 . when @xmath22 is not known , we suggest scaling @xmath7 by the ratio @xmath32 , based on the observed range in our small sample , and the assumptions : ( 1 ) @xmath33 , ( 2 ) @xmath34 , and ( 3 ) the emission and scattering filling factors are similar , @xmath35 . because the @xmath22 parameter depends on effective temperature , the ratio @xmath36 may also be correlated with @xmath37 , although this is not evident in the limited dataset in table [ tab : ef2af ] . a caveat to our approach arises if @xmath37 and @xmath22 rely on data measured over a limited wavelength range . any derived values may not be valid for other wavelengths , or even heliocentric distances . specifically , short ward of @xmath387.5 silicates have limited absorptivity so the 37 continuum is dominated by a warmer continuum from more absorbing carbonaceous grains . in fig . [ fig : sw3 ] , we plot a near - ir spectrum of comet 73p - c / schwassmann - wachmann 3 @xcite , and our proposed model with the parameters @xmath39 , @xmath7=1400 cm , and @xmath40 . the agreement is acceptable for first - order integration time estimations . further refinements , e.g. , wavelength dependent @xmath7 to account for coma color , or wavelength dependent @xmath22 to account for grain composition and size distribution , are not necessary . in addition to the dust continuum , gas emission bands are commonly observed . here , we summarize our method to estimate total band flux . using fluorescence @xmath41-factors ( photons s@xmath2 molecule@xmath2 at 1 au ) from @xcite , we generate total band fluxes in a 04 radius aperture , for either the nirspec integral field unit or micro - shutter assembly , assuming an optically thin coma ( table [ tab : gas - flux ] ) , @xmath42 where @xmath43 is the production rate of the molecule ( molecules s@xmath2 ) , @xmath10 is is the radius of the aperture projected to the distance of the comet , @xmath44 is the planck constant , @xmath45 is the speed of light , @xmath46 is the band central wavelength , @xmath47 m s@xmath2 is the expansion speed of the gas , @xmath12 is the observer - comet distance , and @xmath13 is the heliocentric distance of the comet ( au ) . all parameters have mks units , unless otherwise noted . although we assume an optically thin coma , optical depth effects must be carefully considered when interpreting real spectra . for comet 9p / tempel 1 , @xcite showed that optical depth effects are significant inside of @xmath3810 km for a water production rate of @xmath48 molecules s@xmath2 and a co@xmath0 production rate of @xmath49 molecules s@xmath2 at 1.5 au from the sun . for integration time estimation purposes , we neglect line opacity effects , which should only affect total band fluxes at the 10% level or less ( e.g. , * ? ? ? a _ jwst _ exposure time calculator could provide a two - component model that would be adaptable to a variety of solar system sources . continuum spectra of comet comae , comet nuclei , and asteroids can all be approximated by : ( 1 ) @xmath50 , based on a solar - type spectrum , potentially reddened , representing the reflected or scattered light ; and ( 2 ) @xmath51 , a blackbody spectrum of arbitrary temperature , representing the thermal emission , @xmath52 f_{\lambda , therm } & = c_{therm } \frac{b_\lambda(t,\lambda)}{b_\lambda(t,\lambda_{therm,0})}\end{aligned}\ ] ] where @xmath53 and @xmath54 are the reflected ( or scattered ) and thermal emission normalization factors ; @xmath16 is the solar flux density ( see , e.g. , * ? ? ? * ) ; and @xmath55 is the planck function . the template spectra are normalized to 1.0 at reference wavelengths @xmath56 and @xmath57 . because @xmath7 and @xmath22 can change with wavelength , their values and the normalization wavelengths should be carefully chosen for the observation in consideration . for example , one might choose @xmath58 and @xmath59 for the spectrum in fig . [ fig : sw3 ] . the normalization factors for a comet coma model would be based the @xmath7 and @xmath22 components in eq . [ eq : comet - model ] , @xmath60 c_{therm } & = \epsilon f \rho \frac{\pi \rho}{\delta^2}\ .\end{aligned}\ ] ] thus , comet - specific parameters would not need to be incorporated into an observing tool , but arbitrary scale factors would be a necessity . observers will need to estimate @xmath7 and @xmath22 based on the literature , their own experience , or recent observations . three gas species , h@xmath0o , co@xmath0 , and co , are primarily responsible for driving cometary activity . these molecules have very different levels of volatility ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) , suggesting that each species should dominate the comet s activity at different parts of the orbit , delineated by the abundance of each ice , and the amount of solar energy available for sublimating those ices . co may be released at heliocentric distances out to tens of au , while co@xmath0 sublimation dramatically increases around 68 au and h@xmath0o sublimation near 23 au . however , the limited observations available for these species indicate that cometary behavior is more complex than would otherwise be suggested by a simple energy balance model . the relative abundances not only differ from comet to comet , but also in unexpected ways with respect to heliocentric distance and sometimes before and after perihelion in a single comet @xcite . these variations are due presumably to inherent compositional differences between objects reflecting their origins in the early solar system , heterogeneities in individual nuclei , and evolutionary processing from insolation . comprehensive data sets , i.e. , those covering the production of all three molecules ( h@xmath0o , co@xmath0 , and co ) over a wide range of times or heliocentric distances , typically depend on observations from multiple instruments and techniques . how systematic uncertainties ( instrumental , observational , or theoretical ) affect the relative abundances is therefore of concern . in addition , co@xmath0 can not be observed from the ground , thus little is known regarding its abundance in comets or its role in comet activity outside of recent snapshot surveys @xcite and missions to comets @xcite . the wavelength coverage and sensitivities of nirspec , nircam , and niriss will allow _ jwst _ to simultaneously or contemporaneously measure all three of the primary species over a range of heliocentric distances . such measurements enable studies of activity and composition at various distances and points around a comet s orbit , providing a uniform dataset that can not be achieved by other means . the high - spatial - resolution imaging available with nircam and niriss will also allow the investigation of heterogeneities in these species , revealed in the spatial distribution of the gases around the nucleus and their variations with time . here , we demonstrate the capabilities of time - domain spectroscopy with nirspec . our goal is to observe h@xmath0o , co@xmath0 , and co in a single comet over a wide range of heliocentric distances . we take the orbit of c/2013 a1 ( siding spring ) as a baseline ( perihelion distance @xmath61 au ) to provide a rough idea for potential observing windows and comet brightnesses during an oort cloud comet s journey from 10 au pre - perihelion to 10 au post - perihelion . we assume a coma mixing ratio of 100/10/10 for h@xmath0o / co@xmath0/co at @xmath62 au , approximately equal to the average observation listed in @xcite . using the sublimation model of @xcite , this ratio corresponds to equivalent active areas of 6 km@xmath63 for h@xmath0o , 0.26 km@xmath63 for co@xmath0 , and 0.080 km@xmath63 for co. other model quantities are summarized in table [ tab : gas - model ] . for the sublimation model , we assume a rapidly rotating nucleus with a visual albedo of 0.05 , an infrared emissivity of 1.0 , and the pole direction perpendicular to the orbit . gas production rates are the product of the sublimation rate and the active area . water molecules have the shortest photodissociation length scale of the molecules in our study , but our apertures are no larger than 13% of this scale , so for simplicity , we neglect photodissociation . the continuum arising from dust , a source of noise , is modeled using eq . [ eq : comet - model ] , with the following parameters : @xmath64 cm at 1 au , scaling with @xmath65 , @xmath66 . production rates and @xmath7 values derived from the gas and dust models are listed at approximately 1 au steps in table [ tab : gas - flux ] . we have selected epochs that would be within _ jwst _ s solar elongation constraints ( 85 to 135 ) . the water production rate ( @xmath67 molecules s@xmath2 ) and @xmath7 ( 925 cm ) at 1.5 au are within a factor of two of actual measurements of the comet near perihelion @xcite . comet siding spring s @xmath7 at 4 au ( pre - perihelion ) observed by @xcite is higher than our model by an order of magnitude ( 120 cm versus 2000 cm ) . dynamically new comets tend to have shallow lightcurves upon approach to perihelion @xcite , and we do not incorporate this aspect into our model . using the prototype nirspec exposure time calculator , we compute the integration time required to achieve a peak signal - to - noise ratio ( snr ) of 10 for each band ( table [ tab : gas - flux ] ) and plot it in fig . [ fig : gas ] . for our low - resolution spectroscopy observations , we assumed each band may be approximated by a single line with a full - width at half - maximum of 0.2 . both h@xmath0o and co are detected in a few thousand seconds at production rates @xmath68 molecules s@xmath2 out to @xmath384.5 au ; co@xmath0 is detected in a few thousand seconds at @xmath69 molecules s@xmath2 out to @xmath387 au . thus , _ jwst _ will easily be the best telescope for spectroscopic studies of comet activity at moderate and large heliocentric distance . comet nuclei are aggregates of the planetesimals present in the outer solar system during the epoch of planet formation @xcite . there is an open question concerning the homogeneity of accreted planetesimals for particular comets , i.e. , are individual comets aggregates of a heterogeneous population of planetesimals formed over a range of times and distances ( e.g. , * ? ? ? are comets collisional fragments ( e.g. , * ? ? ? * ; * ? ? ? * ) and less likely to be heterogeneous despite their origins ? high spatial resolution images and spectral data cubes from flyby spacecraft have suggested comet volatiles are not uniformly distributed about nuclei ( e.g. , * ? ? ? * ; * ? ? ? * ) , but the evidence for regions heterogeneous in dust properties relies on poorer resolution telescopic data , usually through multiple observations of a rotating comet ( e.g. , * ? ? ? * ; * ? ? ? a close approach ( here , @xmath70 au ) between a comet and _ jwst _ would allow us to observe coma features on @xmath71 km spatial scales , potentially revealing distinct active regions before they spatially mix in the coma . such a study can test the heterogeneity of comet nuclei by observing and comparing multiple spatially resolved coma jet features . _ jwst _ s miri instrument will be able to simultaneously observe both coma dust and water gas . at @xmath72 , mid - infrared spectroscopy reflects a coma s dust composition , grain sizes , and porosities through solid - state emission features and thermal emission pseudo - continuum . we use the terms grain and aggregate interchangeably . they represent an entire particle of dust , which may be composed of smaller sub - units ( the latter is implied for aggregate ) . typical dust features include a broad plateau from 8 to 12 , primarily arising from anhydrous amorphous silicates ( non - stochiometric olivine and pyroxene ) , and narrow emission peaks ( @xmath73 ) from crystalline silicates , dominated by mg - rich olivine @xcite . miri also covers the @xmath74 water gas emission band at 6 , observed and studied in several comets with the _ infrared space observatory _ and _ spitzer space telescope _ in addition , the nirspec instrument holds the potential to simultaneously detect emission from water gas and hydrocarbons , the latter via the aromatic c h stretch at @xmath383.3 and the aliphatic c h stretch at @xmath383.4 ( see * ? ? ? * ) . we consider comet 46p / wirtanen as an example exercise demonstrating the technical feasibility of an observing program designed to test coma heterogeneity . this comet has a close approach to _ jwst _ in december 2018 . in november 2018 the comet is within the solar elongation constraints ( 85 and 135 ) and quite bright , with integrated flux densities of order 1 jy at 13 ( 05 radius aperture ) . but rather than consider an observing window so soon after _ jwst _ s launch ( october 2018 ) , we instead consider the post - closest approach window starting 2019 march 13 ( @xmath75 au , @xmath76 au ) with non - sidereal rates of @xmath77 hr@xmath2 , well under the _ jwst _ limit of 108 hr@xmath2 . spatial - spectral maps of a comet s dust coma can be performed with either the low resolution spectrometer ( lrs , @xmath78 ) or medium resolution spectrometer ( mrs , @xmath79 ) modes of miri . typically , comet dust studies would use the lrs , because the spectral resolving power of the mrs is much greater than necessary to resolve solid - state emission features . however , we will design an observation for the mrs integral field units ( ifu ) , which can efficiently map a @xmath80 field of view , and have a spectral resolving power appropriate for measuring individual lines in the @xmath74 water band . for our model , we adopt an effective nucleus radius of 0.6 km @xcite , a maximum @xmath7 of 380 cm @xcite , and a maximum water production rate @xmath81h@xmath0o@xmath82 molecules s@xmath2 at perihelion @xmath83 au @xcite . we also assume that the dust production rate scales with water production as @xmath84 @xcite . the ifus can obtain complete spectra from 5 to 28.5 in 3 exposures over a @xmath85 field of view , corresponding to 1500 km @xmath86 1900 km at the comet , with a spatial resolution of about 100 km per resolution element . for comparison , the current best spatial resolution of this comet obtained with mid - infrared spectroscopy is 1300 km with _ spitzer_/irs . comet wirtanen s peak surface brightnesses estimated for the mrs ifus are listed in table [ tab : miri - est ] . the surface brightness is an average within a 04 radius aperture ( corresponding to approximately 10 resolution elements at 6 and a few resolution elements at 23 ) . using the estimated mrs line sensitivities ( in units of w m@xmath15 pixel@xmath2 ) and dividing by the unresolved line width for each module yields continuum sensitivity thresholds in units of flux density . these values are scaled by @xmath87 to an integration time of 100 s and listed in table [ tab : miri - est ] . we expect high signal - to - noise ratios in the central aperture , easily enabling mapping of the inner coma . for the h@xmath0o band at 6 we assume an effective @xmath41-factor of @xmath88 photons s@xmath2 molecule@xmath2 at 1 au for the brightest lines , based on modeled _ spitzer _ spectra of c/2003 k4 ( linear ) , 71p / clark , and c/2004 b1 ( linear ) @xcite . these lines will be affected by optical depth effects , but introducing opacity is not necessary for our first - order feasibility exercise . following eq . [ eq : gas ] , we predict a total line flux @xmath89 w m@xmath15 in a 04 radius aperture , or about @xmath90 w m@xmath15 pixel@xmath2 . the lines will be easily detected above the continuum ( @xmath91 w m@xmath15 pixel@xmath2 ) with the mrs ifus . the morphology of the water gas would be used to locate active areas on the nucleus , and to identify dust that may originate from those areas . additional context with co@xmath0 and co may be obtained with immediate follow - up observations with the near - ir spectrometer . comparisons of a particular active area s dust composition to that of the ambient coma or of other regions can be used to test the heterogeneity of the nucleus . as preserved leftovers from the formation of the solar system , comets serve as one of the best probes for studying primitive water ice in the form it may have existed in the early outer solar system . thus observations of water ice in comets may allow us to test the accretion processes that led to the formation of comet nuclei and therefore of planets ( e.g. , * ? ? ? . contrasting coma ice properties to observations of ice on comet surfaces will allow us to study coma - nucleus interactions , and nucleus surface evolution . water ice has absorption bands in the near - infrared at 1.5 , 2.0 , and 3.0 . the relative strength and shape of these features provides information on aggregate porosities , presence of dust within the ice , total abundance , and the size of the scattering units ( following section [ sec : dust ] , we use the term grain and aggregate interchangeably ) . laboratory measurements show that infrared water - ice absorption bands change position and shape as a function of phase ( crystalline or amorphous ) and temperature : e.g. , crystalline water ice differs from its amorphous counterpart by the presence of a sharp and narrow feature at 1.65 apparent at @xmath92 k @xcite . the best approach to investigate water - ice physical properties is through near - infrared ( 15 ) spectroscopy . water ice has been observed in comae using ground- and space - based telescopes and flyby spacecraft ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? the small sample of comae detections presently available has revealed a range of water - ice characteristics : and sub- water - ice particles have been identified , as well as one case of crystalline water ice ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ; * ? ? ? . however , the limited number of detections prevents development of a classification scheme or taxonomy based on water - ice properties . the sample population is too small to apply statistical analyses with confidence , thereby limiting our understanding of the origins of this diversity . it may relate to the different mechanisms responsible for delivering water ice from the interior into the coma ( sublimation of more volatile ices e.g. , co@xmath0 , co , or through large outbursts in activity ) . alternatively , it may be the outcome of a preserved comet - to - comet heterogeneity . establishing these connections is crucial if water - ice properties are to be used as observationally imposed boundary conditions on comet nucleus formation models . on nucleus surfaces , however , water ice has only been observed with flyby and orbiting spacecraft . the observations of ice on the surfaces of comets 9p / tempel 1 and 103p / hartley 2 point towards sub - units with sizes of order 10 to 100 @xcite . this is much larger than the coma ice summarized above , and the ice excavated from the interior of tempel 1 @xcite . thus , @xcite hypothesize that this ice is more likely due to re - condensation of water gas , rather than recently uncovered interior ice . observations of bare nuclei can help test when this re - condensation occurs : does it require constant replenishment from an active source , or is it deposited by low activity levels as a comet recedes from the sun ? complicating matters , @xcite report the detection of a broad absorption band at 2.9 to 3.6 in _ rosetta _ observations of the surface of comet 67p / churyumov - gerasimenko . the absorption band has a peak depth of 20% located at 3.2 to 3.3 . they suggest this feature is due to the presence of aromatic and aliphatic c - h bonds , carboxylic groups , and/or alcoholic groups . the absorption feature has not been reported on the surfaces of 9p / tempel 1 or 103p / hartley 2 @xcite . whether absorption features are due to water ice or organic molecules , the study of comet surfaces at 34 is best addressed through space - based missions or observatories . nirspec is ideal for the study of water ice in comets . the prism mode covers 0.7 to 5 at a spectral resolving power of @xmath93 , ideal not only to detect the broad absorption features of surface and coma ice but also to determine its chemical phase ( amorphous or crystalline ) . the instrument s higher spectral resolving powers ( @xmath94 ) may be sensitive to trapped co@xmath0 or other volatiles . below we summarize observations of a distant comet nucleus and coma with this instrument to illustrate future science opportunities . at 5 au and beyond , water ice deposited on the cold surface of a comet could persist until vigorous sublimation returns when surface temperatures rise to 160 k or more . where that ice resides would depend strongly on deposition and insolation history , including latitude and surface roughness considerations . cooler nuclei in the outer solar system , e.g. , cometary centaurs , may have surface ice present throughout their orbit . as an example of nirspec s capabilities with respect to spectroscopy of distant nuclei , we consider comet 172p / yeung . this comet was observed to be a point source by _ spitzer _ @xcite at @xmath95 au ( pre - aphelion ) , with an effective nucleus radius of 5.7 km @xcite . we model the comet nucleus thermal emission with the near - earth asteroid thermal model of @xcite and the parameters of @xcite . the scattered light contribution is modeled assuming a geometric albedo of @xmath96 @xcite . our goal is to detect the presence of water ice on a surface . we choose the 3- water ice because of its high relative absorption compared to the other water ice bands in the near - infrared . the same methodology could be applied to the detection of surface organics , similar to those seen at comet 67p / churyumov - gerasimenko @xcite . there is , however , an important caveat : observations of a point source are not necessarily those of a bare nucleus , as a faint and/or unresolved coma may be present . supporting observations may be needed to rule out coma contamination in any particular data set . we target a water - ice band with a band depth of 10% or more at a confidence level of 5-@xmath97 . averaging over a 0.3 - -wide band ( 2.9 to 3.2 ) , and taking a spectral resolving power of 100 , we set our signal - to - noise ratio goal to 20 per resolution element . _ jwst _ can observe comet yeung on 2019 june 30 ( @xmath98 au , @xmath99 au , phase angle @xmath100 ) , when the comet s 3- flux density would be approximately 5.7 @xmath101jy ( the thermal contribution is @xmath1025% at @xmath1034 ) . we meet our snr goal with nirspec s low - resolution , fixed - slit mode in 300 seconds of integration time . since the integration time is short compared to all known comet rotation periods ( @xmath1045 hr ) , longitudinally resolved observations would be possible , providing important motivation to measure this comet s rotation period by 2019 . the centaur 95p / chiron ( 2060 ) is known to have water ice and activity @xcite , and a spectrum taken in jul 2019 ( @xmath105 au , @xmath106 au , @xmath107jy for @xmath108 and 80 km radius , @xcite ) could easily observe a 3- feature as weak as a few percent . in 1.0 hr , an @xmath94 spectrum from 2.9 to 5.0 can be obtained with a 3 snr of 60 per resolution element . this higher resolution spectrum , and complementary spectra at shorter wavelengths , could be used to investigate the structure and composition of the ice ( crystallinity , dirt fraction , trapped volatiles ) , as well as to look for signatures of organics in the 3- to 4- region . in an hour of integration time , we can achieve a snr of @xmath3820 on centaurs as small as @xmath3812 km radius at the distance of chiron , sufficient for detecting a 3- absorption feature at the 10% level . observations of a representative sample of centaur surfaces within _ jwst _ s lifetime will be possible . these observations of centaurs , as transitional objects between the trans - neptunian and jupiter - family comet regions , will yield new insight into the evolutionary processes currently at work in our solar system . for our comet coma case , we move beyond ice detection and instead focus on physical characterization . a signal - to - noise ratio of @xmath109 is necessary to ensure that our observations are sensitive to water - ice fractions of a few percent by area , to constrain the size of the scattering units , and to test the presence of the potentially weak crystalline water - ice feature at 1.65 . as an example case we consider the observations of comet c/2012 s1 ( ison ) by @xcite with the _ hubble space telescope _ when the comet was at 4.15 au from the sun and 4.24 au from earth . the authors report an @xmath21 of 1300 cm at 0.6 within a distance of 5000 km from the nucleus , which corresponds to a flux density of 50 and @xmath110 @xmath101jy at 0.6 and 3 , respectively , within an angular diameter of 04 . we conclude that in 900 s we can obtain a spectrum of comet ison with a snr of 100 using the current nirspec prototype exposure time calculator . @xcite report reddening of the dust coma between 5000 km and 10000 km from the nucleus , compatible with the presence of icy grains close to the nucleus . it would be interesting to test this hypothesis by measuring the coma ice distribution with nirspec . the flux density at 10000 km from the nucleus at 3 is estimated to be 0.4 @xmath101jy . an exposure time of 3 h yields a snr of 30 , high enough quality to test for the presence and size of water ice grains , and to study its evolution from the inner - coma out to 10000 km . this exercise demonstrates that jwst will be able to measure ice properties throughout the coma of moderately bright comets at @xmath111 au . s capabilities will transform our understanding of faint and/or weakly active comets by enabling observations that were heretofore at the limit of experimental techniques . such comets comprise several categories : 1 . main - belt comets ( mbcs ) . activity recurring from orbit to orbit has been identified in a handful of mbcs ( e.g. , * ? ? ? * ; * ? ? ? * ; * ? ? ? * ) , but the mechanism for this activity is unknown . to date , there have been no successful searches for signatures of volatiles on mbcs . nearly inactive and dormant comets . a large number of suspected dead comets ( inactive objects in comet - like orbits ) have been identified in the asteroid population ( e.g. , * ? ? ? * ; * ? ? ? * ) , and occasionally asteroids show cometary activity . for example , _ spitzer_/irac observations of `` asteroid '' ( 3552 ) don quixote exhibited signs of co@xmath0 outgassing @xcite . in addition , there are known comets with extremely weak or only sporadic activity , e.g. , 209p / linear 41 @xcite or 107p / wilson - harrington @xcite . high - sensitivity observations of volatiles can help confirm or better characterize their cometary natures . 3 . very small comets . the population of small near - sun comets are generally thought to be @xmath112 m in radius @xcite , too small or weakly active to be observed beyond the solar coronagraphs on soho or stereo . little is known about these small bodies , including what drives their activity : normal cometary activity via sublimation of ices or the sublimation of dust and refractory grains due to the extreme temperatures experienced near perihelion ( equilibrium temperatures exceeding 1000 k ; * ? ? ? distant comets . many comets show continuous activity outside the so - called water - ice line . their activities may be driven by the more volatile ices co@xmath0 or co @xcite , or by the exothermic annealing of water ice from amorphous to crystalline states @xcite . owing to their large distances , studies of activity are generally limited to observations of dust comae and tails . a common theme in the above cases is faint activity . nircam imaging would be capable of searching for signs of very weak activity ( coma and/or tail , both dust and gas ) and of efficiently characterizing nucleus properties such as size , elongation , albedo , and beaming parameter . nirspec might be able to spectroscopically detect the driving volatiles at production rates lower than previously achievable . the comet - asteroid transition is currently poorly understood , and such results , whether of severely processed sun - grazing / sun - skirting comets , dormant comets , distant comets , or of activated asteroids , would yield insight into the ongoing evolutionary processes of our solar system . to demonstrate _ jwst _ s capabilities for characterizing activity in faint comets , we consider observations of main belt comet p/2010 r2 ( la sagra ) in late aug 2021 ( @xmath113 au , @xmath114 au ) . this observing window is post - perihelion , during a period when comet la sagra has been observed to be active with an @xmath7 near 50 cm @xcite . estimates of the dust production rate by @xcite and @xcite span from 0.1 to 4 kg s@xmath2 . assuming a dust to gas mass production ratio of 1:1 , and assuming water sublimation is driving the activity , this corresponds to a water production rate with an order of magnitude of @xmath115 molecules s@xmath2 . for this exercise , we use nirspec to detect the @xmath116 water band at 2.7 , one of the strongest water bands in comets @xcite and one that can not be observed from the ground @xcite . in a small aperture of radius 02 , the estimated continuum flux density at 2.7 is @xmath117 w m@xmath15 @xmath2 , and the water band flux is @xmath118 w m@xmath15 . with the nirspec prototype exposure time calculator and the @xmath93 prism , we model the water band as a single 0.2 - -wide line . in a 1 hr exposure , the peak of the line is detected above the continuum with a signal - to - noise ratio of 4 . thus , we can expect _ jwst _ to provide one of the best tests for the drivers of main - belt comet activity . _ jwst _ is expected to have the ability to track moving targets with proper motions @xmath119108 hr@xmath2 ( 30 mas s@xmath2 ) @xcite . to estimate the impact this limit may have on cometary science , we searched the nasa jet propulsion laboratory solar system dynamics group database for all comets that reached perihelion during the five - year period 2010 jan 1 to 2015 jan 1 . this list includes 393 comets : 221 short - period comets ( with numbered or provisional `` p/ '' designations ) , and 172 oort cloud and long - period comets ( with provisional `` c/ '' designations ) . for each target , we used the iau minor planet center s ephemeris service to generate ephemerides with 3-day intervals , using the center of the earth as the observer . the differences in ephemerides between the earth and _ jwst _ at the l2 lagrange point ( 1.5@xmath8610@xmath120 km or 0.01 au away ) are not significant for this exercise . the proper motions of all targets within _ jwst _ s solar elongation constraints ( 85135 ) are shown in fig . [ fig : proper ] . not all of the observing windows make sense in a real world situation , e.g. , the plot includes epochs before some comets are discovered , and all epochs are shown independent of the comet s apparent brightness . however , we neglect these and other considerations and focus on the orbital characteristics of the known comet population . in fig . [ fig : proper - hist ] , we show histograms of our comet ephemerides binned by heliocentric distance . oort cloud and long - period comets have higher proper motions in the inner solar system than short - period comets , due to their higher eccentricities and the greater potential for high inclinations , including retrograde orbits . these faster moving comets are , therefore , more likely to exceed a proper motion of 108 hr@xmath2 . less than 50% of the opportunities to observe oort cloud or long - period comets inside of 1.7 au are possible , given the current anticipated tracking limit . inside of 1.3 au from the sun , there is a @xmath12125% probability that any particular comet could be observed . this limitation can significantly affect science programs requiring the maximum possible spatial resolution in studies of nearby comets . any observations dedicated to the near - nucleus environment will have a limited set of objects to consider , if any . if we artificially increase the non - sidereal tracking rate by a factor of 2 to 216 hr@xmath2 , the situation is substantially improved : the 50% limit for oort cloud / long - period comets is reduced to @xmath122 au and most comets can be observed down to 1.1 au from the sun . in table [ tab : comets ] we present a select list of targets observable by _ jwst _ , based on : ( 1 ) comets with perihelion dates during the first 5 years of the observatory s mission , ( 2 ) centaurs with known or suspected cometary activity , ( 3 ) prior and potential spacecraft targets , and ( 4 ) known main - belt comets and activated asteroids . our choice of jupiter - family comets is biased towards previous or potential spacecraft targets , and targets with favorable observing circumstances ( small comet - observer distance , or otherwise easily observable at perihelion ) . this list of comets , as well as undiscovered comets ( especially oort cloud and long - period comets ) , comets near aphelion , and unanticipated outbursts and fragmentation events will provide many targets of opportunity for _ jwst _ observers . we considered simple tools for estimating comet dust and gas brightness based on known comet properties . we then explored four science themes that _ jwst _ will be especially suited to address : ( 1 ) for a moderately bright comet , we can expect studies of the main drivers of cometary activity , h@xmath0o , co@xmath0 , and co , out to a heliocentric distance of at least 4 au ; co@xmath0 can be observed even farther , out to at least 7 au . ( 2 ) we assessed the observatory s potential to detect ice and organics in the comet population . a survey of surface ice in centaurs appears feasible for objects as small as 12 km out to 19 au from the sun . spectroscopy of jupiter - family comets , targeted near aphelion during periods of inactivity , can be used to determine ice deposition rates ( if occurring ) , and further understand the uniqueness of the recent discovery of organics on the surface of comet 67p / churyumov - gerasimenko . the properties of water ice grains can be measured , and spatially resolved throughout the coma of a moderately bright comet at 4 au . such a study would provide observational constraints on the lifetime of coma water ice , which can in turn be compared to their physical properties derived from light scattering . ( 3 ) through the fundamental emission band of water at 2.7 , _ jwst _ will be able to provide direct tests of the cometary nature of main - belt comets . ( 4 ) spatially resolved dust studies in the mid - ir can be used to examine the physical properties of the grains , and to test the heterogeneity of nuclei through their coma dust properties . finally , there will be many opportunities to observe comets during the first five - years of operations , but the proper motion limit significantly affects target availability within 1.7 au of the sun . improved tracking performance is highly desired to enhance the ability of the community to further studies of the near - nucleus environment of comets at the best - spatial resolutions possible . the authors thank an anonymous referee for their insightful critique that improved the manuscript . mspk acknowledges support for this work from nasa ( usa ) grant nnx13ah67 g , and cew acknowledges partial support from grant nnx13aj11 g . this work is supported at the aerospace corporation by the independent research and development program . 1p / halley & 1205 & 1.25 & 3450 & 10.1 & 1.13 & 2.9 & 1985 - 08 - 25.6 , t86 , t88 + 1p / halley & 5710 & 1.25 & 15400 & 10.1 & 1.13 & 2.7 & 1985 - 12 - 13.3 , t86 , t88 + 73p - c / s - w 3 & 243 & 0.7 & 1000 & 13 & 1.07 & @xmath123 & s06 , s11 + 73p - c / s - w 3 & 1400 & 1.0 & 4150 & 4.0 & 1.12 & 3.0 & s11 and this work + 103p / hartley 2 & 233 & 0.445 & 970 & 13 & 1.08 & 4.2 & s10 , m11 + c/1986 p1 ( wilson ) & 4230 & 1.25 & 9660 & 10.1 & 1.11 & 2.3 & 1987 - 06 - 1.2 , h89 + c/2012 k1 ( pan starrs ) & 5731 & 0.64 & 14900 & 19.7 & 1.02 & 2.6 & w15 + 9.90 & 9.20 & 133 & 2668 & 20 & 19.18 & -26.64 & & 24.29 & -20.74 & & 25.27 & -20.86 & + 8.04 & 8.04 & 87 & 2330 & 31 & 21.32 & -24.31 & & 24.90 & -19.93 & 12000 & 25.45 & -20.48 & + 7.05 & 6.55 & 116 & 1901 & 40 & 22.54 & -22.91 & & 25.16 & -19.50 & 2110 & 25.57 & -20.19 & 59300 + 5.00 & 4.90 & 90 & 1421 & 80 & 25.26 & -19.84 & 28200 & 25.69 & -18.62 & 127 & 25.87 & -19.54 & 4100 + 4.01 & 3.63 & 106 & 1052 & 124 & 26.23 & -18.60 & 306 & 25.95 & -18.08 & 32.7 & 26.06 & -19.08 & 841 + 1.92 & 1.68 & 87 & 488 & 540 & 27.62 & -16.39 & 0.731 & 26.69 & -16.53 & 0.942 & 26.70 & -17.62 & 52.4 + 1.47 & 1.08 & 90 & 313 & 925 & 27.94 & -15.71 & 0.148 & 26.93 & -15.92 & 0.288 & 26.94 & -17.02 & 36.8 + 2.39 & 2.16 & 91 & 627 & 350 & 27.33 & -16.94 & 2.79 & 26.48 & -16.99 & 2.56 & 26.51 & -18.06 & 81.2 + 3.03 & 2.65 & 102 & 769 & 218 & 26.93 & -17.59 & 14.9 & 26.25 & -17.46 & 7.61 & 26.31 & -18.51 & 186 + 5.19 & 5.10 & 89 & 1480 & 74 & 25.00 & -20.14 & 98400 & 25.64 & -18.71 & 163 & 25.84 & -19.61 & 5430 + 6.05 & 5.41 & 126 & 1569 & 55 & 23.86 & -21.41 & & 25.42 & -19.05 & 449 & 25.70 & -19.88 & 15400 + 8.09 & 8.03 & 90 & 2329 & 31 & 21.26 & -24.37 & & 24.88 & -19.95 & 13200 & 25.45 & -20.49 & + 9.04 & 8.50 & 120 & 2464 & 24 & 20.12 & -25.61 & & 24.56 & -20.38 & 84600 & 25.35 & -20.69 & + 2p / encke & 2020 - 06 - 26 & potential spacecraft target + & 2023 - 10 - 23 & + 6p / darrest & 2021 - 09 - 18 & potential spacecraft target + 21p / giacobini - zinner & 2018 - 09 - 10 & spacecraft target + 29p / schwassmann - wachmann 1 & 2019 - 03 - 07 & centaur , frequent strong outbursts + 38p / stephan - oterma & 2018 - 11 - 10 & halley - type + 46p / wirtanen & 2018 - 12 - 12 & potential spacecraft target ; historic close approach to earth ( 0.078 au ) + 64p / swift - gehrels & 2018 - 11 - 03 & excellent apparition + 67p / churyumov - gerasimenko & 2021 - 11 - 02 & spacecraft target + 95p / chiron ( 2060 ) & 2046 - 08 - 03 & centaur with jets / arcs / cometary activity @xcite + 99p / kowal 1 & 2022 - 04 - 12 & large @xmath129 ( 4.7 au ) + 103p / hartley 2 & 2023 - 10 - 12 & spacecraft target + 104p / kowal 2 & 2022 - 01 - 08 & excellent apparition + 107p / wilson - harrington & 2022 - 08 - 25 & active asteroid with small @xmath129 ( 0.97 au ) + 117p / helin - roman - alu 1 & 2022 - 07 - 08 & large @xmath129 ( 3.0 au ) + 126p / iras & 2023 - 07 - 05 & halley - type orbit + 176p / linear 52 & 2022 - 11 - 21 & mbc active by sublimation + 133p / elst - pizarro & 2018 - 09 - 21 & mbc active by sublimation + 289p / blanpain & 2019 - 12 - 20 & nearly extinct comet @xcite ; historic close approach to earth ( 0.086 au ) + 311p / panstarrs 23 & 2020 - 10 - 07 & mbc active by rotational breakup @xcite + p/2010 r2 ( la sagra ) & 2021 - 05 - 09 & mbc active by sublimation @xcite + p/2013 r3 ( catalina - panstarrs 1 ) & 2018 - 12 - 06 & mbc active by rotational breakup @xcite + c/2010 u3 ( boattini ) & 2019 - 02 - 26 & oort cloud comet with very large @xmath129 ( 8.5 au ) + c/2014 f3 ( sheppard - trujillo ) & 2021 - 05 - 23 & unusual orbit : low inclination and long ( 60-year ) period + ( 596 ) scheila & 2022 - 05 - 26 & mbc active by impact @xcite + ( 3200 ) phaethon & 2019 - 07 - 03 & active asteroid with small @xmath129 ( 0.14 au ) @xcite + ( 10199 ) chariklo & 2066 - 06 - 26 & centaur with rings @xcite + o , co@xmath0 , and co , given our example model described by section [ sec : gas - coma ] and table [ tab : gas - flux ] . thick lines mark the fictitious _ jwst _ observing windows ( solar elongation between 85 and 135 ) . on the left are pre - perihelion epochs , on the right are post - perihelion epochs.[fig : gas ] ] ) versus heliocentric distance ( @xmath13 ) for all comets in our non - sidereal rate study . only epochs within _ jwst _ s solar elongation constraints ( 85135 ) are shown . a dashed horizontal line marks the observatory s current non - sidereal rate limit of 108 hr@xmath2 . the isolated comet with high proper motion at 5 au is p/2011 p1 ( mcnaught ) . this object was potentially in a large orbit before passing jupiter from a distance of 0.0008 au in december 2010 , and is currently in a jupiter - family comet orbit . [ fig : proper ] ] , versus heliocentric distance , @xmath13 , for our non - sidereal rate study . the solid line includes all comets , the dotted line only considers short - period comets , and the dashed line is limited to long - period and oort cloud comets . bottom : the fraction of epochs with non - sidereal rates within _ jwst _ s observing limits , @xmath9 , versus heliocentric distance . in general , short period comets are more accessible to _ jwst _ , and the observatory s tracking limit significantly impacts observations within @xmath130 au . [ fig : proper - hist],scaledwidth=50.0% ]

though well - studied for over fifty years , the spectral theory of hyperbolic surfaces still presents basic unresolved questions @xcite . for example , does there exist a noncompact , finite area hyperbolic surface whose laplacian has no nonconstant square - integrable eigenfunctions ? this question has been the subject of many investigations including @xcite , @xcite , @xcite , @xcite , @xcite , @xcite , and @xcite . the triangle @xmath0 defined by @xmath1 and @xmath2 . ] as a model problem , phillips and sarnak @xcite suggested studying the neumann eigenvalue problem on the domain @xmath3 pictured in figure [ trisect ] . in this paper , we prove the following : [ thm : main ] for all but at most countably many @xmath40,1[$ ] , the neumann laplacian on the geodesic triangle @xmath0 in the hyperbolic plane has no nonconstant ( square - integrable ) eigenfunction . the group @xmath5 of hyperbolic isometries generated by reflections in the geodesic arcs that bound @xmath0 is discrete if and only if @xmath6 for @xmath7 an integer . for example , if @xmath8 , then @xmath5 contains an index two subgroup that is naturally isomorphic to @xmath9 . it follows from the seminal work of a. selberg @xcite that if @xmath10 or @xmath11 , then the neumann laplacian has infinitely many nonconstant eigenfunctions . in particular , for some special @xmath12 , there do exist square - integrable solutions to the neumann problem . is an integer not equal to @xmath13 , or @xmath11 , phillips and sarnak asked whether the domain @xmath14 has no nonconstant neumann eigenfunctions @xcite . we should point out that since theorem [ thm : main ] allows for countably many exceptional @xmath12 , it does not directly answer their question . ] in @xcite , theorem [ thm : main ] was verified under an additional and as of yet unjustified assumption concerning the spectral multiplicities of the neumann laplacian acting on @xmath15 . the proof consisted of studying the singular perturbation problem associated with letting @xmath12 tend to @xmath16 . similar singular perturbations were studied in the context of degenerating hyperbolic surfaces @xcite and unitary characters over a fixed hyperbolic surface @xcite . in @xcite , @xcite , @xcite , and all prior work on this problem , it was necessary to make assumptions about the multiplicities of the spectrum of the limiting surface . the angles of a geodesic triangle in the hyperbolic plane determine the isometry class of the triangle . the angles of @xmath0 are @xmath17 . it is not difficult to extend theorem [ thm : main ] to triangles with angles @xmath18 ( see [ section : alltriangles ] ) . [ thm : gennocf ] the set of @xmath19 for which the hyperbolic triangle with angles @xmath20 admits a nonconstant neumann laplace eigenfunction has lebesgue measure zero and is contained in a countable union of nowhere dense sets . in other words , the generic hyperbolic triangle with one cusp has no non - constant neumann eigenvalue where ` generic ' can be taken in both a topological and a measurable sense . theorem [ thm : main ] gives the existence of a triple of angles @xmath18 for which there are no nonconstant neumann eigenfunctions . theorem [ thm : gennocf ] then results from applying a general and well - understood principle concerning analytic perturbations ( see , for example , @xcite ) . on the other hand , the proof of theorem [ thm : main ] is much more involved . in particular , the proof will rely upon a refined analysis of ` crossings ' of eigenvalue branches . to prove theorem 1.1 , we further develop the method of ` asymptotic separa- tion of variables ' that we introduced in [ hlrjdg11 ] to study generic simplicity of eigenvalues . this method facilitates the study of real - analytic eigenvalue branches in situations where a geometric domain degenerates onto a lower dimensional domain . there is a vast literature for instance , @xcite , @xcite , @xcite concerning perturbations involving degeneration onto lower dimensional domains , but most of these studies do not address analytic eigenvalue branches . in contrast , our results depend crucially on a study of real - analytic eigenbranches and their crossings . we now describe the content of each section . in [ section : prelim ] , we establish notation and recall basic features of the spectral theory of the laplacian acting on functions on a domain in the hyperbolic plane having one cusp . we describe the fourier decomposition associated to the cusp . the zeroth fourier mode is responsible for an essential spectrum of @xmath21 . following @xcite and @xcite , we will replace the dirichlet quadratic form @xmath22 with a modification @xmath23 obtained by ` truncating ' the zeroth fourier coefficient above @xmath24 . an eigenfunction @xmath25 of @xmath23 corresponds to an eigenfunction of @xmath26 if and only if the zeroth fourier coefficient of @xmath25 vanishes identically . we will call such an eigenfunction a _ cusp form_. , these are ` cusp forms ' in the sense of the theory of automorphic forms , but otherwise there is no discrete group , and hence they are not cusp forms in the traditional sense . in this paper we will always be considering ` even ' cusp forms , that is , eigenfunctions satisfying neumann conditions . ] the operator associated to @xmath23 has compact resolvent and hence discrete spectrum . this makes @xmath23 a much better candidate for the application of methods from spectral perturbation theory . in [ sec : generic ] , we recall and make precise some ideas familiar in the perturbational study of cusp form existence . we consider a real - analytic family , @xmath27 , of quadratic forms that have the same domain as @xmath23 . we say that a real - analytic family @xmath28 of eigenfunctions of @xmath29 is a _ cusp form eigenbranch _ if and only @xmath30 is a cusp form for each @xmath12 . we demonstrate a dichotomy : either the family @xmath27 has a real - analytic cusp form eigenbranch or the set of @xmath12 such that @xmath29 has a cusp form is countable . in [ section : alltriangles ] , we consider arbitrary real - analytic paths in the space of hyperbolic triangles with one cusp . we apply the results of [ sec : generic ] to deduce theorem [ thm : gennocf ] under the assumtption that there exists a triangle with no nonconstant neumann eigenfunction . the remainder of the paper is devoted to proving theorem [ thm : main ] which will give the existence of such a triangle . in [ section : expansion ] we specialize to the family @xmath0 . after renormalizing by a factor of @xmath31 , we find that for each @xmath25 , the function @xmath32 has a taylor expansion at @xmath33 . we compute the leading order terms of this expansion . in [ section : approx ] we show that the method of the asymptotic separation of variables introduced in @xcite may be used to analyse the family of quadratic forms @xmath34 . in particular , we define a reference quadratic form @xmath35 to which separation of variables applies and that is asymptotic to @xmath34 at ` first order ' . by separation of variables we mean that each eigenfunction of @xmath35 is of the form @xmath36 with @xmath37 and @xmath38 a solution to a renormalized form of the equation for a modified bessel function with imaginary parameter . we show that one may deduce a non - concentration estimate ( proposition [ prop : noncon ] ) . in [ section : branches ] we use the non - concentration estimate to derive information concerning the real - analytic eigenbranches @xmath39 of @xmath34 . first , we show that there exists an integer @xmath40 so that @xmath41 limits to @xmath42 as @xmath12 tends to zero . second , we find that if the spectral projection of @xmath30 onto the space @xmath43 spanned by functions of the form the @xmath44 is relatively small , then the derivative @xmath45 is of order @xmath46 . finally , we show that if @xmath47 is a cusp form eigenbranch , then @xmath41 can not limit to zero . in [ sec : proof ] we prove theorem [ thm : main ] . by the dichotomy of [ sec : generic ] , it suffices to show that real - analytic cusp form eigenbranches do not exist . we suppose to the contrary that the real - analytic family @xmath34 has a cusp form eigenbranch @xmath47 . by the results of [ section : branches ] , we have that @xmath41 limits to @xmath42 where @xmath48 . by improving the analysis of @xcite , we show that there exists an eigenbranch , @xmath49 , of @xmath35 that ` tracks ' @xmath41 at order @xmath12 in the sense that @xmath50 we will obtain a contradiction to this estimate by estimating @xmath51 from below . indeed , we show that when the norm of the projection @xmath52 of @xmath30 onto @xmath43 is relatively large with respect to @xmath53 , the function @xmath54 is of order @xmath55 , whereas when @xmath56 is relatively small , the function @xmath54 is of order @xmath57 . by controling the sizes of the sets where @xmath56 is respectively small and large relative to @xmath53 and by integrating , we will contradict ( [ trackingestimate ] ) . the key observation is the following : since @xmath41 limits to @xmath58 , it has to ` cross ' each of the eigenbranches of @xmath35 that limit to zero . we show that near such a crossing , the branch @xmath30 must ` interact ' with the functions in @xmath59 to such an extent that the projection onto @xmath43 can not be too large . the effect of each interaction is made precise by careful estimates of the off - diagonal terms in the quadratic form @xmath60 ( appendix a ) . by summing the effects of these interactions , we eventually prove that there exists @xmath61 so that @xmath62 thus contradicting ( [ trackingestimate ] ) . in this section , we describe some basic spectral theory of the neumann laplace operator acting on the square - integrable functions on domains in the hyperbolic plane with a cusp . we define the laplacian and associated dirichlet quadratic form , describe the fourier decomposition of eigenfunctions along horocycles , construct a modification of the dirichlet form whose eigenfunctions include the eigenfunctions ( cusp forms ) of the standard laplacian but whose spectrum is discrete . the half - plane @xmath63 equipped with the riemannian metric @xmath64 is the poincar - lobachevsky model for the @xmath65-dimensional hyperbolic space @xmath66 . the measure associated to the riemannian metric @xmath67 is given by integrating @xmath68 in the present context , a _ cusp of width @xmath69 and height @xmath70 _ is the subset @xmath71 \times [ y_0 , \infty[$ ] of the upper half plane . a domain @xmath72 is said to _ have one cusp _ if @xmath73 is the union of a cusp and a compact set . we will assume that the boundary of @xmath73 is the the union of finitely many geodesic arcs and that the interior of @xmath73 is connected . let @xmath74 denote the set of functions @xmath75 such that @xmath25 is the restriction to @xmath73 of a compactly supported smooth function defined on some neighbourhood of @xmath73 . the @xmath76-inner product of two functions @xmath25 and @xmath77 in @xmath74 is defined by @xmath78 abusing notation slightly , we will often write @xmath79 in place of @xmath80 . let @xmath81 denote the completion of @xmath74 with respect to the norm @xmath82 . to define the neumann laplacian we consider the bilinear form defined on @xmath74 by @xmath83 where @xmath84 satisfies @xmath85 for all vector fields @xmath86 and smooth functions @xmath87 . let @xmath88 denote the value of the quadratic form @xmath89 . one computes that @xmath90 let @xmath91 denote the completion of @xmath74 with respect to the norm @xmath92 . we will consider @xmath26 as a nonnegative symmetric bilinear form on @xmath93 with domain @xmath91 . as such it is densely defined and closed , and hence there exists a unique , densely defined , self - adjoint operator @xmath94 on @xmath93 such that for each @xmath95 and @xmath25 in the domain of @xmath94 we have @xmath96 the operator @xmath94 is called the _ neumann laplacian_. it can be shown that @xmath97 if and only if @xmath98 and @xmath99 where in the definition of @xmath26 the partial derivatives are to be taken in the distributional sense . it is well - known that @xmath94 has essential spectrum equal to @xmath100 . ( for example , see @xcite or @xcite ) . apart from this essential spectrum , @xmath94 may also have eigenvalues either smaller than @xmath101 or embedded in the continuous spectrum . from ( [ laplacian ] ) we see that @xmath25 is an eigenfunction of @xmath94 with eigenvalue @xmath102 if and only if @xmath97 and for each @xmath95 @xmath103 for each integer @xmath104 , define @xmath105 and define @xmath106 . the collection @xmath107 is an orthonormal basis for @xmath108)$ ] . hence , the functions @xmath109 provide an orthonormal basis of @xmath110)$ ] . for positive @xmath111 , let @xmath112 \times [ y_0 , \infty[$ ] be a cusp of width @xmath69 and height @xmath70 . for each @xmath25 in @xmath113 and almost every @xmath114 , the function @xmath115 belongs to @xmath110)$ ] . thus we can write @xmath116 where @xmath117 belongs to @xmath118 . we refer to @xmath119 as the @xmath120 fourier coefficient of @xmath25 . more generally , if @xmath73 is a domain with a cusp @xmath121 , then we define the @xmath120 fourier coefficient of a function @xmath122 to be the @xmath120 fourier coefficient restriction of @xmath77 to @xmath121 . parseval s theorem gives @xmath123 where @xmath124 denotes the characteristic function of a set @xmath86 . [ lem : ode ] if @xmath125 is a neumann eigenfunction of @xmath26 with eigenvalue @xmath102 , then for each @xmath126 and each @xmath127 , the fourier coefficient @xmath119 satisfies @xmath128 if @xmath77 is a smooth function on @xmath73 , then since @xmath25 is a neumann eigenfunction of @xmath26 , integration by parts gives @xmath129 by letting @xmath130 where @xmath131 is a smooth function with compact support in @xmath132y_0,\infty[$ ] , we find that @xmath133 it follows that @xmath119 satisfies ( [ eq : oden ] ) in the distributional sense in @xmath134 by elliptic regularity , @xmath119 is actually smooth and satisfies ( [ eq : oden ] ) in the strong sense . in particular , @xmath135 for some constants @xmath136 and @xmath137 with @xmath138 . if @xmath139 , then the real part of @xmath140 equals @xmath141 , and hence @xmath142 does not belong to @xmath118 unless both @xmath136 and @xmath137 equal zero . therefore , we have the following . [ coro : cuspform ] if @xmath25 is a neumann eigenfunction with eigenvalue @xmath143 , then the zeroth fourier coefficient @xmath142 vanishes identically on @xmath144 . in the classical spectral theory of a quotient of @xmath145 by a lattice in @xmath146 , a laplace eigenfunction @xmath25 with vanishing zeroth fourier coefficient in each cusp is called a _ ( weight zero ) maass cusp form_. even though most of the domains that we will consider are not fundamental domains for discrete groups of isometries , we will adapt this terminology . if @xmath25 is an eigenfunction for the neumann laplacian on a domain with a cusp , and @xmath147 , then we will call @xmath25 a _ cusp form_. traditionally , the neumann eigenfunctions for @xmath14 are called _ even cusp forms _ whereas the solutions to the dirichlet eigenvalue problem are called _ odd cusp forms_. we will not consider odd cusp forms in this paper . we wish to apply analytic perturbation theory to study the behavior of eigenfunctions of @xmath26 on @xmath0 as we vary @xmath12 . because the eigenvalues of @xmath26 might lie inside the essential spectrum , standard perturbation theory does not apply directly . following @xcite and @xcite , we will use a modification of @xmath26 first constructed by p. lax and r. phillips @xcite @xcite . in this section , we recall the construction , show that the eigenvalues of the modification are isolated , and relate the eigenfunctions of the modification to those of @xmath26 . for @xmath148 , let @xmath149 denote the set of @xmath150 such that for each @xmath151 we have @xmath152 . let @xmath153 denote the hilbert space completion of @xmath149 with respect to @xmath154 . let @xmath155 denote the restriction of @xmath156 to @xmath157 . let @xmath158 denote the hilbert space completion of @xmath149 with respect to the norm @xmath159 . the restriction , @xmath23 , of @xmath26 to @xmath158 is a closed , densely defined quadratic form on @xmath157 . a simple argument shows that @xmath160 and @xmath161 in the sequel it will be more convenient to replace @xmath162 by the following other set . [ def : wbet ] define @xmath163 to be the set of functions @xmath25 in @xmath164 such that * @xmath25 extends to a continuous function on the closure @xmath165 of @xmath166 * @xmath25 is smooth on @xmath167 observe that since @xmath168 , the closure of @xmath163 with respect to the norm @xmath159 is @xmath169 the latter assertion says that @xmath163 is a core of the quadratic form @xmath170 let @xmath171 denote the unique operator such that @xmath172 and that satisfies @xmath173 for each @xmath174 . [ lem : relation ] for each @xmath148 , the resolvent of @xmath171 is compact . hence , the spectrum of @xmath23 with respect to @xmath155 is discrete and each eigenspace is finite dimensional . using the fourier decomposition , one shows that for each @xmath175 , the set of @xmath176 such that @xmath177 and @xmath178 is compact in @xmath157 ( see lemma 8.7 @xcite ) . it follows that @xmath171 has compact resolvent . hence , by standard spectral theory , the spectrum is discrete and the eigenspaces are finite dimensional . [ def : cf ] we will say that an eigenfunction @xmath25 of @xmath23 with respect to @xmath155 is a _ cusp form _ if and only for each @xmath127 we have @xmath152 . [ lem : betacf ] the following assertions are equivalent : 1 . @xmath25 is a cusp form of @xmath26 with respect to @xmath156 . there exists @xmath148 such that @xmath25 is a cusp form of @xmath23 with respect to @xmath155 . 3 . for each @xmath148 , the function @xmath25 is a cusp form for @xmath23 with respect to @xmath155 . @xmath179 : if @xmath25 is an eigenfunction of @xmath26 with eigenvalue @xmath102 , then by lemma [ lem : ode ] the zeroth fourier coefficient @xmath142 satisfies the differential equation @xmath180 . since @xmath181 vanishes for @xmath182 , it must vanish for @xmath183 . @xmath184 : fix a smooth function @xmath185 such that @xmath186 for @xmath187 and @xmath188 for @xmath189 if @xmath152 for each @xmath182 , then @xmath190 and @xmath191 for each @xmath95 . thus , if @xmath192 , then @xmath193 . @xmath194 : follows from the equivalence of ( 1 ) and ( 2 ) . not every eigenfunction @xmath25 of @xmath23 is an eigenfunction of @xmath26 . for example , if @xmath195 and @xmath196 is an eisenstein series whose zeroth fourier coefficient vanishes at @xmath24 , then @xmath197 is an eigenfunction of @xmath23 . this function is not smooth across @xmath24 , but elliptic regularity implies that each eigenfunction of @xmath26 is smooth . let @xmath198\times [ y_0,\infty)$ ] and let @xmath199 in this section , we consider a fixed domain @xmath73 that contains the cusp @xmath200 and a real - analytic family @xmath27 of quadratic forms defined on @xmath201 that represents the cusp of width @xmath202 for @xmath182 and some real - analytic function @xmath203 ( see definition [ def : represent ] below ) . we prove the following dichotomy : either there exists a real - analytic eigenfunction branch consisting of ` cusp forms ' or the set of @xmath12 such that @xmath29 has a ` cusp form ' eigenfunction is countable . this fact is fundamental to the proofs of both theorem [ thm : main ] and theorem [ thm : gennocf ] . let @xmath198\times [ y_0 , \infty[$ ] and let @xmath204 . for any @xmath205 , we can define a transformation @xmath206 between @xmath207 and @xmath208 by asking that , for any @xmath209 the function @xmath210 is defined by @xmath211 since , in the sequel @xmath70 will be fixed , we will drop the index @xmath70 and denote by @xmath212 the cusp of width @xmath213 it is straighforward that @xmath206 is an isometry between @xmath207 and @xmath214 moreover @xmath206 also preserves @xmath164 in the sense that @xmath215 if and only if @xmath216 we may thus define @xmath217 the quadratic form obtained by pulling - back @xmath170 on @xmath212 using @xmath218 this quadratic form is then closed on the domain @xmath219 , and for each @xmath220 @xmath221 [ def : represent ] let @xmath73 be a domain that has @xmath200 as a cusp and @xmath222 let @xmath223 be a quadratic form closed over the domain @xmath224 we will say that @xmath223 represents the cusp of width @xmath69 for @xmath225 if , for any @xmath226 that is supported in @xmath227 we have @xmath228 for such a quadratic form , we will say that an eigenfunction @xmath25 is a cuspform if @xmath181 vanishes on @xmath229 the aim of this section is to prove that being a cuspform is a real - analytic condition . to make this statement precise we have to consider a family @xmath29 of quadratic forms that satisfies the following assumptions . [ ass : qt ] let @xmath230 and @xmath231 . for each @xmath232t_- , t_+[$ ] , let @xmath202 be a positive real - analytic function on @xmath233 let @xmath29 denote a nonnegative , closed quadratic form with domain @xmath234 that represents the cusp of width @xmath202 for @xmath235 lastly , we assume that the family @xmath27 is real - analytic of type ( a ) in the sense of @xcite . that is , for each @xmath236 , the map @xmath32 is real - analytic . a straightforward application of analytic perturbation theory@xcite vii gives the following . [ thm : eigenbasis_exist ] let @xmath27 satisfy the assumptions above [ ass : qt ] . then there exist a collection of real - analytic paths @xmath237 and a collection of real - analytic functions @xmath238 so that for each @xmath12 , the set @xmath239 is an orthonormal basis for @xmath153 , and for each @xmath240 , the function @xmath241 is an eigenfunction of @xmath29 with eigenvalue @xmath242 since the embedding from @xmath243 into @xmath244 is compact , for any @xmath12 the spectrum of @xmath29 consists only in eigenvalues . the proof is then similar to the proof of theorem 3.9 in @xcite viii.3.5 . for @xmath245 , define @xmath246 [ lem : nocf_deriv_criterion ] an eigenfunction @xmath25 of @xmath29 is a cusp form if and only if @xmath247 . @xmath248 is the left - sided derivative of @xmath142 at @xmath249 . since @xmath25 is an eigenfunction , @xmath245 and @xmath142 is a solution to a second order ordinary differential equation on @xmath250 $ ] with @xmath251 . thus , @xmath142 vanishes identically on @xmath250 $ ] if and only if @xmath247 . for real - analytic eigenbranches we have the following . [ lem : lanalytic ] let @xmath252 be an analytic eigenbranch of @xmath29 then the mapping @xmath253 is analytic on @xmath132t_-,t_+[.$ ] the zeroth mode @xmath254 of @xmath30 is a solution to the ode @xmath255 on @xmath256 $ ] with dirichlet boundary condition at @xmath257 denote by @xmath258 the unique solution to this ordinary differential equation that satisfies @xmath259 since the coefficients of the ordinary differential equation depend analytically on the parameter @xmath260 , the mapping @xmath261 is analytic ( with values in @xmath262)$ ] for instance ) . moreover , for each compact set @xmath263t_-,t_+[$ ] we can find @xmath264 such that , for each @xmath265 , @xmath266 . since @xmath30 is a multiple of @xmath267 , we then have @xmath268 analyticity on @xmath269 then follows from the analyticity of @xmath270 and @xmath271 and the choice of @xmath272 . we will say that real - analytic eigenfunction branch @xmath30 of @xmath29 is _ a cusp form eigenbranch _ if and only if for each @xmath273 , the eigenfunction @xmath30 is a cusp form ( see definition [ def : represent ] ) . using the real - analyticity proved in lemma [ lem : nocf_deriv_criterion ] we obtain the following [ coro : cf_discrete ] if @xmath30 is a real - analytic eigenfunction branch that is not a cusp form eigenbranch , then the set of @xmath273 such that @xmath30 is a cusp form is discrete . we now proceed to prove that if @xmath29 has no real - analytic cusp form eigenbranch then , for a generic @xmath12 , the form @xmath29 has no cusp form . as it turns out , we will actually first prove that the spectrum of @xmath29 is generically simple . let @xmath274 denote the set of @xmath273 such that @xmath29 has an eigenspace of dimension at least two . [ prop : mult ] if @xmath29 does not have a real - analytic cusp form eigenbranch , then @xmath274 is countable . let @xmath275 and @xmath276 be as in theorem [ thm : eigenbasis_exist ] . for each @xmath277 , let @xmath278 . since each eigenspace of @xmath29 is spanned by a finite collection of @xmath279 , the union @xmath280 equals @xmath274 . the function @xmath281 is analytic , and hence @xmath278 is either countable or equals @xmath282 . thus to prove the claim , it suffices to show that it is not possible for @xmath283 to equal @xmath282 . suppose that there exists @xmath284 and @xmath40 so that @xmath241 and @xmath285 are real - analytic eigenbranches so that @xmath286 for each @xmath287 . to prove the proposition , it suffices to produce a linear combination @xmath30 of @xmath241 and @xmath285 so that for each @xmath273 , the function @xmath30 is a real - analytic cusp form eikgenbranch . by hypothesis , neither @xmath241 nor @xmath285 are cusp form eigenbranches . by corollary [ coro : cf_discrete ] , the set @xmath288 of @xmath12 such that either @xmath241 or @xmath285 is a cusp form is discrete . for each @xmath289 , define @xmath290 its suffices to show that @xmath291 extends to a real - analytic function on @xmath282 . indeed , since @xmath292 is linear , we have @xmath293 for each @xmath289 . by corollary [ lem : lanalytic ] , the real - analytic extension would satisfy @xmath294 . the order of vanishing of @xmath295 ( resp . @xmath296 ) is finite at each @xmath297 . if the order of vanishing of @xmath298 at @xmath299 is at least the order of vanishing of @xmath300 at @xmath301 , then the ratio @xmath302 has a real - analytic extension near @xmath301 . hence , factorizing @xmath300 we obtain that @xmath303 have real - analytic extensions near @xmath301 . if the order of vanishing of @xmath298 at @xmath301 is at most the order of vanishing of @xmath300 at @xmath301 , then a similar argument applies by factorizing @xmath298 everywhere . thus , @xmath30 extends to a real - analytic cusp form eigenbranch . let @xmath304 denote the set of @xmath273 such that @xmath29 has at least one cusp form eigenfunction . [ alt : alt ] if @xmath29 has no real - analytic cusp form eigenbranch , then @xmath304 is countable . if there exists a cusp form eigenbranch , then @xmath305 . therefore , we have the following dichotomy : either the set @xmath304 countable or the family @xmath27 has a real - analytic cusp form eigenbranch . let @xmath306 be as in theorem [ thm : eigenbasis_exist ] . by corollary [ coro : cf_discrete ] , the zero set @xmath307 is countable . if each eigenspace @xmath102 of @xmath29 is one - dimensional , then there exists a unique @xmath284 such that @xmath102 equals the span of @xmath241 . thus , if @xmath12 does not belong to @xmath274 or to any @xmath308 , then @xmath12 does not belong to @xmath304 . in other words , @xmath309 . by proposition [ prop : mult ] , the set @xmath274 is countable , and hence so is @xmath304 . in this section we use the results of the previous section to explain how theorem [ thm : gennocf ] can be deduced from the existence of a triangle with a cusp that has no nonconstant neumann eigenfunctions . this fact may probably be considered as folklore and follows the general philosophy of using analyticity to prove generic spectral results ( see @xcite ) . the main task here is thus construct a real - analytic family of quadratic forms that is associated with each real - analytic path in the moduli space of triangles . first , we discuss the parametrization of the set of triangles with one cusp . the statement of theorem [ thm : gennocf ] makes use of the fact that hyperbolic triangles with one cusp are parametrized by the two nonzero vertex angles . but in order to prove theorem [ thm : gennocf ] , it will be more convenient to use an alternate set of parameters . for each geodesic triangle @xmath310 in the hyperbolic upper half plane @xmath66 having ( exactly ) one cusp , there exists a unique @xmath3110,1[$ ] and @xmath312 so that @xmath310 is isometric to the domain @xmath313 see figure [ trigeneric ] . the triangle @xmath314 in the upper half plane . ] in this way , the set of hyperbolic triangles may be identified with the euclidean triangle @xmath315 we say that a subset of the set of triangles has measure zero if and only if the corresponding subset of @xmath316 has measure zero . similarly , a subset of the set of triangles is said to be a real - analytic curve if and only if the corresponding subset of @xmath316 is a real - analytic curve . these notions are equivalent to those used in the statement of theorem [ thm : gennocf ] because the relationship between the angles @xmath19 and the parameters @xmath317 is real - analytic . indeed , we have @xmath318 and @xmath319 . see figure [ trigeneric ] . to prove theorem [ thm : gennocf ] , we will apply perturbation theory . the following fact makes this approach feasible . [ prop : allcuspform ] each nonconstant neumann eigenfunction on @xmath314 is a cusp form and hence an eigenfunction of the modified quadratic form @xmath23 . the eigenvalue of a nonconstant neumann eigenfunction on @xmath310 is at least 1/4 @xcite . thus , the claim follows from corollary [ coro : cuspform ] and lemma [ lem : relation ] . let @xmath320 denote the set of @xmath321 such that there exists @xmath322 so that the modified quadratic from @xmath23 has a cusp form . to prove theorem [ thm : gennocf ] it will suffice to show that @xmath320 has measure zero and is a countable collection of nowhere dense sets . to show that @xmath320 is non - generic , we will use analytic perturbation theory and proposition [ alt : alt ] . in order to use analytic perturbation theory we will have to normalize the hilbert space and the domains of the quadratic forms . to accomplish this , we let @xmath323 \times [ 1 , \infty[$ ] and for each @xmath317 we define a @xmath324 diffeomorphism @xmath325 such that onto @xmath200 . the mapping that we define here is more complicated because it must preserve the notion of zeroth fourier coefficient for all @xmath326 above some point . in particular , the vertical displacement of vertical lines should not depend on @xmath327 for large @xmath326 . in @xcite , we considered dirichlet boundary conditions , and in that context there is no need to truncate the zeroth fourier coefficient . ] 1 . the restriction of @xmath328 is the identity for @xmath329 . 2 . for each path @xmath330 , the family @xmath331 is a real - analytic path . to construct @xmath328 , we use the fact that the map @xmath332 defines a fibration of @xmath314 over @xmath333 $ ] and a fibration of @xmath200 over @xmath334 $ ] . we define @xmath328 by sending the fiber over @xmath335 onto the fiber over @xmath336 . [ lem : defnb ] for each @xmath3370,{\underline{\alpha}}[$ ] , there exists a unique cubic polynomial @xmath338 so that * @xmath339 , * @xmath340 . * @xmath341 , * @xmath342 the coefficients of @xmath338 are real analytic for @xmath3430,{\underline{\alpha}}[$ ] . moreover if @xmath344 then , for all @xmath345,$ ] and @xmath346 $ ] , we have @xmath347 . since @xmath348 satisfying the two conditions on @xmath349 is equivalent to the existence of some @xmath136 such that @xmath350 denote by @xmath351 the cubic polynomial defined by @xmath352 by integration , there exists some @xmath353 such that @xmath354 evaluating at @xmath355 and using the condition on @xmath356 we find @xmath357 observe that @xmath358 if @xmath359 and , under this condition , we can solve the last equation on @xmath338 to find @xmath360 we obtain @xmath361\\ & = \frac{1-{\alpha}}{2{\underline{\alpha } } } \left [ { \alpha}^2+{\underline{\alpha}}^2 - 2{\underline{\alpha}}({\alpha}+1)\right],\\ & = \frac{(1-{\alpha})\left [ { \alpha}^2 -2 { \underline{\alpha}}{\alpha}+ { \underline{\alpha}}^2 - 2{\underline{\alpha}}\right ] } { 2{\underline{\alpha } } } \end{split}\ ] ] it follows that we have a unique solution provided @xmath362 and @xmath363 and that the coefficients are real - analytic in @xmath364 we now check the last statement . for @xmath365 , we have @xmath366 so that the claim follows . for @xmath367 we observe that the numerator of @xmath136 is a cubic polynomial that has three roots at @xmath368 thus , if @xmath369 then @xmath16 is the smallest root . since this cubic polynomial is positive for large negative @xmath355 and the denominator also is positive , it follows that @xmath136 is positive for @xmath370 so @xmath371 is a concave function , and by construction @xmath372 and @xmath373 . the claim follows . we will use the notation @xmath374 as well as the notation @xmath375 define @xmath376 by @xmath377 where @xmath378 define @xmath379 by @xmath380 observe that the conditions on @xmath137 imply that @xmath381 and @xmath382 are continuous on @xmath314 so that @xmath328 is @xmath383 we will use this function @xmath328 to normalize the triangle @xmath384 this is made possible by the following lemma . [ lem : bprop ] suppose that @xmath385 for each @xmath386 then the map @xmath328 is a @xmath324 diffeomorphism from @xmath314 onto @xmath200 . + in particular , for each @xmath387 and each @xmath388 , the mapping @xmath328 is a @xmath324 diffeomorphism from @xmath314 onto @xmath389 it suffices to show that the map @xmath390 is a @xmath324 diffeomorphism from @xmath391 onto @xmath392 . by assumption @xmath393 for each @xmath327 . we have @xmath394 , @xmath395 and @xmath396 . since @xmath390 is the identity for @xmath397 , we find that @xmath390 is a @xmath324 diffeomorphism from @xmath391 onto @xmath398 $ ] . for each @xmath399 and each @xmath400 , we define @xmath401 and @xmath402 by @xmath403 we then have [ lem : phi_analytic ] for each @xmath404 , each of the following maps is analytic on @xmath401 : 1 . @xmath405 2 . @xmath406 3 . @xmath407 if , for each @xmath408 the assumption of lemma [ lem : bprop ] holds then the map @xmath409 is also analytic on @xmath402 . + moreover , each restriction extends analytically to an open neighbourhood . the coefficients of the cubic polynomial @xmath338 depend analytically on @xmath410 and hence @xmath411 is analytic . the map @xmath412 is analytic and hence it follows that map ( 1 ) is analytic . maps ( 2 ) and ( 3 ) are therefore analytic . since @xmath411 is analytic and @xmath413 for @xmath414 , the implicit function theorem ( theorem 2.1.2 in @xcite ) implies that there exists a function @xmath415 which is analytic and a solution to @xmath416 we then have @xmath417 and , since @xmath418 is analytic , the claim follows . in the rest of the section , @xmath387 will be fixed so that we can use lemmas [ lem : bprop ] and [ lem : phi_analytic ] . we use the family of diffeomorphisms @xmath328 to define a quadratic form @xmath29 with domain @xmath419 that is unitarily equivalent to @xmath23 on @xmath420 . define @xmath421 by @xmath422 where @xmath423 is the operator that returns the jacobian matrix of a map . [ lem : domain ] the isometry @xmath424 is a unitary isomorphism from @xmath425 onto @xmath426 and it maps @xmath234 onto @xmath427 . + on functions that are supported in @xmath428 , @xmath424 coincides with @xmath218 we have @xmath429 it follows that @xmath424 is a unitary isomorphism from @xmath425 onto @xmath426 . let @xmath430 since @xmath328 is a @xmath431 diffeomorphism and @xmath432 is continuous on @xmath314 and smooth away from @xmath24 , then @xmath433 is continuous and in @xmath434 . the jump formula implies that @xmath435 + since , for @xmath436 , @xmath437 , the last statement is a direct verification . define the quadratic form @xmath438 on @xmath439 by @xmath440 [ lem : cuspequivalence ] @xmath25 is a cusp form for @xmath438 if and only if @xmath441 is a cusp form for @xmath26 on @xmath314 . if @xmath442 , then @xmath443 . it follows that if @xmath442 , then @xmath152 if and only if @xmath444 . for @xmath445 , the function @xmath446 is a solution to a second order ordinary differential equation , and hence @xmath444 for @xmath447 if and only if @xmath444 for @xmath448 . it will be convenient to have the following alternate form for @xmath29 . [ prop : alternate ] we have @xmath449 where @xmath450 is defined by @xmath451 and @xmath452 is defined by @xmath453 moreover , @xmath438 represents the cusp of width @xmath69 for @xmath235 this is a straightforward calculation using the chain rule and the change of variables formula . let @xmath454t_-,t_+[$ ] and let @xmath330 be a real - analytic path . [ thm : analyticity ] the family quadratic forms @xmath455 is analytic of type ( a ) in the sense of @xcite . for each @xmath12 , the quadratic form @xmath456 is a closed form with domain @xmath457 . it suffices to show that for each @xmath236 , the function @xmath458 is real - analytic . by proposition [ prop : alternate ] , we have @xmath459 where @xmath460 if @xmath461 \times [ \underline{\alpha } , \infty[$ ] , then the matrix @xmath462 is given by @xmath463 and @xmath464 . thus , the second integral on the right of ( [ eq : splitintegral ] ) depends analytically on @xmath12 . it remains to consider the integral over @xmath334\times [ 1,\underline{\alpha}]$ ] . the integrand @xmath465 can be expanded into a finite sum of terms of the form @xmath466 where @xmath467 is a function that is obtained by multiplying @xmath468 or its derivatives and the entries of @xmath469 and @xmath69 is one of the @xmath470 functions obtained by making the product @xmath471 where both @xmath472 are either @xmath25 or one of its partial derivatives . by lemma [ lem : phi_analytic ] , the coordinates of @xmath328 and @xmath473 are analytic functions of @xmath317 . it follows that @xmath474 and @xmath475 are analytic ( in a neighborhood of @xmath476 \times [ 1 , \underline{\alpha}]$ ] ) . in all possible choices , the function @xmath467 then is analytic . the analyticity of @xmath458 follows from lemma [ lem : intanalytic ] below . [ lem : intanalytic ] if @xmath477 \times [ 1 , \underline{\alpha}]$ ] is analytic , then for each @xmath478 \times [ 1 , \underline{\alpha}])$ ] , the function @xmath479 is analytic on @xmath282 . there exists an open neighborhood @xmath480 of @xmath476 \times [ 1 , \underline{\alpha}]$ ] such that the map @xmath481 extends to a holomorphic function on @xmath482 . since @xmath334 \times [ 1,\underline{\alpha}]$ ] is compact , @xmath483 converges uniformly to @xmath484 as @xmath12 approaches @xmath140 . it follows that the ( complex ) @xmath12-derivative of the map in ( [ eq : intanalytic ] ) exists at each @xmath485 . given theorem [ thm : analyticity ] , we now explain why the generic triangle @xmath314 has no cusp forms provided that one triangle has none . if there exists a point @xmath486 such that @xmath26 on @xmath487 has no nonconstant eigenfunction , then @xmath320 has measure zero and is a countable union of nowhere dense sets . by proposition [ prop : allcuspform ] , the quadratic form @xmath23 on @xmath487 has no cusp form , and hence by lemma [ lem : cuspequivalence ] , the quadratic form @xmath488 has no cusp form . to show that @xmath320 has measure zero , we apply fubini s theorem in a fashion similar to @xcite : let @xmath489 and apply lemma [ alt : alt ] to find that the set @xmath137 of @xmath69 such that @xmath490 is countable . for each @xmath491 , let @xmath492 and apply lemma [ alt : alt ] to find that the intersection @xmath493 of the line @xmath494 with @xmath320 is countable . hence for each @xmath491 , the set @xmath493 has measure zero with respect to the linear measure @xmath495 . hence , the measure of @xmath320 equals the measure of @xmath496 . since @xmath137 is countable , the measure equals zero . for @xmath497 , let @xmath498 be the set of @xmath321 such that @xmath26 on @xmath499 has a cusp form with eigenvalue at most @xmath500 . using the continuity of @xmath501 and the continuity of linear functional @xmath292 , one can show that @xmath498 is a closed . thus , it suffices to show that @xmath498 is nowhere dense . given a point @xmath502 , let @xmath503 \rightarrow { { \mathcal m}}$ ] be a real - analytic path joining @xmath504 to @xmath317 . since @xmath23 on @xmath487 has no cusp forms , the family @xmath455 has no cusp form eigenfunction branch . it follows from lemma [ alt : alt ] , that for each open neighborhood @xmath482 of @xmath317 , there exists @xmath505 $ ] such that @xmath506 and @xmath507 has no cusp forms . hence @xmath498 is nowhere dense . in the remainder of this paper we consider the specific family of triangles @xmath508 defined in the introduction . in particular , we will study the spectral properties of @xmath509 for small @xmath12 . the family @xmath509 of quadratic forms does not converge as @xmath12 tends to zero nor do its real - analytic eigenbranches . but a simple renormalization will give convergence . fix @xmath322 and @xmath399 such that @xmath510 . let @xmath137 be the function defined in lemma [ lem : defnb ] . when @xmath355 tends to @xmath16 , the function @xmath511 converges to @xmath16 uniformly for @xmath512.$ ] thus , there exists @xmath513 such that if @xmath514 and @xmath515 then @xmath516 choose @xmath301 such that @xmath517 then , for each @xmath518 and each @xmath519 @xmath520 so that @xmath521 we may thus use lemmas [ lem : bprop ] and [ lem : phi_analytic ] . the methods and results of section [ subsec : analyticity ] then apply and we define the quadratic form @xmath509 as previously . for each @xmath522 , define the renormalized quadratic form by @xmath523 with domain @xmath234 . by theorem [ thm : analyticity ] , the family @xmath27 is real - analytic of type ( a ) for @xmath40,1[$ ] . in particular , the results of [ subsec : generic ] apply . to study the limiting properties of the family @xmath29 , we re - express @xmath29 in a more convenient form : for each @xmath324 function @xmath524 define @xmath525 recall that @xmath526 is the inverse of @xmath338 and set @xmath527 . define @xmath528 and @xmath529 where the subscript ( or first argument ) in each @xmath530 and @xmath137 is @xmath531 . when comparing @xmath532 and @xmath533 ( @xmath469 and @xmath534 ) we see that we only miss some powers of @xmath12 than eventually cancel in the computation leading to proposition [ prop : alternate ] . this shows that for each @xmath236 @xmath535 where @xmath536\times [ 1 , \underline{\alpha}]$ ] and @xmath537\times [ \underline{\alpha } , \infty]$ ] . by arguing as in the proof of theorem [ thm : analyticity ] , one can show that @xmath538 is analytic at @xmath33 . is not closed on the domain @xmath234 , the family @xmath29 is _ not _ analytic at @xmath33 in the sense of @xcite . ] we will now compute the first few terms in the taylor series in @xmath12 for @xmath539 and @xmath540 these functions are analytic on a neighbourhood of @xmath541 in particular , in the following , the expressions like @xmath542 are uniform with respect to @xmath543 and may be differentiated with respect to @xmath544 and @xmath545 . we first compute @xmath546 and @xmath547 since @xmath548 is analytic and @xmath549 , it follows from ( [ eq : fexp ] ) that @xmath550 moreover , using analyticity , this asymptotic expansion may be differentiated with respect to @xmath551 we thus obtain , @xmath552 substituting these into ( [ defn : rhotilde ] ) , and differentiating , we find that @xmath553 using ( [ eq : fexp ] ) , ( [ eq : fderivexp ] ) , ( [ exp : yalpha ] ) , and ( [ exp : yderivalpha ] ) we find that @xmath554 where @xmath282 is the identity matrix , @xmath542 is a matrix whose operator norm is bounded by a constant times @xmath31 as @xmath12 tends to zero , and @xmath555 is the polynomial @xmath556 to prove theorem [ thm : main ] we will need to know that @xmath557 . [ lem : p_positive ] @xmath558 . by construction we have @xmath559 by differentiating with respect to @xmath355 and setting @xmath365 we get @xmath560 since @xmath561 the claim follows . in this section we apply the method of asymptotic separation variables developed in @xcite to the family @xmath29 . using the small @xmath12 asymptotics derived in [ section : expansion ] , we approximate @xmath29 to first order with a family of quadratic forms @xmath35 for which separation of variables apply . we also derive a non - concentration estimate for eigenfunctions of @xmath29 . in this section and the following sections , we will use @xmath562 in place of @xmath563 as coordinates for @xmath564 \times [ 1 , \infty[$ ] and unless it is specified otherwise @xmath565 is the norm in @xmath566 we begin by using the expansions obtained in [ section : expansion ] to determine the forms used to approximate @xmath29 . in particular , by substituting the expansions ( [ exp : rho ] ) and ( [ exp : q ] ) into ( [ eq : qdefn ] ) we are led to define @xmath567 and @xmath568 where the operator @xmath569 is defined by ( [ eq : gradient ] ) , @xmath536 \times [ 1 , \underline{\alpha}]$ ] , and @xmath570 we wish to show that @xmath29 is asymptotic to @xmath35 at first order in the sense of @xcite . it will also be used to help derive a key estimate for crossing eigenbranches . however , although @xmath35 is a positive quadratic form , the bottom of its spectrum tends to @xmath571 so that it is more convenient to use the quadratic form @xmath572 that we now define to control quantities . [ def : tildea ] the quadratic form @xmath572 is defined on @xmath573 by @xmath574 the following proposition can be seen as the beginning of an asymptotic expansion for @xmath575 [ prop : q - a - tb ] there exists @xmath353 such that for each @xmath576 @xmath577 we have @xmath578 if @xmath579 , then @xmath580 is identically equal to @xmath16 and @xmath581 is identically equal to @xmath282 . hence , by subsituting ( [ eq : nablarhow ] ) into ( [ eq : qdefn ] ) , we find that @xmath582 is the sum of the following four terms @xmath583 in order to estimate these four terms , we use the asymptotic expansions of [ section : expansion ] . for example , by ( [ exp : q ] ) we have that ( [ leading ] ) is equal to @xmath584 since the operator norm of the matrix @xmath542 is bounded by a constant @xmath353 times @xmath31 , we can apply the cauchy - schwarz inequality to find that the norm of ( [ exp : q ] ) is bounded by @xmath585 . similar arguments show that there is a constant @xmath353 so that * ( [ mixed ] ) is bounded above by @xmath586 * ( [ mixed2 ] ) is bounded above by @xmath587 * ( [ uv ] ) is bounded above by @xmath588 the claim follows . we recall the fourier decomposition of section [ sec : fourier ] . since now @xmath589 we thus have , for each @xmath590 @xmath591 where the latter makes sense for almost every @xmath326 and defines an element of @xmath592 as above , let @xmath593 denote the set of functions @xmath594 such that @xmath77 is the restriction of a compactly supported , smooth function defined in a neighborhood of @xmath200 . if @xmath595 , then each @xmath119 is smooth , and a straightforward computation shows that is the operation defined by @xmath596 . ] @xmath597 we define @xmath598 to be the set of compactly supported , smooth functions defined on @xmath599 . for @xmath600 , each integer @xmath40 , and each @xmath601 , we define @xmath602 for @xmath603 in @xmath604 , the inner product is defined by @xmath605 let @xmath606 denote the subspace consisting of those functions whose support lies in @xmath607 $ ] . for each @xmath126 , the quadratic form @xmath608 extends to a closed , densely defined form on the completion of @xmath609 with respect to @xmath610 . for @xmath611 , we will restrict the domain of @xmath608 to be the completion of those smooth functions whose support lies in @xmath607 $ ] . if @xmath25 is an eigenfunction of @xmath35 with eigenvalue @xmath260 , then for each @xmath77 in the domain of @xmath608 , we have @xmath612 and hence @xmath119 is an eigenfunction of @xmath608 with eigenvalue @xmath260 with respect to @xmath613 . thus , each eigefunction @xmath25 of @xmath35 may be written uniquely as @xmath614 where the @xmath120 fourier coefficient @xmath119 is an eigenfunction of @xmath608 . moreover , the spectrum of @xmath35 with respect to @xmath615 is the union of the spectra of @xmath608 with respect to @xmath613 . in what follows we will often suppress the subscript @xmath326 from the notation . the next two lemmas identify the eigenfunctions of @xmath608 for each @xmath40 . we begin with the nonzero modes . [ lem : nonzero_fourier ] for each @xmath12 and @xmath40 and eigenvalue @xmath260 of @xmath608 with respect to @xmath613 , the associated eigenspace consists of functions of the form @xmath616 such that @xmath617 moreover , when @xmath12 varies , the spectrum is organized into eigenvalues branches @xmath618 each of these is a growing function of @xmath12 and @xmath619 integrate by parts as in the proof of lemma [ zerolemma ] , make the change of variables @xmath620 , and use the boundary conditions . the second part follows from theorem [ thm : eigenbasis_exist ] . the zero modes are given by the following lemma . [ zerolemma ] the spectrum of @xmath621 with respect to @xmath613 is the set @xmath622 the eigenspace associated to @xmath623 is spanned by the eigenvector @xmath624 suppose that @xmath77 is an eigenfunction , that is @xmath625 for all @xmath69 . this implies first that @xmath626 holds in the distributional sense . ellipticity then yield that @xmath77 is smooth . moreover , by integrating by parts against a smooth function that is identically equal to @xmath16 near @xmath627 , we also find that @xmath628 let @xmath140 be such that @xmath629 . then two linearly independent solutions are given by @xmath630 and @xmath631 if @xmath632 and by @xmath633 and @xmath634 if @xmath635 . the condition that @xmath636 is real and nonnegative implies either that @xmath637 with @xmath638 , that @xmath639 or that @xmath635 . if @xmath640 , then the boundary conditions @xmath641 and @xmath642 imply that the solutions take the form given in in ( [ zeroeigenvector ] ) with @xmath643 . if @xmath644 $ ] , then there are no solutions that satisfy the boundary conditions . as a consequence of the identification of the eigenfunctions we have the following poincar type inequality . [ lem : poincare ] for each @xmath645 and each @xmath220 we have @xmath646 we have @xmath647 lemmas [ lem : nonzero_fourier ] and [ zerolemma ] imply @xmath648 and @xmath649 for @xmath650 . in the sequel we will use different kind of projections associated either with the fourier decomposition @xmath651 or with the spectral decomposition of @xmath652 more precisely , for each @xmath653 define the orthogonal projection @xmath654 by @xmath655 and let @xmath656 denote the image of @xmath657 . for each @xmath658 , we define @xmath659 the projection onto @xmath660 . we also define @xmath661 to be the orthogonal projection onto the eigenspace of @xmath35 associated to the eigenvalue @xmath260 of @xmath35 . for a fixed interval @xmath282 , define the @xmath35-spectral projection in the energy interval @xmath282 to be @xmath662 for each eigenvalue @xmath260 of @xmath35 , the associated @xmath35-eigenspace @xmath663 is the orthogonal direct sum @xmath664 where @xmath665 is the @xmath260-eigenspace of @xmath666 and @xmath667 is the span of @xmath668 . it follows that @xmath669 more generally , for a quadratic form @xmath545 the notation @xmath670 will always denote the spectral projection onto the interval @xmath233 we prove that @xmath35 and @xmath29 are asymptotic at first order in the sense of @xcite . in the following , we let @xmath671 ( resp . @xmath672 ) denote the derivative of @xmath29 ( resp . @xmath35 ) in @xmath12 . [ prop : q - a_asymp ] there exists a constant @xmath353 and @xmath301 such that , for all @xmath673 and all @xmath674 @xmath675 one argues as in the proof of proposition [ prop : q - a - tb ] paying a little more attention to the terms ( [ mixed ] ) , ( [ mixed2 ] ) , and ( [ uv ] ) . for example , to estimate ( [ mixed ] ) , use the cauchy - schwarz inequality and ( [ exp : rho ] ) to obtain @xmath676 the cauchy - schwarz inequality and lemma [ zerolemma ] give @xmath677 by combining ( [ est : nabla_rho ] ) and ( [ est : v_nabla ] ) and using ( [ exp : rho ] ) we find that @xmath678 switching the roles of @xmath25 and @xmath77 , we obtain the same bound for the expression in ( [ mixed2 ] ) . similar methods apply to bound the other terms . the estimate for @xmath679 is obtained in a similar way . since @xmath29 is asymptotic to @xmath35 at first order and @xmath35 and @xmath672 are nonnegative quadratic forms , each real - analytic eigenvalue branch @xmath41 of @xmath29 converges to a finite limit @xmath680 as @xmath12 tends to zero ( theorem 3.4 of @xcite ) . for the dirichlet eigenvalue problem on @xmath0 , we showed in @xcite that each limit @xmath680 has the form @xmath42 where @xmath40 is an integer . the methods of @xcite can be applied to show that the same fact is true in the present context . in this section we highlight the necessary modifications . we also show that if the eigenvalue branch is associated to a cusp form , then @xmath40 must be positive . this latter fact will be used crucially in the proof of theorem [ thm : main ] . the proof of convergence depends crucially on the following ` non - concentration ' result proved for the dirichlet problem in @xcite . we will denote by @xmath681 the domain of the quadratic form @xmath682 and by @xmath683 the quadratic form defined for @xmath684 by @xmath685 [ prop : real_noncon](compare proposition 9.1 of @xcite ) let @xmath653 , let @xmath269 be a compact subset of @xmath132(\pi \ell)^2 , \infty[$ ] , and let @xmath686 . there exist positive constants @xmath301 and @xmath687 ( that only depend on @xmath688 and @xmath353 ) so that if @xmath689 , if @xmath518 , and if for each @xmath690 , the function @xmath691 satisfies @xmath692 then @xmath693 if @xmath694 this is proposition 9.1 of @xcite with @xmath695 and @xmath696 . if @xmath697 and we let @xmath698 , then ( [ nonconcentrate ] ) holds . in the language of semi - classical analysis , propostion [ prop : real_noncon ] asserts that a quasimode @xmath77 of order @xmath12 at energy @xmath102 does not concentrate at @xmath699 if @xmath700 . in 12 of @xcite , we used non - concentration to derive indirect estimates for @xmath701 . the following proposition and corollary make these estimates more transparent and simpler to apply . [ prop : noncon ] let @xmath653 and let @xmath702 ( \pi \ell)^2,\infty[$ ] be compact . for each @xmath703 , there exists @xmath704 and @xmath705 such that for each @xmath691 and @xmath518 @xmath706 where @xmath707 from ( [ defn : ak ] ) we find that @xmath708 and hence @xmath709 if the claim is not true , then for each @xmath704 , there exists a sequence @xmath710 tending to zero and sequences @xmath711 @xmath712 such that @xmath713 in particular , since @xmath714 , we have @xmath715 . it follows that for each @xmath716 @xmath717 we fix @xmath718 such that @xmath719+k \subset ( ( \ell \pi)^2 , \infty),$ ] and we set @xmath720}(\tilde{v}_n ) = p_{{\widetilde{a}_t^\ell}}^{[e_n+1-\delta , e_n+1+\delta]}(\tilde{v}_n).$ ] reasoning as in the proof of lemma 2.3 in @xcite we find that @xmath721 observe that the sequences @xmath722 and @xmath723 depend on the initial choice of @xmath724 but the preceding estimate gives a constant @xmath353 that is independent of @xmath724 such that @xmath725 this implies in particular @xmath726 so that , for @xmath727 large enough we have @xmath728 in equation ( [ eq : tvnisqm ] ) we replace the test function @xmath69 by @xmath729}(w)$ ] and use that the spectral projector is self - adjoint and commutes with @xmath730 we obtain that for each @xmath731 , @xmath732}(w ) \right)^{\frac{1}{2}}\\ & \leq c \cdot t_n \cdot \| v_n\| \|w\| , \end{split}\ ] ] where we have used that @xmath733 is controlling @xmath734 and that @xmath735}(w)\right ) \,\leq\ , ( \sup(k)+\delta)\|w\|^2\ ] ] by definition of a spectral projector . since @xmath736 and @xmath737 is a non - negative quadratic form , we also have @xmath738 equation ( [ eq : vnisqm ] ) implies that we may use proposition [ prop : real_noncon ] to find @xmath739 since ( [ eq : vnisqm ] ) also implies @xmath740 using ( [ eq : adot_express ] ) we find that @xmath741 on the other hand , the contradiction assumption implies that @xmath742 the implied constant @xmath353 does not depend on @xmath724 so if we take @xmath743 then choosing @xmath744 small enough yields the contradiction . this proposition yields an estimate for @xmath745 from below in terms of the projection @xmath746 [ coro : noncon ] let @xmath48 and let @xmath747 be a compact subset of @xmath132(\pi k)^2 , \infty[$ ] . for each @xmath748 , there exists @xmath749 @xmath705 such that if @xmath689 , @xmath750 , and @xmath751 , then @xmath752 where @xmath753 the functional @xmath754 is equivalent to the @xmath755-norm of @xmath756 where here @xmath757 is the operator such that @xmath758 for each @xmath759 . since @xmath672 is block diagonal with respect to the sum @xmath760 and @xmath761 , we have @xmath762 we may apply proposition [ prop : real_noncon ] with @xmath763 to each term on the right hand side to find that @xmath764 where @xmath687 is the minimum of the @xmath724 coming from proposition [ nonconcentrate ] . for each @xmath765 , and @xmath766 , we have @xmath767 and hence @xmath768 . we also have , @xmath769 , and the claim follows . the bounds proved in [ sec : noncon ] depend on a bound on @xmath770 . in this subsection , we show that if @xmath69 is an @xmath35-spectral projection of a @xmath29-eigenfunction in an interval containing the eigenvalue @xmath102 , then @xmath770 is of order @xmath12 . we start with a real - analytic eigenfunction branch @xmath30 for @xmath29 with associated real - analytic eigenvalue branch @xmath41 . we let @xmath771 denote the associated spectral projection . let @xmath680 denote the limit of @xmath41 as @xmath12 tends to zero . the following two lemmas express the fact that the projection @xmath772 is an order @xmath12 quasimode for @xmath35 at energy @xmath41 . first we have a lemma that is comparable to lemma 2.3 in @xcite . [ lem : u_w_close ] if @xmath282 is a compact interval whose interior contains @xmath680 , then there exist @xmath705 and @xmath353 such that if @xmath518 , then @xmath773 using the fact that @xmath30 is an eigenfunction of @xmath29 and that @xmath35 and @xmath29 are asymptotic at first order , for each @xmath774 @xmath775 observe that letting @xmath776 yields that @xmath777 moreover the former equation can be rewritten as @xmath778 where $ \tilde{e}_t:= e_t+1.$ we may now follow the proof of lemma 2.3 in \cite{hj10 } observing that $ p_{a_t}^i = p_{{\widetilde{a}_t}}^{i+\{1\}}.$ this yields a constant $ c$ such that \begin{equation * } \begin{split } a_t(u_t - w_t^i)\,+\,\|u_t - w_t^i\|^2\ , & \leq \ , c\cdot t^2 \cdot { \widetilde{a}_t}(u_t)\\ & \leq \frac{c\cdot t^2}{1-c\cdot t^2 } \|u_t\|^2,\\ & \leq c'\cdot t^2 \|u_t\|^2 . \end{split}\ ] ] the claim follows . observe that by orthogonality the preceding lemma also imples the following estimate that roughly expresses the fact that all the mass of @xmath30 is supported by @xmath779 @xmath780 [ lem : quasimode ] if @xmath282 is a compact interval whose interior contains @xmath680 , then there exist @xmath705 and @xmath353 such that if @xmath518 , then @xmath781 for each @xmath782 we have @xmath783 so that the cauchy - schwarz inequality and the preceding lemma imply @xmath784 we also have using cauchy - schwarz and the preceding lemma . @xmath785 we now start again from ( [ eq : afo ] ) . first in the bounding term , we have already seen that we could replace @xmath786 by @xmath787 thus from the triangle inequality , ( [ eq : afo ] ) and the two preceding estimates we obtain @xmath788 the claim follows using ( [ eq : normwequivnormu ] ) . lemma [ lem : quasimode ] has the following corollary that expresses , in the language of semiclassical analysis , that @xmath779 is an order @xmath12 quasimode . [ coro : wisqm ] if @xmath282 is a compact interval that contains @xmath680 there exists @xmath353 and @xmath705 such that , for @xmath518 and each @xmath789 , we have @xmath790 since @xmath791 is a spectral projector , we have @xmath792 and hence lemma [ lem : quasimode ] implies @xmath793 since @xmath794 , the claim follows . by combining lemma [ lem : quasimode ] with corollary [ coro : noncon ] we prove the following . [ thm : limitsq ] let @xmath47 be an eigenbranch of @xmath29 then there exists @xmath795 such that @xmath796 suppose to the contrary that @xmath680 is not of the form @xmath797 where @xmath40 is an integer . let @xmath798 . choose a compact interval @xmath799(n-1)^2\pi^2,n^2\pi^2[$ ] whose interior contains @xmath680 . let @xmath30 be a real - analytic eigenfunction branch of @xmath29 associated to @xmath41 . let @xmath800 be the projection of @xmath30 onto the modes of @xmath35 that have energy lying in @xmath282 . if @xmath801 , then each eigenvalue of @xmath666 is at least @xmath802 . ( this follows from lemma [ lem : lambda_asymp ] or simply the nonnegativity of @xmath672 . ) thus , since @xmath803 , equation ( [ eq : projection_relation ] ) implies that @xmath804 let @xmath353 be as in lemma [ lem : quasimode ] , and apply lemma [ coro : noncon ] with @xmath805 to obtain @xmath687 so that @xmath806 it follows that @xmath807 is not integrable . this contradicts theorem 4.2 of @xcite . the next proposition will be the starting point of the contradiction argument in the following sections . it says that a cusp form eigenbranch can not limit to @xmath571 . heuristically , the zeroth fourier coefficient of a cusp form vanishes identically whereas an eigenvalue branche that limit to @xmath571 must eventually live have nontrivial zeroth fourier mode . however , because we have made a nontrivial change of variable , this fact requires an argument . [ prop : cflimnot0 ] if @xmath41 is a real - analytic cusp form eigenvalue branch of @xmath29 , then the integer @xmath40 appearing in ( [ eq : elimit ] ) is positive . suppose to the contrary that @xmath808 . set @xmath809 $ ] and consider @xmath810 defined in ( [ defn : wti ] ) . if @xmath694 , the restriction of @xmath811 to @xmath656 is bounded below by @xmath812 , thus we have @xmath813 on the other hand , the projection of @xmath30 onto @xmath814 equals @xmath815 . let @xmath816 denote the projection of @xmath817 onto @xmath59 . since each @xmath656 is a direct sum of eigenspaces of @xmath811 , we have @xmath818 the quadratic form @xmath811 is nonnegative and the restriction of @xmath35 to @xmath814 is bounded below by @xmath819 . hence @xmath820 . by lemma [ lem : u_w_close ] we have @xmath821 therefore , @xmath822 , and hence @xmath823 for small @xmath12 . lemma [ lem : cuspequivalence ] implies that @xmath824 is a cusp form for @xmath26 . in particular , for each @xmath825 $ ] @xmath826 where @xmath827 and @xmath828 . inspection shows that there exists @xmath353 such that @xmath829\times[1,\beta]\right \ } \ , \leq \ , c \cdot t \ ] ] for small @xmath12 . therefore , using the cauchy - schwarz inequality we find that @xmath830 using the fundamental theorem of calculus and the cauchy - schwarz inequality we find that @xmath831 thus , since there exists @xmath832 so that @xmath833\times[1,\beta]\right \ } \leq c ' \cdot t\ ] ] for small @xmath12 , we have @xmath834 since @xmath35 and @xmath29 are asymptotic at first order @xmath835 for sufficiently small @xmath12 . by combining these estimates , we find that @xmath836 since @xmath837 we can combine ( [ eq : zero_zero ] ) , ( [ est : etaminus1 ] ) and ( [ est : gradient ] ) to find that @xmath838 squaring and integrating over @xmath825 $ ] , we find that @xmath839 without loss of generality , we have @xmath840 . therefore , if @xmath41 limits to zero , then ( [ est : zerosmall ] ) will contradict ( [ est : zerolarge ] ) for small @xmath12 . we can also use the nonconcentration of the spectral projection to give an @xmath841 lower bound on the first variation @xmath842 in the case that the projection onto the small modes is significant . [ prop : adotbound ] let @xmath282 be a compact interval whose interior contains @xmath680 and @xmath843 . for each @xmath844 , there exists @xmath704 and @xmath705 so that if @xmath845 and @xmath846 then @xmath847 using the cauchy - schwarz inequality and the nonnegativity of @xmath672 we have @xmath848 it follows from ( [ eq : defa_t ] ) that for all @xmath849 @xmath850 and hence @xmath851 . moreover , by combining this with lemma [ lem : u_w_close ] we find @xmath852 thus , from ( [ est : adotu_adotw ] ) we obtain @xmath853 hence by applying lemma [ lem : u_w_close ] we have @xmath854 for @xmath12 sufficiently small . as in the proof of theorem [ thm : limitsq ] , we have @xmath855 . since @xmath701 is non - negative and ` block - diagonal ' we have @xmath856 let @xmath353 be as in lemma [ lem : quasimode ] , and apply lemma [ coro : noncon ] with @xmath857 to obtain @xmath687 so that @xmath858 for @xmath12 sufficiently small . estimate ( [ eq : normwequivnormu ] ) implies that @xmath859 for @xmath12 small , and therefore by combining the above inequalities , we prove the claim . in contrast to proposition [ prop : adotbound ] , we have the following . [ edotupperestimate ] there exists @xmath705 and @xmath353 such that if @xmath845 , then @xmath860 it follows from ( [ eq : defa_t ] ) that for all @xmath849 @xmath850 from ( [ eq : q - a ] ) , there exists @xmath353 so that for sufficiently small @xmath12 @xmath861 and @xmath862 thus , if @xmath30 is the real - analytic eigenfunction branch of @xmath29 associated to @xmath41 , then @xmath863 by choosing @xmath301 sufficiently small , we obtain the claim . proposition [ alt : alt ] reduces the proof of theorem [ thm : main ] to the following . [ thm : nobranch ] the family @xmath27 does not have a real - analytic cusp form eigenbranch . the proof of theorem [ thm : nobranch ] will be by contradiction . we will assume that there exists a real - analytic cusp form eigenvalue branch , @xmath41 . by the results of [ section : branches ] , we have @xmath864 where @xmath40 is positive . we aim to contradict the positivity of @xmath40 . the proof of theorem [ thm : nobranch ] will rely on the estimates of solutions to ordinary differential equations made in appendix [ appendix : airyb ] . to make the estimates less tedious , we will choose @xmath249 to be sufficiently close to 1 , where ` sufficiently close ' will be determined by the integer @xmath40 that appears in ( [ eq : limitk ] ) . however , the construction of the quadratic form @xmath29 depends on @xmath249 . therefore , the integer @xmath40 that appears in ( [ eq : limitk ] ) depends a priori on @xmath249 . in order to avoid circularity of reasoning , we will prove the following . [ prop : choosebeta ] let @xmath41 be a real - analytic eigenvalue branch associated to a real - analytic cusp form eigenfunction branch of @xmath865 . for each @xmath866 , the family @xmath867 has a real - analytic cusp form eigenfunction branch with associated eigenvalue branch @xmath41 .. ] for each fixed @xmath12 , since @xmath41 corresponds to a cusp form , it belongs to the spectrum of @xmath868 for all @xmath869 ( see lemma [ lem : betacf]).by theorem [ thm : eigenbasis_exist ] , there exists a real - analytic eigenvalue branch @xmath870 of @xmath871 such that @xmath872 . since @xmath873 has only countably many real - analytic eigenvalue branches , there exists some branch @xmath870 such that the set of @xmath874 with @xmath875 has an accumulation point . thus , by real - analyticity , we have @xmath875 for all @xmath874 . if for each @xmath12 , the dimension of the eigenspace @xmath876 of @xmath877 associated to @xmath878 is greater than one , then one can argue as in the proof of proposition [ prop : mult ] to obtain a real - analytic cusp form eigenfunction branch of @xmath877 associated to @xmath41 . otherwise , by real - analyticity , for each @xmath12 in the complement of discrete set @xmath136 of @xmath12 we have @xmath879 . let @xmath880 be a real - analytic eigenfunction branch of @xmath877 associated to @xmath41 . let @xmath881 denote a real - analytic eigenfunction branch of @xmath882 associated to @xmath41 . for each @xmath12 , the pull - back of @xmath30 by the diffeomorphism @xmath883 is a cusp form of @xmath26 on @xmath0 . in turn , for each @xmath12 , the pull - back of @xmath884 by @xmath885 is a cusp form eigenfunction of @xmath877 . hence , if @xmath886 , then the eigenfunction @xmath887 is a cusp form for @xmath886 . thus , by corollary [ coro : cf_discrete ] , the branch @xmath880 is a real - analytic cusp form eigenbranch of @xmath877 . as a consequence of proposition [ prop : choosebeta ] , we may fix @xmath249 to satisfy is most probably not necessary but it will simplify the arguments in the appendix . in particular it implies that , on @xmath888 $ ] and for @xmath889 , the sturm - liouville equations associated with @xmath890 have no turning point . ] @xmath891 it follows that for each @xmath889 and @xmath892 $ ] we have @xmath893 as soon as @xmath12 is small enough . in what follows , we will drop @xmath249 from the notation for @xmath882 . in this section we show that there exists a real - analytic eigenvalue branch @xmath894 of @xmath608 such that @xmath895 is at most of order @xmath12 . in [ sec : relative ] , we will show to the contrary that @xmath895 is at least of order @xmath896 . this will provide the desired contradiction . [ tracking ] if @xmath41 is a cusp form eigenvalue branch with positive limit @xmath58 , then there exists @xmath705 , @xmath686 , and a real - analytic eigenvalue branch @xmath49 of @xmath608 so that for each @xmath518 , @xmath897~ = ~ \{\lambda^*_t\}. \ ] ] it suffices to show that there exist positive constants @xmath301 and @xmath353 so that the distance from @xmath41 to the spectrum of @xmath608 is at most @xmath898 for @xmath518 . indeed , if this were so , then for each @xmath12 we would have an eigenvalue @xmath894 of @xmath608 so that @xmath899 . suppose that there existed a sequence @xmath900 and eigenvalues @xmath901 of @xmath902 so that @xmath903 . then we would have @xmath904 this would contradict the ` super - separation ' of eigenvalues given in theorem 10.4 in @xcite . hence there exists @xmath905 such that if @xmath906 , then at most one eigenvalue @xmath894 of @xmath608 satisfies @xmath899 . since the eigenvalues of the family @xmath608 belong to real - analytic branches , the function @xmath907 would be real - analytic . we first proceed to prove that there exists a sequence @xmath744 going to @xmath571 such that @xmath41 is at a distance of order @xmath12 of the spectrum of @xmath908 let @xmath30 denote a real - analytic eigenfunction branch associated to @xmath41 . by corollary [ coro : wisqm ] and the fact that @xmath909 is an orthogonal projection that commutes with @xmath35 there exists constants @xmath910 and @xmath301 such that for each @xmath77 and @xmath911 then @xmath912 let @xmath137 denote the set of @xmath601 so that the distance from @xmath41 to spectrum of @xmath608 is at least @xmath913 , where @xmath910 is the constant that appears in ( [ quasimodetrack ] ) . by the remark above , we want to show that there exists @xmath914 so that @xmath9150,t[$ ] is empty . for each @xmath9160,s[$ ] , we apply a resolvent estimate to ( [ quasimodetrack ] ) and obtain @xmath917 . orthogonality then implies that for each @xmath9160,s[$ ] , there exists @xmath918 such that @xmath919 by applying corollary [ coro : noncon ] and using the fact that @xmath701 is both nonnegative and ` block - diagonal ' with respect to @xmath920 , we find that @xmath921 for all @xmath9220 , t_3[\ , \cap b$ ] . by theorem 4.2 in @xcite , the function @xmath923 is integrable over @xmath1320 , t_3[$ ] . thus , by ( [ adotbound ] ) , the function @xmath46 is integrable over @xmath9150,t_3[$ ] . it follows that there exists a sequence @xmath924 with @xmath900 . let @xmath925 be a real - analytic eigenbranch of @xmath902 with @xmath926 . suppose that there exists a sequence @xmath927 with @xmath928 . we claim that there exists another constant @xmath353 and another sequence @xmath929 with @xmath930 so that @xmath931 . to see this , first note that by taking subsequences if necessary , we may assume that @xmath932 for each @xmath727 . let @xmath933 $ ] be the connected component of @xmath934)$ ] that contains @xmath744 . the graph of @xmath925 must cross the lines @xmath935 , and hence @xmath936 this establishes the existence of the desired sequence @xmath937 . by lemma 4.3 in @xcite , the function @xmath938 is bounded . thus , since @xmath939 , we can apply ( [ adotbound ] ) to find a constant @xmath749 so that @xmath940 for all sufficiently large @xmath727 . hence , @xmath941 , and thus @xmath942 but this contradicts lemma [ lambdadot ] below . [ lambdadot ] let @xmath650 and let @xmath744 be a sequence converging to zero . for each @xmath943 , let @xmath925 be a real - analytic eigenbranch of the family @xmath608 . if @xmath944 , then @xmath945 lemma [ lem : lambda_asymp ] as stated does not directly apply since the eigenvalue branch @xmath925 depends on @xmath727 . by keeping track of the constants in the proof of lemma [ lem : lambda_asymp ] , one can produce a version that directly implies lemma [ lambdadot ] . we prefer to give a direct proof here . let @xmath946 be a unit norm eigenfunction of @xmath902 with eigenvalue @xmath947 . by the standard variational formula @xmath948 since @xmath946 is an eigenfunction of @xmath902 with eigenvalue @xmath947 , we have @xmath949 it suffices to show that the right hand side of ( [ variation ] ) tends to zero as @xmath727 tends to infinity . let @xmath703 . since @xmath947 tends to @xmath42 , there exists @xmath844 so that if @xmath950 , then @xmath951 and thus @xmath952 to estimate the remaining integral over @xmath953 , we will apply a standard convexity estimate from the theory of ordinary differential equations . if @xmath727 is sufficiently large , @xmath954 , and hence there exists @xmath955 so that if @xmath956 , then @xmath957 it follows that there exists @xmath301 so that if @xmath751 , then @xmath958 since @xmath959 for sufficiently large @xmath727 , we may combine ( [ deltaplus ] ) with ( [ delta ] ) to show that ( [ variation ] ) is less than the given @xmath960 for sufficiently large @xmath727 . in this subsection , we show that @xmath961 is smaller than @xmath962 for @xmath140 near a _ crossing time _ , a value of the parameter @xmath12 such that @xmath41 belongs to the spectrum of @xmath621 . then we show that there exists a sequence of crossing times @xmath744 and intervals of width @xmath963 about the crossing times on which @xmath56 is smaller than @xmath53 . the proof of the first result depends on the analysis contained in appendix [ appendix : airyb ] . [ prop : xings - kpart ] given @xmath964 , there exists @xmath955 and @xmath705 such that if @xmath518 and @xmath965 then @xmath966 let @xmath967 be an eigenfunction of @xmath621 with eigenvalue @xmath968 satisfying @xmath969 . we have @xmath970 and hence @xmath971 in appendix [ appendix : airyb ] , we prove that there exists @xmath972 so that @xmath973 hence by applying the cauchy - schwarz inequality to the left hand side of ( [ eq : brhs ] ) , we find that @xmath974 let @xmath975 , and use ( [ dist53 ] ) to find that @xmath976 the claim follows by choosing @xmath301 sufficiently small . [ prop : xings ] for all @xmath977 there exists @xmath978 such that if @xmath979 , @xmath980 , and @xmath981 $ ] , then @xmath982 let @xmath983 be the eigenvalue branch of @xmath621 such that @xmath984 . by lemma [ zerolemma ] , we have @xmath985 for some @xmath61 , and hence @xmath986 . using the fact that @xmath35 and @xmath29 are asymptotic and the fact that @xmath701 is non negative , there exists a constant @xmath353 such that @xmath987 for all sufficiently small @xmath12 . thus , for even smaller @xmath12 we obtain @xmath988 since @xmath989 , integration over @xmath990 $ ] and exponentiation gives @xmath991 if @xmath992 , then @xmath993 where the last inequality holds for @xmath994 . by combining this with ( [ est : ratio_eigen ] ) , one finds that for @xmath995 $ ] , we have @xmath996 using lemma [ edotupperestimate ] , and @xmath997 , we have that @xmath998 an argument similar to the one above gives that @xmath999 for @xmath1000 $ ] . since by assumption @xmath1001 , there exists @xmath1002 so that if @xmath1003 , then @xmath1004 . therefore , by ( [ est : lambda_first ] ) , we have that @xmath1005 is bounded above by @xmath1006 for @xmath1007 . in sum , if @xmath1008 and @xmath995 $ ] , then @xmath1009 the claim follows by chosing @xmath718 carefully . we wish to estimate from below the size of the set of @xmath12 for which ( [ dist53 ] ) holds true . this is accomplished by the following proposition . [ prop : seqxings ] let @xmath844 . there is a sequence @xmath744 of crossing times such that @xmath1010 and if @xmath1011 are large enough , then the intervals @xmath1012 $ ] and @xmath1013 $ ] are disjoint . by lemma [ lem : lambda_asymp ] , there exists , @xmath1014 so that @xmath1015 . it also follows from proposition [ tracking ] that there exists @xmath1016 so that @xmath1017 for sufficiently small @xmath12 . by lemma [ zerolemma ] , the eigenvalues of @xmath621 have the form @xmath1018 where @xmath1019 and @xmath1020 is the increasing sequence of positive solutions to the equation @xmath1021 . standard asymptotic analysis shows that @xmath1022 fix @xmath1023 for each @xmath1024 $ ] and each @xmath1025 , there exists a unique @xmath1026 so that @xmath1027 we drop the dependence in @xmath1028 from the notation for a moment . if we set @xmath1029 and @xmath1030 then ( [ eqn : imp_+ ] ) becomes @xmath1031 by the analytic implicit function theorem , for @xmath327 near @xmath571 , there exist a unique analytic function @xmath1032 so that @xmath1033 by inspecting of the first few coefficients in the taylor expansion of @xmath1034 , we find that @xmath1035 thus , since @xmath1036 , and @xmath1037 we find that @xmath1038 where @xmath1039 . choose @xmath703 so that if @xmath1040 , then @xmath1041 . define @xmath1042 . by applying the intermediate value theorem to @xmath1043 , there exists @xmath1044t_n^- , t_n^+[$ ] so that @xmath1045 . see figure [ intermediate ] . since @xmath1046 is increasing to infinity , the sequence @xmath744 is decreasing to zero . . [ intermediate ] ] moreover , since @xmath1047 @xmath1048 where @xmath1049 . from ( [ eqn : r_n_asymp ] ) we have @xmath1050 where @xmath1051 . by substituting this into ( [ eq : tn+_asymp ] ) and ( [ eq : tn_asymp ] ) we find that @xmath1052 and @xmath1053 the first claim follows . moreover , since @xmath1054 , we have @xmath1055 it follows that , for all sufficiently large @xmath727 , we have @xmath1056\subset [ t_n^-,t_n^+]$ ] . since the intervals @xmath1057\}$ ] are disjoint , the claim is proven . in this section we derive the desired contradiction . in particular , we prove the following . [ prop : dt ] if @xmath41 is a cusp form eigenvalue branch with a positive limit , then there exists @xmath705 and @xmath1058 so that if @xmath845 , then @xmath1059 the proof will consist of two types of lower estimates . the first depends on the fact that near each crossing the ` relative variation ' @xmath1060 is at least of order @xmath841 . the second shows that away from the crossings the relative variation is not too negative . define @xmath10610,t]~ \left|~ \left\| w_s^k \right\|~ \leq~ \rho \cdot \|u_s\| \right . \right\}.\ ] ] if @xmath964 , then it follows from proposition [ prop : adotbound ] that there exists @xmath749 so that for @xmath1062 we have @xmath1063 hence , since @xmath1064 , there exists @xmath1065 so that if @xmath1066 and @xmath964 , then @xmath1067 for each @xmath1062 . we will integrate this estimate near the crossings to obtain the following . [ lem : onk ] for each @xmath964 , there exists @xmath705 and @xmath1068 so that for each @xmath518 , we have @xmath1069 by ( [ eq : edotkappa ] ) the integrand is positive on @xmath1070 , it suffices to show the same estimate holds for a subset @xmath1071 of @xmath1070 . to define this subset , we first combine proposition [ prop : xings - kpart ] , proposition [ prop : xings ] , and proposition [ prop : seqxings ] to find @xmath844 , @xmath1072 , and a monotone sequence @xmath1073 so that for each @xmath727 @xmath1074 \ ] ] belongs to @xmath1075 , the intervals @xmath1076 and @xmath1077 are disjoint , and for each @xmath1078 @xmath1079 where @xmath1080 . the subset @xmath1071 will be defined as a union of @xmath1081 over sufficiently large @xmath727 . we have @xmath1082 and hence there exists @xmath1083 so that if @xmath1084 we have @xmath1085 thus , from ( [ t_n_bound ] ) we find that if @xmath1086 , then @xmath1087 since the intervals @xmath1088 are disjoint , by combining ( [ eq : edotkappa ] ) , ( [ eq : intoveri ] ) , and ( [ eq : sumn53 ] ) , we find that @xmath1089 where @xmath1090 and @xmath1091 let @xmath1092 . if @xmath845 , then @xmath1093 $ ] for some @xmath1086 . we have @xmath1094 and hence @xmath1095 and @xmath1096 . therefore , ( [ est : int_g ] ) implies the claim . to bound the relative variation on the complement of @xmath1070 , we will use the following . [ prop : fdotbetweenxings ] there exists @xmath353 and @xmath705 such that , if @xmath518 , then @xmath1097 by arguing as in ( [ est : e_dot_lower ] ) we have @xmath1098 let @xmath1099 denote the orthogonal projection of @xmath52 onto the eigenfunction branch of @xmath35 that corresponds to @xmath1100 from theorem [ tracking ] . let @xmath1101 . we have @xmath1102 , @xmath1103 and @xmath1104 where @xmath1105 is an eigenfunction of @xmath608 with eigenvalue @xmath1100 . using the cauchy - schwarz inequality and the nonnegativity of @xmath672 we have @xmath1106 since @xmath1107 is an eigenfunction , we have @xmath1108 . using ( [ defn : ak ] ) and the fact that @xmath1107 is an eigenfunction that is orthogonal to @xmath1109 , we find that @xmath1110 since @xmath1111 , we have @xmath1112 the large @xmath326 asymptotics of @xmath1113 and @xmath1114 can be analysed using the same methods used in appendix [ appendix : airyb ] for @xmath1115 . in particular , the estimate proven in lemma [ prop : expdecay ] holds with @xmath1116 estimated in lemma [ lem : restimate ] . in particular , for each @xmath1117 , there exists a constant @xmath353 so that for sufficiently small @xmath12 @xmath1118 and @xmath1119 if @xmath1120 , then @xmath1121 for sufficiently small @xmath12 . therefore , by splitting the domain of integration into @xmath1122 $ ] and @xmath1123 and using the cauchy - schwarz inequality , we find that @xmath1124 for sufficiently small @xmath12 . we claim that @xmath1125 . indeed , by applying lemma [ lem : quasimode ] with @xmath1126 , we have @xmath1127 for some constant @xmath353 . thus , since the eigenvalue @xmath1128 satisfies @xmath1129 we find that @xmath1130 by definition , @xmath1109 is a projection onto eigenspaces of @xmath608 whose associated eigenvalues are distinct from @xmath1128 . by lemma [ lem : lambda_asymp ] , there exists @xmath844 so that such eigenvalues differ from @xmath1128 by at least @xmath1131 . because of ( [ est : preresolv ] ) , we can thus apply a resolvent estimate ( e.g. lemma 2.1 @xcite ) to find that @xmath1132 by substituting ( [ est : resolvent_est ] ) into ( [ est : pre_resolvent ] ) and setting @xmath1133 , we find a constant @xmath832 so that @xmath1134 by combining this estimate with ( [ eq : at_ak_inner ] ) , ( [ est : adot ] ) , and ( [ eq : edot ] ) , we obtain a constant @xmath1135 so that @xmath1136 by orthogonality @xmath1137 , and hence by ( [ est : resolvent_est ] ) and lemma [ lem : lambda_asymp ] , we have a constant @xmath1138 so that @xmath1139 the desired result follows . [ coro : nottoonegative ] there exists @xmath832 such that for each @xmath11400,1[$ ] , there exists @xmath705 such that if @xmath1141 , then @xmath1142 \setminus k(t,\rho ) } \left(\dot{e}_s~ -~ \dot{\lambda}^*_s\right)~ ds~ \geq~ c ' \cdot \left(\rho^2- 1 \right ) \cdot t^{{\frac{2}{3}}}. \ ] ] by definition , if @xmath1143\setminus k(t,\rho)$ ] , then @xmath1144 and hence from lemma [ prop : fdotbetweenxings ] , we find that @xmath1145 by using lemma [ lem : lambda_asymp ] , we find @xmath832 and @xmath301 so that for @xmath518 @xmath1146 the claim follows from integration . finally , we use lemma [ lem : onk ] and corollary [ coro : nottoonegative ] to prove theorem [ prop : dt ] . this will complete the proof of the main theorem . apply lemma [ lem : onk ] with @xmath1147 . then apply corollary [ coro : nottoonegative ] with @xmath1148 such that @xmath1149 since @xmath1150 is positive on @xmath1151 , we find that @xmath1152 since @xmath1153 , we have the desired conclusion . in this appendix , we compute the asymptotics of each real - analytic eigenvalue branch of @xmath890 for each @xmath1154 . [ lem : lambda_asymp ] let @xmath1155 and let @xmath907 be a real - analytic eigenvalue branch of @xmath666 for @xmath601 . then @xmath1156 where @xmath1157 and @xmath1158 is a zero of the derivative of the airy function @xmath1159 defined in ( [ eq : airy_asymp ] ) . moreover , @xmath1160 to prove proposition ( [ lem : lambda_asymp ] ) , we will first transform the eigenvalue problem into an eigenvalue problem that is easier to analyse . if @xmath77 is an eigenfunction of @xmath666 with respect to @xmath1161 with eigenvalue @xmath260 , then for each @xmath1162 and @xmath601 , we have @xmath1163 where we have set @xmath1164 hence @xmath1165 by making the change of variable @xmath1166 , letting @xmath1167 and @xmath1168 , and dividing by @xmath1169 , we find that @xmath1170 where @xmath1171 and @xmath1172 this leads us to set @xmath1173 and define for each @xmath1174 the quadratic forms @xmath1175 and @xmath1176 define @xmath1177 and @xmath1178)$ ] ( i.e. the set of functions @xmath1179 such that the distributional derivative also is in @xmath76 ) . then for each @xmath1180 , the form @xmath1181 is a bounded quadratic form on @xmath1182 and @xmath1183 is a closed quadratic form on @xmath1182 with domain @xmath1184 . since @xmath1185 maps bijectively @xmath1186)$ ] onto @xmath1187)$ ] , the function @xmath1188 is an eigenfunction of @xmath1183 with respect to @xmath1181 with eigenvalue @xmath1189 . it follows from the perturbation theory of generalized eigenvalue problems ( see vii.6 in @xcite ) that the eigenvalues of @xmath1183 with respect to @xmath1181 can be organized into real - analytic eigenvalue branches for @xmath1180 . the domains of @xmath1183 and @xmath1181 change , and hence analytic perturbation theory can not be applied . ] since the generalized eigenvalue problem @xmath1190 corresponds to a sturm - liouville problem with neumann condition at @xmath1191 , the eigenspaces are 1-dimensional . hence , we may enumerate the real - analytic eigenvalue branches @xmath1192 so that for each @xmath1193 and @xmath1180 we have @xmath1194 [ lem : bounded_branches ] for each @xmath1195 , there exists @xmath1196 and @xmath353 so that if @xmath979 then @xmath1197 in particular , there exists @xmath811 so that for small @xmath1180 @xmath1198 moreover , @xmath1199 is a zero of the derivative of the airy function @xmath1159 . first , we show that each @xmath1200 is bounded . to this end , define @xmath1201 since @xmath1202 is bounded above by 2 , we have @xmath1203 for each @xmath1180 and @xmath1174 . note that for each @xmath1204 , we have @xmath1205 and hence @xmath1206 integration by parts shows that the eigenfunctions of @xmath1207 with respect to @xmath1208 are solutions to the sturm - liouville problem @xmath1209 standard convexity estimates on solutions to ordinary differential equations imply that each eigenfunction belongs to the domain @xmath1184 of @xmath1183 for each @xmath1180 . in particular , the sum of the first @xmath1195 eigenspaces of @xmath1207 with respect to @xmath1208 belongs to @xmath1184 . therefore , using ( [ est : btoa ] ) , the minimax principle , and ( [ order ] ) we find that @xmath1192 is bounded by the @xmath1210 eigenvalue of @xmath1207 with respect to @xmath1208 . in the remainder of the argument we drop the superscript @xmath1195 and focus on an individual real - analytic eigenfunction branch @xmath1211 with eigenvalue @xmath1212 . for @xmath1180 , we have @xmath1213 where @xmath1214 indicates differentiation with respect to @xmath140 . a computation gives that for each @xmath69 @xmath1215 and @xmath1216 let @xmath1211 be a real - analytic eigenfunction branch of @xmath1183 with respect to @xmath1181 associated to the real - analytic eigenvalue branch @xmath1212 . integration by parts gives @xmath1217 let @xmath1016 be the upper bound on @xmath1212 proven above . if @xmath1204 and @xmath1218 , then @xmath1219 and hence @xmath1220 for @xmath1204 . it follows that @xmath1221 for @xmath1222 . thus , since @xmath1223 is finite , we find that for @xmath1224 @xmath1225 integrating from @xmath1226 to @xmath1227 , we find a constant @xmath353 ( that depends on @xmath1226 ) such that , for @xmath1228 we have @xmath1229 from this we find constants @xmath353 such that that @xmath1230 a similar argument shows that @xmath1231 therefore , ( [ est : nu_dot ] ) holds , and via integration we find @xmath811 so that ( [ eq : nu_asymp ] ) holds true . continuity of solutions to ordinary differential equations with respect to coefficients applies to ( [ eq : a - n - ode ] ) with fixed initial conditions @xmath1232 and @xmath1233 . in particular , we have a solution @xmath1234 to @xmath1235 it follows that @xmath1236 is a solution to @xmath1237 . estimate ( [ a - n - convexity ] ) applies to @xmath1234 , and hence it follows from ( [ eq : airy_asymp ] ) that @xmath77 is a multiple of @xmath1159 . the function @xmath25 satisfies the neumann condition @xmath1238 and hence @xmath1239 as desired . if @xmath1240 is a real - analytic eigenfunction branch of of @xmath666 associated to the eigenvalue branch @xmath894 , then @xmath1241 is a real - analytic eigenfunction branch of @xmath1183 with eigenvalue branch @xmath1242 . lemma [ lem : bounded_branches ] implies that @xmath1243 by differentiating @xmath1244 , we find that @xmath1245 by lemma [ lem : bounded_branches ] , both @xmath1246 and @xmath1212 are bounded . therefore , @xmath1247 and hence @xmath1248 . let @xmath39 be a real - analytic eigenbranch of @xmath29 such that @xmath1249 for some positive integer @xmath40 . for a fixed constant @xmath686 , let @xmath1250.\ ] ] as in [ sec : nota ] , let @xmath202 denote the orthogonal projection of @xmath30 onto the sum of the eigenspaces of @xmath35 whose eigenvalues lie in @xmath282 . [ prop : blower ] let @xmath955 . there exists @xmath749 , @xmath1251 and @xmath705 such that , if @xmath518 and if @xmath1252 is an eigenfunction of @xmath621 with eigenvalue @xmath1253 satisfying @xmath1254 then @xmath1255 the quadratic form @xmath1257 is controlled by @xmath35 meaning that there exists a constant @xmath353 such that for @xmath1258 @xmath1259 thus , proposition [ lem : u_w_close ] and the fact that @xmath1260 imply that observe also that proposition [ lem : u_w_close ] also implies that @xmath1263 in the limit @xmath1264 so that we can freely replace @xmath53 by @xmath1265 and vice - versa in each ( multiplicative ) estimate . by the discussion [ subsec : separation ] and [ sec : nota ] , for each @xmath12 we can uniquely write @xmath1266 where each @xmath1267 is an eigenfunction of @xmath666 with eigenvalue @xmath1268 . set @xmath1269 and @xmath1270 by linearity @xmath1271 from ( [ eq : defb_t ] ) we have @xmath1272 where @xmath1273 $ ] and @xmath1274 is defined in ( [ defn : p ] ) . since @xmath1275 , and @xmath1276 , we find that @xmath1277 if @xmath1278 , then @xmath1279 , and so @xmath1280 . for @xmath1281 , apply lemma [ lem : bwellpsi ] below to find that @xmath1282 since @xmath1283 and @xmath1284 are orthogonal if @xmath1285 , we have @xmath1286 thus , by summing ( [ eq : vlower ] ) over @xmath1287 , we obtain @xmath1288 for @xmath1289 , we have @xmath1290 thus , from lemma [ lem : bwkpsi ] and lemma [ lem : p_positive ] , there exists @xmath704 so that @xmath1291 for some @xmath704 . the latter estimate , combined with ( [ eq : linearity ] ) , ( [ eq : sumvl ] ) , and the triangle inquality , yield the claim . multiply both sides of ( [ eq : rdefn ] ) by a smooth function with compact support @xmath1302 and integrate over @xmath1303 , then integrate by parts to obtain @xmath1304 observe that @xmath1305 so that by applying lemma [ coro : wisqm ] to the test function @xmath1306 , there exists @xmath705 and @xmath832 such that for @xmath518 , we have @xmath1307 recalling that the @xmath76-norm on the right hand side has the weight @xmath1308 this implies that @xmath1309 the claim follows since @xmath1310 on the interval over which we integrate . the strategy of the proof of lemma [ lem : bwellpsi ] is as follows . by ( [ eq : rdefn ] ) , the function @xmath1311 is a solution to the inhomogeneous equation @xmath1312 where @xmath1313 and @xmath1314 the function @xmath1315 is a solution to the homogeneous equation @xmath1316 where @xmath1317 . our choice of @xmath249 in ( [ eq : beta ] ) implies that @xmath1318 is bounded below by a constant @xmath1319 for all small @xmath12 , @xmath889 , and @xmath1320 . hence we can use wkb type estimates to find a basis @xmath1321 of solutions to the homogeneous equation ( [ eq : homoode ] ) . we will then use ` variation of parameters ' to express each solution to to ( [ eq : inhomoode ] ) in terms of this basis , and we use lemma [ lem : restimate ] to provide control of the inhomogeneous term @xmath1116 . finally , we will estimate the integral in ( [ eq : ellsmall ] ) using a riemann - lebesgue type estimate . for @xmath889 and @xmath1320 , we have @xmath1322 , and hence we can apply theorem 6.2.1 in @xcite , to obtain a basis @xmath1323 of solutions to the homogeneous equation ( [ eq : homoode ] ) that satisfy @xmath1324 and @xmath1325 where , for @xmath1320 and @xmath889 , the smooth functions @xmath1326 and @xmath1327 have @xmath1328 norm that is uniformly @xmath1329 . observe that @xmath1330 have @xmath1331)$ ] norms that are uniformly bounded above and away from @xmath1332 moreover , since @xmath1333 and @xmath1334 are highly oscillatory for small @xmath12 , an integration by parts argument shows that the @xmath1331)$ ] inner product @xmath1335 is @xmath1329 . it follows that there exists @xmath1336 such that if @xmath1337 , then @xmath1338 for all sufficiently small @xmath12 . here @xmath1339 denotes the @xmath1331)$ ] norm . in particular , for each @xmath765 and each @xmath12 , there exists @xmath1345 so that the function @xmath38 of ( [ eq : vt ] ) satisfies @xmath1346 the eigenfunction @xmath1252 of @xmath1347 satisfies ( [ eq : generic_inhomo ] ) with @xmath1348 , and hence there exists @xmath1349 so that @xmath1350 the integral in ( [ eq : ellsmall ] ) is equal to @xmath1351 by expanding the product of sums , one obtains a sum of @xmath1352 integrals . by substituting the expressions ( [ eq : wkbderivate ] ) and ( [ eq : wkb ] ) , integration by parts , and applying standard estimates , we find that each integral is @xmath1329 . for example , consider the terms of the form @xmath1353 since @xmath694 , an elementary computation shows that there exists @xmath844 so that if @xmath1354 $ ] and @xmath12 is sufficiently small , then @xmath1355 thus , we may integrate by parts to find a constant @xmath353 so that the integral in ( [ eq : aa ] ) is at most @xmath1356 . it follows that all the terms of this form are bounded above by from ( [ eq : wkb ] ) and ( [ eq : wkbderivate ] ) we find that there exists @xmath1360 so that if @xmath1361 , then @xmath1362 . from ( [ eq : wkb ] ) and ( [ eq : ftl - defn ] ) we find that for each @xmath765 @xmath1363}~ \left|v_{e_t,\mp}^{\ell}(y ) \right|~ \leq~ \frac{2}{\sqrt{\delta_1 } } \ ] ] for all sufficiently small @xmath12 . hence , using the cauchy - schwarz inequality and lemma [ lem : restimate ] we have , for all @xmath1364 $ ] , @xmath1365 for all sufficiently small @xmath12 . for the derivative of @xmath1366 , we have @xmath1367 applying cauchy - schwarz and lemma [ lem : restimate ] gives @xmath1368 finally , we apply integration by parts to ( [ eq : aa ] ) . the resulting terms that do not contain @xmath1369 have uniformly bounded @xmath1328 norm . the term that contains @xmath1369 can be bounded using ( [ eq : boundadotpm ] ) . it follows that all the term of this form are bounded by @xmath1370 the final step consists in bounding @xmath1371 by @xmath1372 to control the terms of eq . ( [ eq : type1 ] ) . using ( [ eq : basis_equivalence ] ) and ( [ eq : defv_var ] ) we have @xmath1373}\{|h_+(y)|+|h_-(y)|\}\ ] ] by orthogonality we have @xmath1374 and , using the bound on @xmath1375 we finally obtain @xmath1376 this finishes the proof . as in the previous subsection , the function @xmath1377 is a solution to the inhomogeneous equation ( [ eq : inhomoode ] ) with @xmath1313 and @xmath1116 defined by ( [ eq : rdefn ] ) . however , for @xmath1289 , the function @xmath1378 is negative for large @xmath326 . in fact , since @xmath41 decreases to @xmath42 , the function @xmath1379 changes sign nearer and nearer to @xmath627 . since the solution @xmath1377 belongs to @xmath1380 , we expect it to decay exponentially as soon as @xmath326 moves away from @xmath16 . for @xmath326 near 1 , we will approximate @xmath38 using airy functions . in this subsection we will make these approximations precise and use them to give a proof of lemma [ lem : bwkpsi ] by lemma [ zerolemma ] , @xmath1252 is a constant multiple of @xmath1381 defined in ( [ zeroeigenvector ] ) . because both sides of the estimate in lemma [ lem : bwkpsi ] are homogeneous functions of degree 1 in @xmath1252 , it suffices to assume that @xmath1382 . we have @xmath1388 where @xmath1389 @xmath1390 and @xmath1391 . in particular , for sufficiently small @xmath12 @xmath1392 thus , since @xmath1393 and @xmath1394 the triangle inequality applied to ( [ eq : psi_0_decomp ] ) implies that ( [ est : psi_0_sup ] ) holds for sufficiently small @xmath12 . by proposition [ tracking ] and lemma [ lem : lambda_asymp ] , there exists @xmath353 so that if @xmath12 is sufficiently small , then @xmath1402 hence , for @xmath1403 , one finds that @xmath1404 thus , since @xmath1405 , there exists @xmath905 such that if @xmath1406 then for all @xmath1407 we have @xmath1408 for each smooth function @xmath1409 with support in @xmath1321+t^{\alpha } , \infty[$ ] , define @xmath1410 extend @xmath1411 to a self - adjoint operator on @xmath1412 . it follows from ( [ eq : lowerbound ] ) that the spectrum of @xmath1411 lies in @xmath1413 . hence @xmath1411 is invertible and the operator norm of @xmath1414 is bounded above by @xmath1415 therefore @xmath1416 the function @xmath1417 is a solution to ( [ eq : inhomoode ] ) on @xmath1418 , and hence @xmath1419 is a solution to the homogeneous equation ( [ eq : homoode ] ) . it follows from ( [ eq : lowerbound ] ) that @xmath1420 if @xmath1421 . in particular , @xmath1422 is convex , moreover @xmath1422 is non - negative and in @xmath1423 since @xmath1424 and @xmath1425 this implies that @xmath1426 . indeed , since @xmath1422 is convex , @xmath1427 has a limit @xmath1428 in @xmath1429 if this limit is positive then it implies that @xmath1430 for large @xmath326 and this contradicts the fact that @xmath1431 is integrable . in particular , @xmath1427 is bounded so that by integrating ( [ est : w2convex ] ) we find that @xmath1432 the argument also shows that @xmath1427 is non positive for large @xmath326 so that @xmath1422 has a limit when @xmath1433 since @xmath1422 is integrable , this limit is @xmath1332 for each @xmath1434 the function @xmath1435 that is defined by @xmath1436 satisfies @xmath1437 with @xmath1438 and @xmath1439 . therefore , by comparison with ( [ est : w2convex ] ) , and using the maximum principle , we find that if @xmath1440 , then @xmath1441 . applying this to @xmath1442 we find that for each @xmath1443 @xmath1444 for small @xmath12 , the function @xmath1455 has a simple zero near @xmath627 . thus , to approximate solutions of @xmath1456 near @xmath627 we will use solutions to airy s differential equation @xmath1457 where @xmath1458 . we first describe the link to airy s equation . if we define @xmath1459 , then we have @xmath1460 where @xmath1461 . let @xmath532 be the smooth function that satisfies @xmath1462 by substituting the latter expression into ( [ eq : ode2 ] ) and by dividing by @xmath1463 , we find that @xmath1464 setting @xmath1465 we have @xmath1466 where @xmath1467 we have @xmath1472 , and hence by lemma [ lem : restimate ] , @xmath1473 let @xmath1474 be the tracking eigenvalue branch associated to @xmath1475 . the eigenvalue @xmath49 corresponding to @xmath1476 is in a @xmath1329 neighbourhood of @xmath1477 moreover , when @xmath12 tends to @xmath571 , proposition ( [ lem : lambda_asymp ] ) implies that @xmath894 is at a distance of order @xmath896 of the rest of the spectrum . using ( [ eq : lambda_asymp ] ) , ( [ eq : rdefn ] ) , lemma [ lem : restimate ] and a resolvent estimate , we have @xmath1478 , and hence @xmath1479 the tracking eigenfunction @xmath1476 satisfies ( [ eq : inhomoode ] ) with @xmath1348 . hence by proposition [ prop : expdecay ] , if @xmath1480 , then @xmath1481 observe that , by orthogonality , @xmath1482 , hence @xmath1483 since @xmath1484 ( because @xmath1470 ) , we may thus take @xmath1485 and the biggest term is then of order @xmath1486 . the claim follows . we first construct solutions to the associated homogeneous equation @xmath1489 using airy functions . in particular , it is well - known that there exists a basis @xmath1490 of solutions to @xmath1491 such that @xmath1492 and @xmath1493 one checks that @xmath1494 defines a basis of solutions to ( [ eq : homoairy ] ) . it follows from well - known identities that the wronskian of @xmath1495 is @xmath65 . hence the wronskian of @xmath1496 is @xmath1497 . therefore , by the method of variation of constants , for each @xmath1498 , the function @xmath1499 is a solution to ( [ eq : inhomoairy ] ) . [ lem : estwp ] for each compact set @xmath1500 , there exists @xmath353 such that , for each @xmath1498 , if @xmath1501 and @xmath1502 $ ] , then @xmath1503 and @xmath1504 where @xmath1505 denotes the @xmath76 norm of @xmath1506 over @xmath1507 $ ] . moreover , there exists @xmath1016 so that if @xmath1508$],then @xmath1509 and @xmath1510 using the cauchy - schwarz inequality , we have for @xmath1511 $ ] @xmath1512 and @xmath1513 thus , from ( [ eq : wparticular ] ) and the triangle inequality , we have @xmath1514 estimate ( [ est : wparticular ] ) then follows from lemma [ lem : wdt ] below . the proof is a straightforward consequence of the continuity and known asymptotics of @xmath1525 and @xmath1526 . from ( [ eq : airy_asymp ] ) and integration by parts we find that @xmath1527 as @xmath25 tends to @xmath1528 . thus there exists @xmath1529 so that if @xmath1530 , then @xmath1531 therefore , for @xmath1530 , @xmath1532 and , using ( [ eq : airy_deriv_asymp ] ) , @xmath1533 the expressions on the left hand sides of ( [ est : aplusaminus ] ) and ( [ est : aplusaminusderiv ] ) are continuous in @xmath25 , and hence are bounded by a constant @xmath353 for @xmath1534 where @xmath1535 . moreover , from ( [ est : aplusaminus ] ) and ( [ est : aplusaminusderiv ] ) we have @xmath1541 and @xmath1542 provided @xmath1543 . let @xmath1544 . if @xmath1545 , then @xmath1546 and @xmath1547 . the desired estimates in the @xmath1548/@xmath1549 case follow . the estimates in the @xmath1549/@xmath1548 case are proved in a similar fashion . the function @xmath1555 satisfies the inhomogeneous equation ( [ eq : inhomoairy ] ) with @xmath1556 , and the inhomogeneity @xmath1469 satisfies ( [ eq : boundaw ] ) . to estimate @xmath1555 and hence ( [ eq : small_x_integral ] ) we write @xmath1557 where @xmath1558 is the particular solution to ( [ eq : inhomoairy ] ) defined by ( [ eq : wparticular ] ) for @xmath1559 . the function @xmath1560 is a solution to the associated homogeneous equation . integration by parts gives @xmath1562 using lemmas [ lem : airyapprox ] and [ lem : estwp ] , one finds that there exists @xmath353 such that , for @xmath1563 $ ] @xmath1564 thus , using ( [ lem_psi_0_bound ] ) , we conclude that the first term on the right side of ( [ eq : wpt_integral ] ) is @xmath1565 . to bound the second term on the right side of ( [ eq : wpt_integral ] ) , we take @xmath1566 and estimate the integrals over @xmath1567 $ ] and @xmath1568 $ ] separately . observe that since @xmath1569 then @xmath1570 using lemmas [ lem : airyapprox ] and [ lem : estwp ] , we find @xmath353 so that for all sufficiently small @xmath12 @xmath1571 and @xmath1572 by combining these estimates and using ( [ eq : psi_zero_deriv ] ) , we find that @xmath1573 the claim will follow provided we can choose @xmath1574 and @xmath1575 so that each power of @xmath12 appearing on the righthand side is greater than @xmath1576 . this gives us a solution set which is the open triangle in @xmath1577 bounded by the lines @xmath1578 , @xmath1579 , and @xmath1580 . the two latter lines intersect for @xmath1581 the claim follows . as above we consider @xmath1586 and take some @xmath1587 using lemmas [ lem : airyapprox ] and [ lem : estwp ] , one finds that @xmath1588 } & \leq c \cdot t^{\frac{5{\alpha}}{2}-\frac{3{\widetilde{\alpha}}}{4}-\frac{1}{6 } } \|u_t\| \\ \| w_{p , t}\|_{[t^{{\widetilde{\alpha}}},3t^{\alpha } ] } & \leq c \cdot t^{2{\alpha}-\frac{{\widetilde{\alpha}}}{2}-\frac{2}{3 } } \|u_t\| . \end{split}\ ] ] the claim will follow provided we can find @xmath1589 and @xmath1590 such that @xmath1591 and @xmath1592 the solution set is here a quadrilateral whose projection on the @xmath355-axis is the interval @xmath132\frac{7}{33},{\frac{1}{3}}[.$ ] finally , we consider the integral corresponding to the homogeneous part @xmath1560 of @xmath1555 : @xmath1593 there exist constants @xmath1594 , @xmath1595 , depending on @xmath12 , such that @xmath1596 where @xmath1597 and @xmath1598 are as defined in ( [ eq : defn_w ] ) with the parameter @xmath140 and @xmath1599 defined in ( [ eq : s_z_s ] ) . [ lem : norms ] for all @xmath1400\frac{7}{33 } , \frac{1}{3}[,$ ] there exists @xmath1584 such that @xmath1602}~ = ~ o(t^\infty ) \cdot \| u_t\| \ ] ] where @xmath1603 is a function that is of order @xmath1604 for each @xmath727 , and @xmath1605}~ \geq~ { \frac{1}{2}}\cdot \left ( \|w_t^k\| -c\cdot t^\delta \|u_t\|\right ) . \ ] ] using the behavior of the norm of @xmath1606 we find that @xmath1607}~ = ~ o(t^\infty ) \cdot \|a_+w_+\|_{[2t^{\alpha},3t^{\alpha } ] } \ ] ] and @xmath1608}~ \leq~ c \cdot \|a_- w_-\|_{[0,2t^{\alpha}]}. \ ] ] we thus have @xmath1609 } & = & o(t^\infty ) \cdot \|a_+w_+\|_{[2t^{\alpha},3t^{\alpha}]}\\ & \leq & o(t^\infty ) \cdot \left ( \|w\|_{[2t^{\alpha},3t^{\alpha}]}\,+\ , \|a_- w_-\|_{[2t^{\alpha},3t^{\alpha}]}\,+\,\|w_{p , t}\|_{[2t^{\alpha},3t^{\alpha}]}\right ) \\ & \leq & o(t^\infty ) \cdot \left ( \|u_t\| \,+\,\|w_-\|_{[0,2t^{\alpha } ] } \,+\ , t^\delta \|u_t\|\right ) \\ & \leq & o(t^\infty ) \cdot \left ( \|u_t\| \,+\ , \|w\|_{[0,2t^{\alpha}]}\,+\ , \|a_+w_+\|_{[0,2t^{\alpha}]}\,+\,\|w_{p , t}\|_{[0,2t^{\alpha}]}\right ) \\ & \leq & o(t^\infty)\left ( \|a_+w_+\|_{[0,2t^{\alpha}]}\,+\,\|u_t\|\right ) . \ ] ] estimate ( [ est : a+w+ ] ) then follows by absorbing the norm of @xmath1610 into the left hand side . to prove estimate ( [ est : a - w- ] ) , we first observe that by using the triangle inequality we find that @xmath1611 } \leq \|w_{p , t}\|_{[0,2t^{\alpha } ] } \,+\,\|a_- w_-\|_{[0,2t^{\alpha}]}\,+\,\|a_+w_+\|_{[0,2t^{\alpha}]}.\ ] ] the first term on the righta hand side is @xmath1612 and the last one is @xmath1613 so that we obtain @xmath1614 } \geq \left ( \|w\|_{[0,2t^{\alpha } ] } - o(t^\delta ) \|u_t\|\right ) .\ ] ] the claim then follows by observing that corollary [ coro : normlocal ] implies that @xmath1615 } \geq { \frac{1}{2}}\left ( \|w_t^k\| - o(t^{1-{\alpha } } ) \|u_t\|\right).\ ] ] from ( [ eq : psi_prime ] ) , we have @xmath1617 thus , the integral we want to estimate can be written as @xmath1618 where we have set @xmath1619 denote by @xmath1620 the integral @xmath1621 integration by parts shows that @xmath1622 where we have set @xmath1623 since @xmath1624 and @xmath140 is of order @xmath12 , and since @xmath1159 is rapidly decreasing , the boundary term at @xmath1625 is @xmath1626 observe that we have a global @xmath1627 prefactor in front of the integral term . thus , when the @xmath1628 is applied to @xmath1629 , we gain @xmath1630 , that is , something of order @xmath1631 when @xmath1628 hits the airy function , we lose a @xmath1632 so that the global prefactor is of order @xmath1633 which is @xmath1634 . summarizing , integrating by parts gains at least a prefactor @xmath1635 by repeated integration by parts we thus observe that we can write , for each @xmath500 @xmath1636 where the @xmath1637 are some constants and the remainder term @xmath1638 can be written @xmath1639 for some smooth functions @xmath1640 if we fix some order @xmath1641 then , using that @xmath1159 and all its derivatives are rapidly decreasing we can find @xmath500 such that the remainder @xmath1638 is @xmath1642 this tells us that @xmath1620 admits a complete asymptotic expansion of the form @xmath1643 from the first integration by parts we see that @xmath1644 and the second term is then of order @xmath1635 the claim follows by taking the imaginary part . let @xmath1158 be the zero of @xmath1159 that comes from lemma [ lem : nonzero ] . since @xmath1159 is a nontrivial solution to a second order differential equation , @xmath1159 can not vanish at a zero of the derivative @xmath1649 . hence , for sufficiently for small @xmath12 , we have @xmath1652 . by arguing as in the proof of lemmas [ lem : nonzero ] and [ lem : wdt ] and using @xmath1653 , we find @xmath1654 so that @xmath1655 where @xmath1656 and @xmath12 is sufficiently small . in particular , @xmath1657 hence the claim follows from lemma [ w- ] . c. m. judge . _ small eigenvalues and maximal laminations on complete surfaces of negative curvature . _ in the tradition of ahlfors - bers . iv , 93 - 105 , contemp . , 432 , amer . , providence , ri , 2007 .

interference between wireless links has been taken into account as a critical problem in wireless communication systems . recently , interference alignment ( ia ) was proposed for fundamentally solving the interference problem when there are two communication pairs @xcite . it was shown in @xcite that the ia scheme can achieve the optimal degrees of freedom ( dof ) , which is equal to @xmath4 , in the @xmath1-user interference channel with time - varying channel coefficients . since then , interference management schemes based on ia have been further developed and analyzed in various wireless network environments : multiple - input multiple - output ( mimo ) interference networks @xcite , x networks @xcite , and cellular networks @xcite . on the one hand , following up on these successes for single - hop networks , more recent and emerging work has studied multihop networks with multiple source - destination ( s d ) pairs . for the 2-user 2-hop network with 2 relays ( referred to as the @xmath5 interference channel ) , it was shown in @xcite that interference neutralization combining with symbol extension achieves the optimal dof . a more challenging network model is to consider @xmath1-user two - hop relay - aided interference channels , consisting of @xmath1 source - destination ( s d ) pairs and @xmath2 helping relay nodes located in the path between s d pairs , so - called the @xmath3 channel . several achievability schemes have been known for the network , but more detailed understanding is still in progress . by applying the result from @xcite to the @xmath0 channel , one can show that @xmath4 dof is achieved by using orthogonalize - and - forward relaying , which completely neutralizes interference at all destinations if @xmath2 is greater than or equal to @xmath6 . another achievable scheme , called aligned network diagonalization , was introduced in @xcite and was shown to achieve the optimal dof in the @xmath0 channel while tightening the required number of relays . the scheme in @xcite is based on the real interference alignment framework @xcite . in @xcite , however , the system model under consideration assumes that there is no interfering signal between relays and the relays are full - duplex . moreover , in @xcite , the @xmath5 interference channel with full - duplex relays interfering with each other was characterized and its dof achievability was shown using aligned interference neutralization . interference channel @xcite . ] on the other hand , there are lots of results on the usefulness of fading in the literature , where one can obtain the multiuser diversity gain in broadcast channels : opportunistic scheduling @xcite , opportunistic beamforming @xcite , and random beamforming @xcite . such opportunism can also be fully utilized in multi - cell uplink or downlink networks by using an opportunistic interference alignment strategy @xcite . various scenarios exploiting the multiuser diversity gain have been studied in cooperative networks by applying an opportunistic two - hop relaying protocol @xcite and an opportunistic routing @xcite , and in cognitive radio networks with opportunistic scheduling @xcite . in addition , recent results @xcite have shown how to utilize the opportunistic gain when there are a large number of channel realizations . more specifically , to amplify signals and cancel interference , the idea of opportunistically pairing complementary channel instances has been studied in interference networks @xcite and multi - hop relay networks @xcite . in cognitive radio environments @xcite , opportunistic spectrum sharing was introduced by allowing secondary users to share the radio spectrum originally allocated to primary users via transmit adaptation in space , time , or frequency . in this paper , we study the _ @xmath0 channel with interfering relays _ , which can be taken into account as one of practical multi - source interfering relay networks and be regarded as a fundamentally different channel model from the conventional @xmath0 channel in @xcite . then , we introduce an _ opportunistic network decoupling ( ond ) _ protocol that achieves full dof with comparatively easy implementation under the channel model . this work focuses on the @xmath0 channel with one additional assumption that @xmath2 _ half - duplex _ ( hd ) relays interfere with each other , which is a more feasible scenario . the scheme adopts the notion of the multiuser diversity gain for performing interference management over two hops . more precisely , in our scheme , a scheduling strategy is presented in time - division duplexing ( tdd ) two - hop environments with time - invariant channel coefficients , where a subset of relays is opportunistically selected in terms of producing the minimum total interference level . to improve the spectral efficiency , the _ alternate relaying _ protocol in @xcite is employed with a modification , which eventually enables our system to operate in _ virtual full - duplex _ mode . as our main result , it turns out that in a high signal - to - noise ratio ( snr ) regime , the ond protocol asymptotically achieves the min - cut upper bound of @xmath1 dof even in the presence of inter - relay interference and half - duplex assumption , provided the number of relays , @xmath2 , scales faster than @xmath7 , which is the minimum number of relays required to guarantee our achievability result . numerical evaluation also indicates that the ond protocol has higher sum - rates than those of other relaying methods under realistic network conditions ( e.g. , finite @xmath2 and snr ) since the inter - relay interference is significantly reduced owing to the opportunistic gain . for comparison , the ond scheme without alternate relaying and the max - min snr scheme are also shown as baseline schemes . note that our protocol basically operates with local channel state information ( csi ) at the transmitter and thus is suitable for distributed / decentralized networks . our main contributions are fourfold as follows : * in the @xmath0 channel with interfering relays , we introduce a new achievability scheme , termed ond with virtual full - duplex operation . * under the channel model , we completely analyze the optimal dof , the required relay scaling condition , and the decaying rate of the interference level , where the ond scheme is shown to approach the min - cut upper bound on the dof . * our achievability result ( i.e. , the derived dof and relay scaling law ) is validated via numerical evaluation . * we perform extensive computer simulations with other baseline schemes . the rest of this paper is organized as follows . in section [ sec:2 ] , we describe the system and channel models . in section [ sec:3 ] , the proposed ond scheme is specified and its lower bound on the dof is analyzed . section [ sec:4 ] shows an upper bound on the dof . numerical results of the proposed ond scheme are provided in section [ sec:5 ] . finally , we summarize the paper with some concluding remarks in section [ sec:6 ] . throughout this paper , @xmath8 , @xmath9 $ ] , and @xmath10 indicate the field of complex numbers , the statistical expectation , and the ceiling operation , respectively . unless otherwise stated , all logarithms are assumed to be to the base 2 . moreover , table [ tab : notations ] summarizes the notations used throughout this paper . some notations will be more precisely defined in the following sections , where we introduce our channel model and achievability results . .summary of notations [ cols="^,^ " , ] as one of two - hop cooperative scenarios , we consider the @xmath11 channel model with interfering relays , which fits into the case where each s d pair is geographically far apart and/or experiences strong shadowing ( thus requiring the response to a huge challenge for achieving the target spectral efficiency ) . in the channel model , it is thus assumed that each source transmits its own message to the corresponding destination only through one of @xmath2 relays , and thus there is no direct path between an s d pair . note that unlike the conventional @xmath12 channel , relay nodes are assumed to interfere with each other in our model . there are @xmath1 s d pairs , where each receiver is the destination of exactly one source node and is interested only in traffic demands of the source . as in the typical cooperative relaying setup , @xmath2 relay nodes are located in the path between s d pairs so as to help to reduce path - loss attenuations . suppose that each node is equipped with a single transmit antenna . each relay node is assumed to operate in half - duplex mode and to fully decode , re - encode , and retransmit the source message i.e. , decode - and - forward protocol is taken into account . we assume that each node ( either a source or a relay ) has an average transmit power constraint @xmath13 . unlike the work in @xcite , @xmath2 relays are assumed to interfere with each other . channel can also be applied here . ] to improve the spectral efficiency , the alternate relaying protocol in @xcite is employed with a modification . with alternate relaying , each selected relay node toggles between the transmit and listen modes for alternate time slots of message transmission of the sources . if @xmath2 is sufficiently large , then it is possible to exploit the channel randomness for each hop and thus to obtain the opportunistic gain in multiuser environments . in this work , we do not assume the use of any sophisticated multiuser detection schemes at each receiver ( either a relay or a destination node ) , thereby resulting in an easier implementation . now , let us turn to channel modeling . let @xmath14 , @xmath15 , and @xmath16 denote the @xmath17th source , the corresponding @xmath17th destination , and the @xmath18th relay node , respectively , where @xmath19 and @xmath20 . the terms @xmath21 denote the channel coefficients from @xmath14 to @xmath16 and from @xmath16 to @xmath15 , corresponding to the first and second hops , respectively . the term @xmath22 indicates the channel coefficient between two relays @xmath16 and @xmath23 . all the channels are assumed to be rayleigh , having zero - mean and unit variance , and to be independent across different @xmath18 , @xmath17 , @xmath24 , and hop index @xmath25 . we assume the block - fading model , i.e. , the channels are constant during one block ( e.g. , frame ) , consisting of one scheduling time slot and @xmath26 data transmission time slots , and changes to a new independent value for every block . channel with interfering relays . , scaledwidth=57.0% ] in this section , we describe the ond protocol , operating in virtual full - duplex mode , in the @xmath0 channel with interfering relays . then , its performance is analyzed in terms of achievable dof along with a certain relay scaling condition . the decaying rate of the interference level is also analyzed . in addition , the ond protocol with no alternate relaying and its achievability result are shown for comparison . in this subsection , we introduce an ond protocol as the achievable scheme to guarantee the optimal dof of the @xmath12 channel with inter - relay interference , where @xmath27 relay nodes among @xmath2 candidates are opportunistically selected for data forwarding in the sense of producing a sufficiently small amount of interference level . the proposed scheme is basically performed by utilizing the channel reciprocity of tdd systems . suppose that @xmath28 and @xmath29 denote the indices of two relays communicating with the @xmath17th s d pair for @xmath19 . in this case , without loss of generality , assuming that the number of data transmission time slots , @xmath26 , is an odd number , the specific steps of each node during one block are described as follows : * time slot 1 : sources @xmath30 transmit their first encoded symbols @xmath31 , where @xmath32 represents the @xmath33th transmit symbol of the @xmath17th source node . a set of @xmath1 selected relay nodes , @xmath34 , operating in receive mode at each odd time slot , listens to @xmath31 ( note that a relay selection strategy will be specified later ) . other @xmath35 relay nodes and destinations @xmath36 remain idle . * time slot 2 : the @xmath1 sources transmit their encoded symbols @xmath37 . the @xmath1 selected relays in the set @xmath38 forward their first re - encoded symbols @xmath39 to the corresponding @xmath1 destinations . if the relays in @xmath38 successfully decode the corresponding symbols , then @xmath40 is the same as @xmath41 . another set of @xmath1 selected relay nodes , @xmath42 , operating in receive mode at each even time slot , listens to and decodes @xmath37 while being interfered with by @xmath43 . the @xmath1 destinations receive and decode @xmath39 from @xmath44 @xmath45 . the remaining @xmath46 relays keep idle . * time slot 3 : the @xmath1 sources transmit their encoded symbols @xmath47 . the @xmath1 relays @xmath48 forward their re - encoded symbols @xmath49 to the corresponding @xmath1 destinations . another @xmath1 relays in @xmath38 receive and decode @xmath50 @xmath51 while being interfered with by @xmath52 . the @xmath1 destinations receive and decode @xmath49 from @xmath53 @xmath54 . the remaining @xmath46 relays keep idle . * the processes in time slots 2 and 3 are repeated to the @xmath55th time slot . * time slot @xmath26 : the @xmath1 relays in @xmath56 forward their re - encoded symbols @xmath57 @xmath58 to the corresponding @xmath1 destinations . the @xmath1 sources and the other @xmath35 relays remain idle . at each odd time slot @xmath33 ( i.e. , @xmath59 ) , let us consider the received signal at each selected relay for the first hop and the received signal at each destination for the second hop , respectively . for the first hop ( phase 1 ) , the received signal @xmath60 at @xmath61 is given by @xmath62 where @xmath32 and @xmath63 are the @xmath33th transmit symbol of @xmath14 and the @xmath64th transmit symbol of @xmath65 , respectively . as addressed earlier , if relay @xmath66 successfully decodes the received symbol , then it follows that @xmath67 . the received signal @xmath68 at @xmath61 is corrupted by the independent and identically distributed ( i.i.d . ) and circularly symmetric complex additive white gaussian noise ( awgn ) @xmath69 having zero - mean and variance @xmath70 . note that the second term in the right - hand side ( rhs ) of ( [ eq : y1 ] ) indicates the inter - relay interference , which occurs when the @xmath1 relays in the set @xmath38 , operating in receive mode , listen to the sources , the relays are interfered with by the other set @xmath56 , operating in transmit mode . note that when @xmath71 , relays have no symbols to transmit , and the second term in the rhs of ( [ eq : y1 ] ) becomes zero . similarly when @xmath72 , sources do not transmit symbols , and the first term in the rhs of ( [ eq : y1 ] ) becomes zero . for the second hop ( phase 2 ) , assuming that the @xmath1 selected relay nodes transmit their data packets simultaneously , the received signal @xmath73 at @xmath15 is given by @xmath74 where @xmath75 is the i.i.d . awgn having zero - mean and variance @xmath70 . we also note that when @xmath71 , there are no signals from relays . likewise , at each even time slot ( i.e. , @xmath76 ) , the received signals at @xmath77 and @xmath15 ( i.e. , the first and second hops ) are given by @xmath78 and @xmath79 respectively . the illustration of the aforementioned ond protocol is geographically shown in fig . [ fig : dig ] ( two terms @xmath80 and @xmath81 are specified later in the following relay selection steps ) . now , let us describe how to choose two types of relay sets , @xmath38 and @xmath56 , among @xmath2 relay nodes , where @xmath2 is sufficiently large ( the minimum @xmath2 required to guarantee the dof optimality will be analyzed in section [ subsec : ana ] ) . let us first focus on selecting the set @xmath82 , operating in receive and transmit modes in odd and even time slots , respectively . for every scheduling period , it is possible for relay @xmath16 to obtain all the channel coefficients @xmath83 and @xmath84 by using a pilot signaling sent from all of the source and destination nodes due to the channel reciprocity before data transmission , where @xmath20 and @xmath19 ( note that this is our local csi assumption ) . when @xmath16 is assumed to serve the @xmath17th s d pair @xmath85 , it then examines both i ) how much interference is received from the other sources and ii ) how much interference is generated by itself to the other destinations , by computing the following scheduling metric @xmath86 : @xmath87 where @xmath88 and @xmath89 . we remark that the first term @xmath90 in ( [ eq : il ] ) denotes the sum of interference power received at @xmath16 for the first hop ( i.e. , phase 1 ) . on the other hand , the second term @xmath91 indicates the sum of interference power generating at @xmath16 , which can be interpreted as the _ leakage of interference _ to the @xmath92 receivers expect for the corresponding destination , for the second hop ( i.e. , phase 2 ) under the same assumption . suppose that a short duration cts ( clear to send ) message is transmitted by the destination who finds its desired relay node ( or the master destination ) . then according to the computed metrics @xmath86 in ( [ eq : il ] ) , a timer - based method can be used for the relay selection similarly as in @xcite . times that of the single s - d pair case @xcite . ] note that the method based on the timer is considerably suitable in distributed systems in the sense that information exchange among all the relay nodes can be minimized . at the beginning of every scheduling period , the relay @xmath16 computes the set of @xmath1 scheduling metrics , @xmath93 , and then starts its own timer with @xmath1 initial values , which can be set to be proportional to the @xmath1 metrics . thus , there exist @xmath94 metrics over the whole relay nodes , and we need to compare them so as to determine who will be selected . the timer of the relay @xmath95 with the least one @xmath96 among @xmath94 metrics will expire first , where @xmath97 and @xmath98 . the relay then transmits a short duration rts message , signaling its presence , to the other @xmath99 relays , where each rts message is composed of @xmath100 bits to indicate which s d pair the relay wants to serve . thereafter , the relay @xmath95 is first selected to forward the @xmath101th s d pair s packet . all the other relays are in listen mode while waiting for their timer to be set to zero ( i.e. , to expire ) . at the stage of deciding who will send the second rts message , it is assumed that the other relays are not allowed to communicate with the @xmath101th s d pair , and thus the associated metrics @xmath102 are discarded with respect to timer operation . if another relay has an opportunity to send the second rts message of @xmath103 bits in order to declare its presence , then it is selected to communicate with the corresponding s when such @xmath1 rts messages , consisting of at most @xmath104 bits , are sent out in consecutive order , i.e. , the set of @xmath1 relays , @xmath82 , is chosen , the timer - based algorithm for the first relay set selection terminates , yielding no rts collision with high probability . we remark that when @xmath105 ( i.e. , the single s d pair case ) , @xmath1 relay nodes are _ arbitrarily _ chosen as the first relay set @xmath38 since there is no interference in this step . now let us turn to choosing the set of @xmath1 relay nodes ( among @xmath35 candidates ) , @xmath106 @xmath107 , operating in receive and transmit modes in even and odd time slots , respectively . using @xmath1 rts messages broadcasted from the @xmath1 relay nodes in the set @xmath38 , it is possible for relay node @xmath108 to compute the sum of inter - relay interference power generated from the relays in @xmath38 , denoted by @xmath109 . when @xmath16 is again assumed to serve the @xmath17th s d pair @xmath85 , it examines both i ) how much interference is received from the undesired sources and the selected relays in the set @xmath38 for the first hop and ii ) how much interference is generated by itself to the other destinations by computing the following metric @xmath110 , termed _ total interference level ( til ) _ : @xmath111 where @xmath88 and @xmath89 . we note that steps 1 and 2 can not be exchangeable due to the fact that the inter - relay interference term @xmath112 is measured after determining the first relay set @xmath38 . if the relay set selection order is switched , then the metric til in ( [ eq : til ] ) will not be available . according to the computed til @xmath110 , we also apply the timer - based method used in step 1 for the second relay set selection . the relay @xmath113 computes the set of @xmath1 tils , @xmath114 , and then starts its timer with @xmath1 initial values , proportional to the @xmath1 tils . thus , we need to compare @xmath115 til metrics over the relay nodes in the set @xmath116 in order to determine who will be selected as the second relay set . the rest of the relay set selection protocol ( i.e. , rts message exchange among relay nodes ) almost follows the same line as that of step 1 . the timer - based algorithm for the second relay set selection terminates when @xmath1 rts messages are sent out in consecutive order . then , @xmath1 relay nodes having a sufficiently small amount of til @xmath110 are selected as the second relay set @xmath56 . owing to the channel reciprocity of tdd systems , the sum of inter - relay interference power received at any relay @xmath117 , @xmath118 , also turns out to be sufficiently small when @xmath2 is large . that is , it is also guaranteed that @xmath1 selected relays in the set @xmath38 have a sufficiently small amount of til . the overhead of each scheduling time slot ( i.e. , the total number of bits required for exchanging rts messages among the relay nodes ) can be made arbitrarily small , compared to one transmission block . from the fact that @xmath1 rts messages , consisting of at most @xmath104 bits , are sent out in each relay set selection step , only @xmath119 bit transmission could suffice . the @xmath27 selected relays request data transmission to their desired source nodes . each source ( @xmath14 ) then starts to transmit data to the corresponding destination ( @xmath15 ) via one of its two relay nodes alternately ( @xmath120 or @xmath66 ) , which was specified earlier . if the tils of the selected relays are arbitrarily small , then i ) the associated undesired source relay and relay destination channel links and ii ) the inter - relay channel links are all in deep fade . in section [ subsec : ana ] , we will show that it is possible to choose such relays with the help of the multiuser diversity gain . at the receiver side , each relay or destination detects the signal sent from its desired transmitter , while simply treating interference as gaussian noise . thus , no multiuser detection is performed at each receiver , thereby resulting in an easier implementation . in this subsection , using the scaling argument bridging between the number of relays , @xmath2 , and the received snr ( refer to @xcite for the details ) , we shall show 1 ) the lower bound on the dof of the @xmath121 channel with interfering relays as @xmath2 increases and 2 ) the minimum @xmath2 required to guarantee the achievability result . the total number of dof , denoted by @xmath122 , is defined as @xcite @xmath123 where @xmath124 denotes the transmission rate of source @xmath14 . using the ond framework in the @xmath125 channel with interfering relays where @xmath26 transmission slots per block are used , the achievable @xmath122 is lower - bounded by @xmath126 where @xmath127 denotes the received signal - to - interference - and - noise ratio ( sinr ) at the relay @xmath128 and @xmath129 denotes the received sinr at the destination @xmath15 when the relay @xmath128 transmits the desired signal ( @xmath130 and @xmath19 ) . more specifically , the above sinrs can be formally expressed as , the third term in the denominator of @xmath131 ( i.e. , the inter - relay interference term ) becomes zero . ] @xmath132 where the second term in the denominator of @xmath133 indicates the interference power at relay @xmath134 received from the sources while the third term indicates the inter - relay interference , and the second term in the denominator of @xmath135 indicates the interference power at the destination @xmath136 received from the active relays . here , @xmath137 , i.e. , @xmath138 if @xmath139 , and vice versa . we focus on the first relay set @xmath140 s perspective to examine the received sinr values according to each time slot . let us first denote @xmath141 for @xmath142 . for the first hop , at time slot @xmath143 ( i.e. , each odd time slot ) , @xmath144 , the received @xmath131 at @xmath145 is lower - bounded by @xmath146 where @xmath147 indicates the scheduling metric in ( [ eq : il ] ) when @xmath61 is assumed to serve the @xmath18th s d pair ( @xmath148 , @xmath136 ) . for the second hop , at time slot @xmath149 ( i.e. , each even time slot ) , @xmath150 , the received @xmath151 at @xmath152 is lower - bounded by @xmath153 where the second inequality holds due to the channel reciprocity . the term @xmath154 in the denominator of ( [ eq : sinr_1_pi1 ] ) and ( [ eq : sinr_2_pi1 ] ) needs to scale as @xmath155 , i.e. , @xmath156 , so that both @xmath157 and @xmath158 scale as @xmath159 with increasing snr , which eventually enables to achieve the dof of @xmath160 per s d pair from ( [ eq : dof_ond ] ) . means that there exist constant @xmath161 and @xmath162 such that @xmath163 for all @xmath164 . ii ) @xmath165 if @xmath166 . iii ) @xmath167 means that @xmath168 @xcite . ] even if such a bounding technique in ( [ eq : sinr_1_pi1 ] ) and ( [ eq : sinr_2_pi1 ] ) leads to a loose lower bound on the sinr , it is sufficient to prove our achievability result in terms of dof and relay scaling law . now , let us turn to the second relay set @xmath169 . similarly as in ( [ eq : sinr_1_pi1 ] ) , for the first hop , at time slot @xmath149 , @xmath170 , the received @xmath171 at @xmath172 is lower - bounded by @xmath173 where @xmath174 indicates the til in when @xmath77 is assumed to serve the @xmath18th s d pair ( @xmath148 , @xmath136 ) . for the second hop , at time slot @xmath175 , @xmath144 , the received @xmath176 at @xmath152 can also be lower - bounded by @xmath177 the next step is thus to characterize the three metrics @xmath86 , @xmath178 , and @xmath110 ( @xmath179 and @xmath180 ) and their cumulative density functions ( cdfs ) in the @xmath0 channel with interfering relays , which is used to analyze the lower bound on the dof and the required relay scaling law in the model under consideration . since it is obvious to show that the cdf of @xmath178 is identical to that of @xmath110 , we focus only on the characterization of @xmath110 . the scheduling metric @xmath86 follows the chi - square distribution with @xmath181 degrees of freedom since it represents the sum of i.i.d . @xmath182 chi - square random variables with 2 degrees of freedom . similarly , the til @xmath110 follows the chi - square distribution with @xmath183 degrees of freedom . the cdfs of the two metrics @xmath86 and @xmath110 are given by @xmath184 respectively , where @xmath185 is the gamma function and @xmath186 is the lower incomplete gamma function ( * ? ? ? ( 8.310.1 ) ) . we start from the following lemma . [ lem:1 ] for any @xmath187 , the cdfs of the random variables @xmath86 and @xmath110 in and are lower - bounded by @xmath188 and @xmath189 , respectively , where @xmath190 and @xmath191 is the gamma function . the detailed proof of this argument is omitted here since it essentially follows the similar line to the proof of ( * ? ? ? * lemma 1 ) with a slight modification . in the following theorem , we establish our first main result by deriving the lower bound on the total dof in the @xmath192 channel with interfering relays . [ thrm:1 ] suppose that the ond scheme with alternate relaying is used for the @xmath121 channel with interfering relays . then , for @xmath26 data transmission time slots , @xmath193 is achievable if @xmath194 . from ( [ eq : dof_ond])([eq : sinr_2_pi2 ] ) , the ond scheme achieves @xmath195 provided that the two values @xmath196 and @xmath197 are less than or equal to some constant @xmath198 , independent of snr , for all s d pairs . then , a lower bound on the achievable @xmath122 is given by @xmath199 which indicates that @xmath200 dof is achievable for a fraction @xmath201 of the time for actual transmission , where @xmath202 we now examine the relay scaling condition such that @xmath201 converges to one with high probability . for the simplicity of the proof , suppose that the first and the second relay sets @xmath140 and @xmath169 are selected out of two mutually exclusive relaying candidate sets @xmath203 and @xmath204 , respectively , i.e. , @xmath205 , @xmath206 , @xmath207 , @xmath208 , and @xmath209 . then , we are interested in how @xmath210 and @xmath211 scale with snr in order to guarantee that @xmath201 tends to one , where @xmath212 denotes the cardinality of @xmath213 for @xmath130 . from ( [ eq : pond : def0 ] ) , we further have @xmath214 let @xmath215 with @xmath216 and @xmath217 be the candidate set associated with the second relay set and the @xmath218th s - d pair and its cardinality , respectively . for a constant @xmath198 , we can bound the second term in as follows : @xmath219 where the inequality @xmath220 holds from the de morgan s law ; @xmath221 follows from the union bound ; @xmath222 follows since @xmath223 ; @xmath224 follows since @xmath225 are the i.i.d . random variables @xmath226 for a given @xmath18 , owning to the fact that the channels are i.i.d . variables ; and @xmath227 follows from lemma [ lem:1 ] with @xmath228 since @xmath229 as @xmath230 and from the fact that @xmath231 . we now pay our attention to the first term in , which can be bounded by @xmath232 where the equality follows from the fact that @xmath233 and @xmath234 for @xmath235 are the functions of different random variables and thus are independent of each other . by letting @xmath236 , by the definition of @xmath233 , we have @xmath237 where the inequality follows from the fact that for any random variables @xmath238 and @xmath239 , @xmath240 @xmath241 @xcite . in the same manner , let @xmath242 with @xmath243 and @xmath244 be the candidate set associated with the first relay set and the @xmath218th s d pair and its cardinality , respectively . then , we can bound the first two terms in the rhs of as follows : @xmath245 where the last inequality follows from lemma [ lem:1 ] with @xmath246 . finally , from , it follows that @xmath247 tends to zero as @xmath248 grows large by noting that @xmath249 due to the reciprocal property of tdd systems . from , , and , it is obvious that if @xmath250 and @xmath248 scale faster than @xmath251 and @xmath7 , respectively , then @xmath252 therefore , @xmath201 asymptotically approaches one , which means that the dof of @xmath253 is achievable with high probability if @xmath254 . this completes the proof of the theorem . note that the lower bound on the dof asymptotically approaches @xmath1 for large @xmath26 , which implies that our system operates in virtual full - duplex mode . the parameter @xmath2 required to obtain full dof ( i.e. , @xmath1 dof ) needs to increase exponentially with the number of s d pairs , @xmath1 , in order to make the sum of @xmath255 interference terms in the til metric ( [ eq : til ] ) non - increase with increasing snr at each relay . scales slower than @xmath7 , it does not affect the performance in terms of dof and relay scaling laws . ] here , from the perspective of each relay in @xmath56 , the snr exponent @xmath255 indicates the total number of interference links and stems from the following three factors : the sum of interference power received from other sources , the sum of interference power generated to other destinations , and the sum of inter - relay interference power generated from the relays in @xmath38 . from theorem [ thrm:1 ] , let us provide the following interesting discussions regarding the dof achievability . @xmath1 dof can be achieved by using the proposed ond scheme in the @xmath0 channel with interfering links among relay nodes , if the number of relay nodes , @xmath2 , scales faster than @xmath7 and the number of transmission slots in one block , @xmath26 , is sufficiently large . in this case , all the interference signals are almost nulled out at each selected relay by exploiting the multiuser diversity gain . in other words , by applying the ond scheme to the interference - limited @xmath0 channel such that the channel links are inherently coupled with each other , the links among each s d path via one relay can be completely decoupled , thus enabling us to achieve the same dof as in the interference - free channel case . it is not difficult to show that the centralized relay selection method that maximizes the received sinr ( at either the relay or the destination ) using global csi at the transmitter , which is a combinatorial problem with exponential complexity , gives the same relay scaling result @xmath194 along with full dof . however , even with our ond scheme using a decentralized relay selection based only on local csi , the same achievability result is obtained , thus resulting in a much easier implementation . in this subsection , we analyze the til decaying rate under the ond scheme with alternate relaying , which is meaningful since the desired relay scaling law is closely related to the til decaying rate with respect to @xmath2 for given snr . let @xmath256 denote the @xmath1th smallest til among the ones that @xmath2 selected relay nodes compute . since the @xmath1 relays yielding the til values up to the @xmath1th smallest one are selected , the @xmath1th smallest til is the largest among the tils that the selected relays compute . similarly as in @xcite , by markov s inequality , a lower bound on the average decaying rate of @xmath256 with respect to @xmath2 , @xmath257 $ ] , is given by @xmath258\ge \frac{1}{\epsilon}\pr(l_{k\text{th - min}}\le \epsilon),\end{aligned}\ ] ] where the inequality always holds for @xmath259 . we denote @xmath260 as the probability that there are only @xmath1 relays satisfying @xmath261 , which is expressed as @xmath262 where @xmath263 is the cdf of the til . since @xmath264 is lower - bounded by @xmath265 , a lower bound on the average til decaying rate is given by @xmath258\ge \frac{1}{\epsilon}\mathcal{p}_k(\epsilon ) . \label{eq : til_min}\end{aligned}\ ] ] the next step is to find the parameter @xmath266 that maximizes @xmath260 in terms of @xmath267 in order to provide the tightest lower bound . [ lem:2 ] when a constant @xmath266 satisfies the condition @xmath268 , @xmath269 in ( [ eq : p_k ] ) is maximized for a given @xmath2 . to find the parameter @xmath266 that maximizes @xmath260 , we take the first derivative with respect to @xmath267 , resulting in @xmath270 which is zero when @xmath271 the parameter @xmath266 is the unique value that maximizes @xmath260 since @xmath272 which completes the proof of the lemma . now , we establish our second main theorem , which shows a lower bound on the til decaying rate with respect to @xmath2 . [ thrm:2 ] suppose that the ond scheme with alternate relaying is used for the @xmath0 channel with interfering relays . then , the decaying rate of til is lower - bounded by @xmath258 \ge \theta\left(n^{\frac{1}{3k-2}}\right).\end{aligned}\ ] ] as shown in ( [ eq : til_min ] ) , the til decaying rate is lower - bounded by the maximum of @xmath273 over @xmath267 . by lemma [ lem:2 ] , @xmath274 is maximized when @xmath275 . thus , we have @xmath258 & \ge \frac{1}{\mathcal{f}_l^{-1}(k / n)}{{n}\choose{k}}\left(\frac{k}{n}\right)^k \left(1-\frac{k}{n}\right)^{n - k } \nonumber\\ & \ge \frac{1}{\mathcal{f}_l^{-1}(k / n)}\left(\frac{n - k+1}{n}\right)^k \!\!\!\left(1-\frac{k}{n}\right)^{n - k } \nonumber\\ & \ge \frac{1}{\mathcal{f}_l^{-1}(k / n ) } \left(\frac{1}{k}\right)^k e^{-k } \nonumber\\ & \ge \theta\left(n^{\frac{1}{3k-2}}\right ) , \nonumber\end{aligned}\ ] ] where the second and third inequalities hold since @xmath276 and @xmath277 , respectively . by lemma [ lem:1 ] , it follows that @xmath278 , where @xmath279 is given by ( [ eq : c1 ] ) . hence , the last inequality also holds , which completes the proof of the theorem . from theorem [ thrm:2 ] , the following valuable insight is provided : the smaller snr exponent of the relay scaling law , the faster til decaying rate with respect to @xmath2 . this asymptotic result will be verified in a finite @xmath2 regime via numerical evaluation in section [ sec:5 ] . for comparison , the ond scheme without alternate relaying is also explained in this subsection . it is worth noting that there exists a trade - off between the lower bound on the dof and the minimum number of relays required to guarantee our achievability result by additionally introducing the ond protocol without alternate relaying . in the scheme , the first relay set @xmath38 only participates in data forwarding . that is , the second relay set @xmath56 does not need to be selected for the ond protocol without alternate relaying . specifically , the steps of each node during one block are then described as follows : * time slot 1 : sources @xmath30 transmit their first encoded symbols @xmath31 , where @xmath32 represents the @xmath33th transmitted symbol of the @xmath17th source node . a set of @xmath1 selected relay nodes , @xmath34 , operating in receive mode at each odd time slot , listens to @xmath31 . other @xmath35 relay nodes and destinations @xmath36 remain idle . * time slot 2 : the @xmath1 relays in the set @xmath38 forward their first re - encoded symbols @xmath280 @xmath281 to the corresponding @xmath1 destinations . the @xmath1 destinations receive from @xmath44 @xmath45 and decode @xmath39 . the remaining @xmath35 relays keep idle . * the processes in time slots 1 and 2 are repeated to the @xmath55th time slot . * time slot @xmath26 : the @xmath1 relays in @xmath38 forward their re - encoded symbols @xmath282 @xmath283 to the corresponding @xmath1 destinations . the @xmath1 sources and the other @xmath35 relays remain idle . when @xmath16 is assumed to serve the @xmath17th s d pair @xmath85 ( @xmath20 and @xmath19 ) , it computes the scheduling metric @xmath86 in ( [ eq : il ] ) . according to the computed @xmath86 , a timer based method is used for relay selection as in section [ subsub:1strelay ] . because there is no inter - relay interference for the ond scheme without alternate relaying , it is expected that the minimum required @xmath2 to achieve the optimal dof is significantly reduced . our third main theorem is established as follows . [ thrm:3 ] suppose that the ond scheme without alternate relaying is used for the @xmath121 channel . then , for @xmath26 data transmission time slots , @xmath284 is achievable if @xmath285 . the detailed proof of this argument is omitted here since it basically follows the same line as the proof of theorem [ thrm:1 ] . in section [ sec:5 ] , it will be also seen that in a finite @xmath2 regime , there exists the case even where the ond without alternate relaying outperforms that of the ond with alternate relaying in terms of achievable sum - rates via computer simulations . in this section , to show the optimality of the proposed ond scheme in the @xmath0 channel with interfering relays , which consists of @xmath1 s d pairs and @xmath2 relay nodes , we derive an upper bound on the dof using the cut - set bound @xcite as a counterpart of the lower bound on the total dof in section [ subsec : ana ] . suppose that @xmath286 relay nodes are active , i.e. , receive packets and retransmit their re - encoded ones , simultaneously , where @xmath287 . this is a generalized version of our transmission since it is not characterized how many relays need to be activated simultaneously to obtain the optimal dof . we consider the two cuts @xmath288 and @xmath289 dividing our network into two parts in a different manner . let @xmath290 and @xmath291 denote the sets of sources and destinations , respectively , for the cut @xmath292 in the network ( @xmath293 ) . for the @xmath0 channel model with interfering relays , we now use the fact that there is no direct path between an s then , it follows that under @xmath288 , @xmath1 transmit nodes in @xmath294 are on the left of the network , while @xmath286 active relay nodes and @xmath1 ( final ) destination nodes in @xmath295 are on the right and act as receivers . in this case , we can create the @xmath296 multiple - input multiple - output ( mimo ) channel between the two sets of nodes separated by the cut . similarly , the @xmath297 mimo channel are obtained under the cut @xmath289 . it is obvious to show that dof for the two mimo channels is upper - bounded by @xmath1 . hence , it turns out that even with the half - duplex assumption , our lower bound on the dof based on the ond with alternate relaying asymptotically approaches this upper bound on the dof for large @xmath26 . note that this upper bound is generally derived regardless of whether the number of relays , @xmath2 , tends to infinity or not , whereas the scaling condition @xmath298 is included in the achievability proof . when @xmath299 and @xmath300 in the @xmath121 channel . it is assumed that @xmath300 and @xmath301 are used for the ond schemes with and without alternate relaying , respectively . , scaledwidth=57.0% ] when @xmath302 in the @xmath121 channel . , scaledwidth=57.0% ] in this section , we perform computer simulations to validate the achivability result of the proposed ond scheme in section [ sec:3 ] for finite parameters @xmath2 and snr in the @xmath0 channel model with interfering relays . in our simulation , the channel coefficients in ( [ eq : y1 ] ) and ( [ eq : y2 ] ) are generated @xmath303 times for each system parameter . figure [ fig:1 ] shows the achievable sum - rates of the @xmath304 channel for the ond schemes with and without alternate relaying according to @xmath305 in db scale when @xmath299 . note that @xmath2 is set to a different scalable value according to @xmath305 , i.e. , @xmath306 for the ond with alternate relaying and @xmath307 for the ond without alternate relaying , respectively , to see whether the slope of each curve follows the dof in theorems [ thrm:1 ] and [ thrm:2 ] . in the figure , the dotted green lines are also plotted to indicate the first order approximation of the achievable rates with a proper bias , where the slopes are given by @xmath1 and @xmath4 for the ond schemes with and without alternate relaying , respectively . in fig . [ fig:2 ] , the log - log plot of the average til of the ond with alternate relaying versus @xmath2 is shown for the @xmath308 channel when @xmath302 . is taken into account to precisely see some trends of curves varying with @xmath2 . ] it can be seen that the til tends to decrease linearly with @xmath2 . it is further seen how many relays are required with the ond scheme with alternate relaying to guarantee that the til is less than a small constant for a given parameter @xmath1 . in this figure , the dashed lines are also plotted from theoretical results in theorem [ thrm:2 ] with a proper bias to check the slope of the til . we can see that the til decaying rates are consistent with the relay scaling law condition in theorem [ thrm:1 ] . more specifically , the til is reduced as @xmath2 increases with the slope of 0.25 for @xmath299 and 0.143 for @xmath309 , respectively . when @xmath299 and @xmath310 in the @xmath121 channel . both ond schemes with and without alternate relaying are compared . , scaledwidth=57.0% ] when @xmath299 and @xmath310 in the @xmath121 channel . the ond scheme with alternate relaying and the max - min snr scheme are compared . , scaledwidth=57.0% ] figure [ fig:3 ] illustrates the achievable sum - rates of the @xmath121 channel for the ond schemes with and without alternate relaying versus @xmath305 ( in db scale ) when @xmath299 and @xmath311 . we can see that in a finite @xmath2 regime , there exists the case where the ond without alternate relaying outperforms that of the ond with alternate relaying . this is because for finite @xmath2 , the achievable sum - rates for the alternate relaying case tend to approach a floor with increasing snr faster than no alternate relaying case due to more residual interference in each dimension . we can also see that the crossing points slightly move to the right as @xmath2 increases ; this is due to the fact that our ond scheme with alternate relaying always benefits from having more relays for selection , thus resulting in more multiuser diversity gain . this highly motivates us to operate our system in a switched fashion when the relay selection scheme is chosen between the ond schemes with and without alternate relaying depending on the operating regime of our system . to further ascertain the efficacy of our scheme , a performance comparison is performed with a baseline scheduling . specifically , in the _ max - min snr _ scheme , each s d pair selects one relay @xmath16 ( @xmath20 ) such that the minimum out of the desired channel gains of two communication links ( either from @xmath14 to @xmath16 or from @xmath16 to @xmath15 ) becomes the maximum among the associated minimum channel gains over all the unselected relays . this max - min snr scheme is well - suited for relay - aided systems if interfering links are absent . the achievable sum - rates are illustrated in fig . [ fig:4 ] according to @xmath305 ( in db scale ) when @xmath299 and @xmath310 . we can see that our ond scheme with alternate relaying outperforms this baseline scheme beyond a certain low snr point . we also see that the rate gaps increase when @xmath2 increases in the high snr regime . on the other hand , for fixed @xmath2 , the sum - rates of the max - min scheme are slightly changed with respect to @xmath305 due to more residual interference in each dimension . an efficient distributed ond protocol operating in virtual full - duplex mode was proposed for the @xmath0 channel with interfering relays , referred to as one of multi - source interfering relay networks . a novel relay scheduling strategy with alternate half - duplex relaying was presented in two - hop environments , where a subset of relays is opportunistically selected in terms of producing the minimum total interference level , thereby resulting in network decoupling . it was shown that the ond protocol asymptotically achieves full dof even in the presence of inter - relay interference and half - duplex assumption , provided that the number of relays , @xmath2 , scales faster than @xmath7 . numerical evaluation was also shown to verify that our scheme outperforms the other relay selection methods under realistic network conditions ( e.g. , finite @xmath2 and snr ) with respect to sum - rates . i. shomorony and a. s. avestimehr , degrees of freedom of two - hop wireless networks : everyone gets the entire cake `` , '' in _ proc . 50th annual allerton conf . control , and computing _ , monticello , il , oct . t. gou , c. wang , and s. a. jafar , aligned interference neutralization and the degrees of freedom of the @xmath5 interference channel with interfering relays , " in _ proc . 49th annual allerton conf . commun . , control , and computing _ , monticello , il , sep . t. gou , c. wang , and s. a. jafar , degrees of freedom of a class of non - layered two unicast wireless networks , " in _ proc . 45th asilomar conf . signals , systems and computers _ , pacific grove , ca , nov . 2011 . h. j. yang , w .- y . shin , b. c. jung , and a. paulraj , `` opportunistic interference alignment for mimo interfering multiple access channels , '' _ ieee trans . wireless commun . _ , vol . 12 , no . 5 , pp . 21802192 , may 2013 . h. j. yang , w .- y . shin , b. c. jung , c. suh , and a. paulraj , `` opportunistic downlink interference alignment , '' in _ proc . inf . theory ( isit ) _ , honolulu , hi , jun./jul . 2014 , pp . 15881592 . c. shen and m. p. fitz , opportunistic spatial orthogonalization and its application to fading cognitive radio networks , " , vol . 5 , no . 1 , pp . 182189 , feb . 2011 . t. w. ban , w. choi , b. c. jung , and d. k. sung , multi - user diversity in a spectrum sharing system , " , vol . 1 , pp . 102106 , jan . 2009 . r. zhang and y. c. liang , exploiting multi - antennas for opportunistic spectrum sharing in cognitive radio networks , " , vol . 1 , pp . 88102 , feb . 2008 . y. fan , c. wang , j. thompson , and h. v. poor , recovering multiplexing loss through successive relaying using repetition coding , " , vol . 6 , no . 12 , pp . 44844493 , dec . 2007 . shin , h. j. yang , and b. c. jung , opportunistic network decoupling in multi - source interfering relay networks , " in _ proc . ieee conf . commun . ( icc ) _ , sydney , australia , jun . 2014 , pp . 26712676 .

models with more than four dimensions are interesting , because all the physical ingredients of the universe might be unified in a higher dimensional theory . string theory is the most promising scenario where quantum gravity is included by the requirement of additional dimensions and supersymmetry . the idea of large extra dimension@xcite is important , because it might solve the hierarchy problem . in this case , the observed planck mass is obtained by the relation @xmath0 , where @xmath1 and @xmath2 denote the fundamental scale of gravity and the volume of the @xmath3-dimensional compact space , respectively . in the scenarios of large extra dimension , fields in the standard model are expected to be localized on a wall - like structure , while the graviton propagates in the bulk . in the context of string theory , a natural embedding of this picture is realized by brane construction . the brane models are interesting from both phenomenogical and cosmological viewpoints . analyses of cosmological defect formation are important in brane models as well as in usual cosmological models . one may consider three types of brane defects . * defects are branes . + in this case , cosmological defects are formed by the branes that have less than three spacial dimensions in the uncompactified spacetime . in the previous discussions , it was concluded that the cosmological production of monopoles and domain walls are negligible . in this paper , however , we reconsider the cosmological formation of monopoles and domain walls and show explicitly how they can be formed by the brane dynamics . * defects are formed by the spatial deformation of branes . + in this case , cosmological defects are formed by the continuous deformation of branes . first in ref.@xcite , and later in ref.@xcite , the fields that parameterize the positions of the branes are shown to fluctuate in spatial directions to form the cosmological defects . cosmic strings are constructed in ref.@xcite , where the singularity is resolve by smearing the brane . monopoles , strings and domain walls are discussed in @xcite , where the relative positions between branes are utilized . * the center of the localized matter fields are shifted in the defect . + in ref.@xcite , the localization of the matter fields on a fat domain wall is discussed to explain the small interactions . this idea is important when the suppression factor is required in the interaction , so that it does not violate the constraints from the proton lifetime . in ref.@xcite , we have constructed defect configurations , which induce the shift of the center of the localized matter fields . these defects are used to produce the baryon asymmetry of the universe@xcite . in this paper , we reconsider the cosmological formation of brane defects , which are described by branes which have less than three spacial dimensions in the uncompactified space . in our model , the defect branes are extended between branes . we will also consider a cosmological brane creation induced by the excitation of a sphaleron - like brane configuration . first we will review the essence of the previous arguments about the cosmological formation of brane defects . in the original scenario of brane inflation@xcite , inflationary expansion is driven by the potential between d - brane and anti - d - brane evolving in the bulk . then the scenario of the inflating branes at a fixed angle is studied in ref.@xcite , where the slow - roll condition is improved by introducing a small angle . the end of brane inflation is induced by the brane collision where brane annihilation ( or recombination ) proceeds through tachyon condensation@xcite . during brane inflation , tachyon is trapped in the false vacuum , which may result in the formation of lower - dimensional branes after brane inflation . the production of cosmological brane defects was discussed in ref.@xcite , where it was concluded that cosmic strings are copiously produced in these scenarios , but monopoles and domain walls are negligible . in ref.@xcite , however , it was discussed that all kinds of defects are produced , and the conventional problems of domain walls and monopoles arise . later in ref.@xcite , the formation of cosmological defects was reexamined and the conclusion was different from @xcite and @xcite . as was noted in ref.@xcite , the effect of compactification is significant for the defect formation , if it is induced by tachyon condensation after brane inflation . since the compactification radius must be small compared to the horizon size during inflation , any variation of a field in the compactified direction is suppressed . then the daughter brane , which is the defect formed by tachyon condensation on the worldvolume of the mother brane , wraps the same compactified space as the mother brane . as a result , the codimension of the daughter brane should lie within the uncompactified space . since the number of the codimension must be even , the cosmological defect is inevitably a cosmic string . in ref.@xcite , it was discussed that the analysis does not fully account for the effects of compactification , because the directions transverse to the mother brane had not been considered . in ref.@xcite , the effect of the rr fields , which are extended to the compactified dimensions , were discussed . their conclusion was that the creation of the gradients of the rr fields in the bulk of the compactified space is costly in energy , so that the creation of the daughter d brane is suppressed if they do not fill all the compactified dimensions . here it should be noted that the above arguments are not fully reliable . the most obvious example is the inconsistency of the tension of strings produced after angled brane inflation . in the original argument in ref.@xcite , the @xmath4-dependence of the tension of the string did not coincide with the one calculated from the effective lagrangian . what is wrong with the above arguments ? the reason is clearly described in ref.@xcite . in the original argument of ref.@xcite , it was discussed that the daughter branes are created on the mother brane . in fact the statement is true , but we should be more careful about the process of the brane recombination . when the distance between the mother branes becomes shorter than a critical distance , tachyon starts to condensate on a mother brane . therefore the tachyon forms strings on the mother brane when the mother branes start to recombine . however , as we have depicted in fig.[fig : split ] , the daughter @xmath5-brane must be pulled out of the @xmath6-branes . as a result , the daughter @xmath5 brane does not wrap the same compactified space as the mother brane . the tension of the cosmic string is calculated in ref.@xcite , which depends on @xmath4 and matches to the effective action . therefore , the daughter brane in fig.[fig : split ] satisfies the property required from analysis of the effective lagrangian . thus for the angled inflation model , it is obvious that the cosmic string does not wrap the same compactified space as the mother brane . from the above discussions , it is appearent that the original argument of ref.@xcite fails because there is no obvious reason that in their final state the daughter branes wrap the same compactified space as the mother brane . the idea , which supports the formation of daughter branes that are extended between mother branes , is quite important for our discussions in this paper . since the previous arguments can not fully account for the generic processes of cosmological defect formation , one can hardly accept their conclusions without reconsiding the production of monopoles and domain walls paying careful attention to the brane dynamics . is it really impossible to produce sufficient amount of cosmological monopoles and domain walls by the branes ? as we have discussed in ref.@xcite , crucial modifications are required for the previous arguments , since it is possible to produce daughter branes that do not wrap the same compactified space as the mother brane . in this paper we will reconsider the formation of cosmological monopoles and domain walls by using the new idea . for the extended branes to be produced between mother branes , the distance between mother branes are required to vanish at a moment . the obvious examples are ; * recombination after angled brane inflation . + at the time when the recombination starts after angled brane inflation , the distance between branes vanishes as required . thus , daughter branes can be pulled out of splitting branes , which finally become the cosmological defect in the effective action . as a result , the cosmological defect in the effective action is the brane extended between mother branes . two examples are already discussed in ref.@xcite and ref.@xcite , where the formation of cosmic strings are considered . * two branes coincide during inflation . + in the effective action , the brane distance plays the role of a trigger field of hybrid inflation . this type of brane inflation is already discussed in ref.@xcite and ref.@xcite , in which the formation of unstable semilocal strings are suggested . in this paper , using the model in ref.@xcite , we will show that a point - like object that corresponds to a monopole in the effective action can be produced as a daughter @xmath7 brane . the daughter @xmath7 brane is extended between @xmath8 branes . we calculate the tension of the extended object and show that it coincides with the mass of a monopole in the effective action . * branes collide or oscillate with huge kinetic energy . + after brane collision , there would be chaotic processes , which include oscillation , recombination and production / annihilation of branes . therefore , it seems natural to expect that there could be a stage of preheating after inflation , which induce large occupation numbers for long - wavelength configurations that is necessary for the domain wall production from sphalerons . then , extended objects could be formed efficiently between oscillating branes . as usual , these non - equilibrium processes are always important for the discussions about cosmological defect formation . as in the case of the conventional electroweak phase transition , we do not exclude the possibility that the final state of the oscillating branes are separated at a distance , which implies a spontaneous symmetry breaking in the effective lagrangian . the defect branes could be created by the tachyon condensation on the mother brane , or could have been existed from the beginning and have been dissolved in the mother brane . in section 2 , we consider the model of brane inflation in ref.@xcite and study the cosmological formation of monopoles after brane inflation . in section 3 , we discuss the formation of cosmological domain walls . as expected , there are notable differences between the previous arguments and our conclusions . the cosmological formation of monopoles and domain walls are not always negligible in models of brane cosmology . our results are consistent with the analyses of the effective action . in this section , we discuss the formation of monopoles in a specific example of brane inflation . we first repeat the previous arguments in ref.@xcite , which suggest that monopoles are not produced by the usual daughter brane creation after inflation . as is noted in ref.@xcite , the effect of compactification is significant for the daughter brane creation . since the compactification radius is small compared to the horizon size during inflation , any variation of a field in the compactified direction is suppressed . as a result , the daughter brane wraps the same compactified dimensions as the mother brane . the codimension of the daughter branes should lie within the uncompactified space , which suggests that the defect is inevitably a cosmic string . seeing the above arguments , one might conclude that the monopoles are not produced after brane inflation . is the production of monopoles really impossible ? to answer this question , we consider a simple example where monopoles are produced by the daughter brane creation after brane inflation . to avoid the ambiguity of the brane products , we consider a model in which the daughter brane formation is described without such ambiguity . the schematic representation of our idea is given in fig.[fig : monopole ] . the typical configuration of fig.[fig : monopole ] will also appear in generic cosmological processes where branes collide , recombine or stick together by thermal effects . in any case , if the production of monopoles or domain walls is suggested in the analysis of the effective action , it is natural to think that it should be possible to explain how they can be formed by the cosmological brane dynamics . moreover , even if the production of such defects is not suggested in the effective action , they might be produced by the pure brane dynamics beyond the cut - off scale of the effective action . first we make a brief review of the halyo s idea for brane inflation on conifolds@xcite . the important aspect of this scenario is that the location of the inflating branes coincide at the end of inflation , then the branes separate along the @xmath9 compactified space . the model is described as a d - brane inflation on fractional @xmath7 branes transverse to a resolved and deformed conifold . the d - term inflation is induced by the conifold that is resolved by the blow - up of its tip . in the effective action , the model looks like a model of hybrid d - term inflation . the slow roll is described by the slow motion of the two fractional @xmath7 branes approaching each other along a compactified direction , and the trigger field parameterizes the distance between the two branes along the other compactified space in the base of the conifold . for example , one can consider a conifold that is described as a cone over @xmath10 , where a @xmath7 brane , which is transverse to the conifold , is separated into two fractional d3 branes on the @xmath11 . the conifofd singularity is resolved by replacing the tip of the cone with an @xmath9 of finite size , which induces inflation . the inflaton mass arises from another deformation of the conifold towards @xmath12 and results in a slow motion of the two fractional branes towards each other . as we are not interested in the parameter space of inflation itself , we spare the discussions about the condition for successful inflation . for us , the important point is to understand the formation of daughter branes after inflation , which finally become extended between the separating mother branes . as was already discussed in ref.@xcite , one usually expect that the cosmic strings , which has codimension 2 in the uncompactified space , are formed after inflation . if the inflating @xmath7 brane is a @xmath8 brane wrapped on the @xmath13 with radius @xmath14 , the tension of the @xmath7 brane is @xmath15 where @xmath16 and @xmath17 are the @xmath13 metric and ns - ns field on the @xmath13 respectively . taking @xmath18 for simplicity , one can find @xmath19 the tension of the inflating @xmath7 brane must be equal to the energy density during inflation , which corresponds to the anomalous d - term in the effective lagrangian . thus in this case , the relation @xmath20 is required , where @xmath21 is the gauge coupling in the effective action . the cosmic strings are produced by tachyon condensation , which are the daughter @xmath7 branes on the worldvolume of the inflation brane . the daughter brane wraps around the same compactified space as the mother brane . in this case , however , the symmetry of the corresponding field in the effective action allows only semilocal strings , which is unstable to dissolve in space , as is expected from the brane dynamics . let us consider a configuration of fig.[fig : monopole ] . if the seed of the daughter brane fluctuate between the two coincidental mother branes , it will be pulled out of the mother branes . the branes that are extended between mother branes are point - like objects in the effective action on the mother brane . it is easy to calculate the tension of such daughter branes and compare it with the mass of the monopoles in the effective action . string ending on a @xmath7 brane provides a magnetic source for the three - brane worldvolume gauge field . our example is a simple modification of the scenario . however , it was still unclear if such extended branes are produced after brane inflation . as we have repeated above , the previous arguments were negative to the cosmological production of such monopoles . ] in the final state of brane inflation , two branes are separated at a distance of @xmath22 . the effective mass of the extended daughter brane is @xmath23 where the gauge coupling is given by the formula @xmath24 it is easy to see that the mass of the extended daughter brane coincides with the mass of the monopole in the effective action@xcite . we can extend the above analysis to the cases where more than two branes are oscillating around each other ( or stick together by the thermal effects ) and then fall apart along a direction of the compactified space . in this case , one can construct @xmath25 monopoles , which are the daughter branes extended between @xmath26 mother branes . a schematic representation is given in fig.[fig : monopole_n ] . seeing the above arguments , we can not help thinking about the m theory fivebrane version of qcd ( mqcd ) , where the quantum effects are explained by the brane dynamics . in the most basic model of mqcd , which is depicted in fig.[fig : mqcdmonopole ] , a monopole in type iia string theory@xcite is a rectangular @xmath27 brane with two boundaries on @xmath28 branes and other two on ns brane . the mass of such monopole is proportional to the minimal area of the hole between the branes . in realistic models , such brane configuration is expected to be embedded in the compactified space . is the cosmological formation of the @xmath27 monopoles impossible ? in this case , the following conditions are required for the production of the @xmath27 brane monopoles . * there was a period when the two @xmath28 branes coincide . * @xmath27 brane is created on the worldvolume of the @xmath28 brane . when the area shrinks to zero , the monopole becomes the massless excitation . the @xmath27 brane is produced on the worldvolume of the @xmath28 branes , due to the usual mechanism of tachyon condensation . in fig.[fig : monopoled2 ] , we show why the @xmath27 brane can be extended between the separating @xmath28 branes . as usual , the daughter brane does not fluctuate in any direction of the compactified space , while it fluctuate in the uncompactified directions . as a result , the @xmath27 monopoles are formed when the @xmath28 branes start to fall apart . the most natural situation is that the symmetry restoration is induced by the thermal effects , which break supersymmetry and glue branes together . in the effective action , the restoration of the symmetry is naturally induced by the thermal effects . the repulsive force between the two @xmath28 branes are induced by the usual supersymmetry breaking . the non - equilibrium production of such monopoles is possible during brane oscillation . the massless monopoles gain mass when the branes start to fall apart . as a result , we conclude that the production of monopoles after brane inflation is quite natural , even if the monopoles are the daughter branes created by the usual mechanism of tachyon condensation . in the previous arguments in ref.@xcite , it was concluded that the formation of cosmological domain walls are always negligible after brane inflation , because the daughter branes must wrap the same compactified space as the mother brane . is it impossible to produce daughter branes that do not wrap the same compactified space as the mother brane ? first , we repeat the previous discussions , which suggest that the cosmological formation of domain walls is forbidden . as is discussed in ref.@xcite and @xcite , the effect of compactification is believed to play significant roles in the defect formation due to tachyon condensation . since the compactification radius should be small compared to the horizon size during inflation , any variation of a field in the compactified direction must be suppressed . as a result , the daughter brane will wrap the same compactified space as the mother brane . at the same time , since the codimension of the daughter branes should lie within the uncompactified space , the defect is inevitably a cosmic string . on the other hand , as we have discussed in ref.@xcite for cosmic strings and in the previous section for monopoles , it is possible to produce daughter branes that are extended between branes , which correspond to the d - term strings or monopoles in the effective action . as we have seen in the previous section , efficient production of such monopoles is allowed in generic situations . for cosmic strings , the idea of the extended daughter brane formation was used to solve the problem of the angle - dependence of the string tension in the scenario of angled brane inflation@xcite . the cosmological formation of the extended branes plays important roles in the analyses of strings and monopoles . finally , we will examine the formation of cosmological domain walls . our purpose in this section is to examine the cosmological formation of domain walls . actually , it is possible to show an example of the domain wall formation , although this example is not a pure creation of branes induced by the tachyon condensation on the mother brane . please imagine that two coincident 5-branes are placed on top of each other . both of them have four space - time dimensions(uncompactified ) and also wrap two - dimensional compactified space . these 5-branes look like 3-branes in the limit where the wrapped space is very small . let us assume that after a phase transition these two 5-brnes finally fall apart to the true vacuum . in this case , it is possible to add an additional 3-brane by hand , which does not wrap the same compactified space as the above 5-branes . in this case , one can assume that the latter 3-brane is not produced by the tachyon condensation on the 5-brane , but could be present from the beginning . the latter 3-brane have four ( uncompactified ) space - time dimensions , and dissolves in the 5-branes . then , it is quite natural to think that the additional 3-brane sticks to either 5-brane when the 5-branes finally fall apart . then there appears two kinds of domains in the universe , depending on which 5-brane the additional 3-brane sticks to . since the domain walls can not interact across the scale much larger than the hubble radius , one can expect that at least one domain wall is produced in one huble horizon . the resulting domain wall is the 3-brane extended between the 5-branes . we can discuss brane creation from brane sphalerons . as far as the effective action is applicable , it seems possible to create new branes that do not wrap the same compactified radius as the mother brane , even if it is initially created on the worldvolume of the mother brane . our second example for the formation of cosmological domain walls is the domain walls produced by the excitation of sphaleron - like brane configuration@xcite . brane could have their boundaries on the mother @xmath29 branes . ] the most obvious difference from the scenario of tachyon condensation is that the sphaleron - like configuration is not required to wrap the same compactified space as the mother brane , even when they are first created on the worldvolume of the mother brane . in this case , for example , cosmological domain walls can be the @xmath7 branes extended between @xmath8 branes . the schematic picture of the domain wall is given in fig.[fig : sphaleron1 ] . if there is a non - equilibrium process that enhances the long - wavelength configurations of the sphalerons , the efficient production of the sphaleron excitations will happen . then the sphaleron - like branes will recombine , as is shown in fig.[fig : sphaleron2 ] . although the production of massive sphalerons is suppressed by an exponential factor in conventional thermal processes , the production of sphalerons are not suppressed in non - equilibrium processes such as preheating after inflation . thus , the extended branes that are formed by the sphaleron - like excitations could be produced if there was a period when the long - wavelength configurations of the sphalerons are enhanced . in this case , one can not ignore the production of domain walls . the period when branes oscillate is inevitable in generic cosmological models , as we have discussed in the previous section . since the conventional electroweak sphaleron is quite similar to the sphaleron - like brane configuration in the effective action , we start the discussion comparing the electroweak sphalerons to the brane - sphalerons . first , we briefly review the sphaleron interaction in the electroweak gauge theory . at zero temperature , the rate per unit volume of baryon number violating processes is exponentially suppressed by the factor @xmath30 in the high temperature , however , the baryon number violating processes that are mediated by sphalerons are not exponentially suppressed@xcite . the sphalerons have typically the sizes given by the magnetic correlation length @xmath31 in this case , one may think that the space is divided up into cells of this size , although these massless sphalerons look nothing like sphalerons because the `` barrier '' does not make sense when the symmetry is restored by the thermal effects . there is another idea for the sphaleron production@xcite , where the non - equilibrium sphaleron transition is used to explain the baryon asymmetry of the universe . for the brane dynamics , sphaleron - like brane configuration is discussed in ref.@xcite . the configuration is described by an appropriate excitation of the transverse coordinate field , which corresponds to the higgs expectation value in the discussion of the electroweak sphalerons . here we do not repeat the analysis in ref.@xcite , because the analogy comparing the sphaleron - like brane configuration to the electroweak sphaleron seems appropriate and sufficient for qualitative discussions , as far as the effective action is applicable . the sphaleron - like brane configuration is excited when the following conditions are satisfied . * more than two branes coincide during inflation . + after inflation , the branes fall apart toward the true vacuum configuration . in the effective action , this corresponds to the symmetry restoration during hybrid inflation . * brane inflation ends with brane collision . + if the brane collision is accompanied by the brane oscillation , sphaleron - like branes will be produced during this period . in the effective action , this corresponds to the non - equilibrium sphaleron production , which is discussed in ref.@xcite . * symmetry restoration is induced by the thermal effects . + spontaneous breaking of the symmetry in the effective action is explained by brane separation or brane recombination . sphalerons are excited when they are massless . sphalerons - like brane configuration is expanded between mother branes and obtains mass when the mother branes fall apart . in general , one can not simply ignore the production of the sphaleron - like brane configurations , irrespective of their stability . even if there was no non - equilibrium processes that makes the sphalerons form stable global structure of massive domain wall networks , there is a possibility that unstable defects ( and their decay products ) play important roles in the evolution of the universe . is unstable , because it must shrink to a point as it obtains mass . however , in the special cases that multiple sphalerons are produced at the same time within the same horizon , multiple sphalerons will recombine into larger or smaller cells , as is depicted in fig.[fig : sphaleron2 ] . of course , it is not trivial if the sphalerons could interact each other to form the significant global structure of the domain wall networks . if the correlation length is short , sphalerons are small even if they are massless . if the nucleation rate of the sphalerons is too small to make them interact each other to form larger structure of the domain wall networks , sphalerons must shrink to a point after the phase transition . in this case , sphalerons can not become stable domain walls . therefore , it is required that a non - trivial mechanism that enhances the correlation length(or the nucleation rate of the sphalerons ) so that the sphalerons interact each other to seed the domain walls . in the scenario of preheating , the domain wall formation is already convinced by simulation . in the dynamics of brane sphalerons , the brane distance is corresponding to the higgs field . in this respect , the domain wall formation is likely to occur in the brane - motivated models , when some conditions are satisfied . as a result , it is important to note that one can not simply ignore the domain wall production after brane inflation . ] in ref.@xcite , we have discussed the formation of the domain walls , which are induced by the spatial deformation of the @xmath28 branes . the domain walls formed by the spatial deformation of the branes are , at least in principle , different from the ones that are formed by the brane creation . however , since the cosmological requirement for their formation is the same@xcite , the actual cosmological defects are inevitably the mixture of the two different kinds of brane defects . it was believed that monopoles and domain walls can not be produced by the daughter brane creation after brane inflation , because the daughter brane must wrap the same compactified space as the mother brane . as we have discussed in this paper , the above statement is not fully reliable . in the final state , daughter branes are not always expected to wrap the same compactified space as the mother brane . the most obvious example is found in the discussion about the inconsistency of the tension of the strings produced after angled brane inflation . in the original argument in ref.@xcite , the @xmath4-dependence of the tension of the string did not coincide with the the analysis of the effective lagrangian . in the original argument@xcite , it was discussed that the daughter branes are created on the mother brane . in fact the statement is true , but we should be more careful about the process of the brane recombination . when the distance between the mother branes becomes shorter than a critical distance , tachyon starts to condensate on a mother brane . therefore the tachyon forms strings on the mother brane and wraps the same compactified space as the mother brane . however , as we have depicted in fig.[fig : split ] , the daughter brane must be pulled out if there was a recombination of the mother branes . as a result , the daughter brane does not wrap the same compactified space as the mother brane . thus for the angled inflation model , the cosmic strings can depend on @xmath4 , which is the property required from analysis of the effective lagrangian . therefore the original argument of ref.@xcite fails because there is no obvious reason that in their final state the daughter branes wrap the same compactified space as the mother brane . in this paper , we have reconsidered the production of monopoles and domain walls , paying special attention to the brane dynamics . as a result , we have found that monopoles are produced by daughter brane creation . we have also suggested that domain walls could be produced by sphaleron - like brane creation if there is an enhancement of long - wavelength configurations of the brane - sphalerons . our conclusions are consistent with the analyses of the effective action . another type of brane defects@xcite , which are formed by the deformation of branes , can also be produced by the same cosmological processes . therefore , the actual cosmological relics are the mixture of the two kinds . arguments about the evolution of the mixed brane defects are interesting and deserve further investigation . we wish to thank k.shima for encouragement , and our colleagues in tokyo university for their kind hospitality . 1 i. antoniadis , n. a - hamed , s. dimopoulos , and g. r. dvali , _ new dimensions at a millimeter to a fermi and superstrings at a tev , phys.lett.b436(1998)257_ [ hep - ph/9804398 ] ; i. antoniadis , _ a possible new dimension at a few tev , phys.lett.b246(1990)377_ ; n. a - hamed , s. dimopoulos and g. r. dvali , _ the hierarchy problem and new dimensions at a millimeter , phys.lett.b429(1998)263_ [ hep - ph/9803315 ] . n. arkani - hamed , s. dimopoulos , n. kaloper , and j. march - russell , _ rapid asymmetric inflation and early cosmology in theories with submillimeter dimensions , nucl.phys.b567(2000)189_ [ hep - ph/9903224 ] ; r. n. mohapatra , a. perez - lorenzana , and c. a. de s. pires , _ inflation in models with large extra dimensions driven by a bulk scalar field , phys.rev.d62(2000)105030_ [ hep - ph/0003089 ] ; a. mazumdar , _ extra dimensions and inflation , phys.lett.b469(1999)55_ [ hep - ph/9902381 ] ; a. m. green and a. mazumdar , _ dynamics of a large extra dimension inspired hybrid inflation model , phys.rev.d65(2002)105022_ [ hep - ph/0201209 ] ; d. h. lyth , _ inflation with tev scale gravity needs supersymmetry , phys.lett.b448(1999)191_ [ hep - ph/9810320 ] ; p. kanti and k. a. olive , _ on the realization of assisted inflation , phys.rev.d60(1999)043502_ [ hep - ph/9903524 ] ; p. kanti and k. a. olive , _ assisted chaotic inflation in higher dimensional theories , phys.lett.b464(1999)192_ [ hep - ph/9906331 ] ; t. matsuda , _ kaluza - klein modes in hybrid inflation , phys.rev.d66(2002)107301_ [ hep - ph/0209214 ] ; t. matsuda , _ successful d term inflation with moduli , phys.lett.b423(1998)35_ [ hep - ph/9705448 ] . t. matsuda , _ topological hybrid inflation in brane world , jcap 0306(2003)007 _ [ hep - ph/0302204 ] ; t. matsuda , _ q ball inflation , phys.rev.d68(2003)127302_ [ hep - ph/0309339 ] . g. r. dvali , g. gabadadze , _ nonconservation of global charges in the brane universe and baryogenesis , phys.lett.b460(1999)47_ [ hep - ph/9904221 ] ; a. masiero , m. peloso , l. sorbo , and r. tabbash , _ baryogenesis versus proton stability in theories with extra dimensions , phys.rev.d62(2000)063515_ [ hep - ph/0003312 ] ; a. pilaftsis , _ leptogenesis in theories with large extra dimensions , phys.rev.d60(1999)105023_ [ hep - ph/9906265 ] ; r. allahverdi , k. enqvist , a. mazumdar and a. perez - lorenzana , _ baryogenesis in theories with large extra spatial dimensions , nucl.phys . b618(2001)377 _ [ hep - ph/0108225 ] ; s. davidson , m. losada , and a. riotto , _ a new perspective on baryogenesis , phys.rev.lett.84(2000)4284_ [ hep - ph/0001301 ] . t. matsuda , _ activated sphalerons and large extra dimensions , phys.rev.d66(2002)047301_ [ hep - ph/0205331 ] ; t. matsuda , _ defect mediated electroweak baryogenesis and hierarchy , j.phys.g27(2001)l103_ [ hep - ph/0102040 ] . t. matsuda , _ hybridized affleck - dine baryogenesis , phys.rev.d67(2003)127302_ [ hep - ph/0303132 ] ; t. matsuda , _ affleck - dine baryogenesis after thermal brane inflation , phys.rev.d65(2002)103501_ [ hep - ph/0202209 ] ; t. matsuda , _ affleck - dine baryogenesis in the local domain , phys.rev.d65(2002)103502_ [ hep - ph/0202211 ] ; t. matsuda , _ electroweak baryogenesis mediated by locally supersymmetry breaking defects , phys.rev.d64(2001)083512_ [ hep - ph/0107314 ] . t. matsuda , _ brane q - ball , branonium and brane inflation _ [ hep - ph/0402223 ] . t. matsuda , _ formation of cosmological brane defects _ [ hep - ph/0402232 ] . g. r. dvali , i. i. kogan and m. a. shifman , _ topological effects in our brane world from extra dimensions , phys.rev.d62(2000)106001_ [ hep - th/0006213 ] . t. matsuda , _ incidental brane defects , jhep 0309(2003)064 _ [ hep - th/0309266 ] . n.arkani-hamed , m.schmaltz , _ hierarchies without symmetries from extra dimensions , phys.rev.d61(2000)033005_ [ hep - ph/9903417 ] . t. matsuda , _ baryon number violation , baryogenesis and defects with extra dimensions , phys.rev.d66(2002)023508_ [ hep - ph/0204307 ] ; t. matsuda , _ enhanced baryon number violation due to cosmological defects with localized fermions along extra dimensions , phys.rev.d65(2002)107302_ [ hep - ph/0202258 ] . g. r. dvali and s. h. henry tye _ brane inflation , phys.lett.b450(1999)72_ [ hep - ph/9812483 ] . c. herdeiro , s. hirano and r. kallosh , _ string theory and hybrid inflation / acceleration , jhep0112(2001)027 _ [ hep - th/0110271 ] ; k. dasgupta , c. herdeiro , s. hirano and r. kallosh , _ d3 / d7 inflationary model and m theory , phys.rev.d65(2002)126002_ [ hep - th/0203019 ] , j. garcia - bellido , r. rabadan and f. zamora , _ inflationary scenarios from branes at angles , jhep 0201(2002)036 _ [ hep - th/0112147 ] . a. sen , _ rolling tachyon , jhep 0204(2002)048 _ [ hep - th/0203211 ] ; a. mazumdar , s. panda and a. perez - lorenzana , _ assisted inflation via tachyon condensation , nucl.phys.b614(2001)101_ [ hep - ph/0107058 ] . t. matsuda , _ f term , d term and hybrid brane inflation , jcap 0311(2003)003 _ [ hep - ph/0302078 ] ; t. matsuda , _ nontachyonic brane inflation , phys.rev.d67(2003)083519_ [ hep - ph/0302035 ] t. matsuda , _ thermal hybrid inflation in brane world , phys.rev.d68(2003)047702_ [ hep - ph/0302253 ] . n. jones , h. stoica , and s. h. h. tye , _ brane interaction as the origin of inflation , jhep 0207(2002)051 _ [ hep - th/0203163 ] ; s. sarangi , s. h. h. tye , _ cosmic string production towards the end of brane inflation , phys.lett.b536(2002)185_ [ hep - th/0204074 ] ; l. pogosian , s. h. h. tye , i. wasserman and m. wyman , _ observational constraints on cosmic string production during brane inflation , phys.rev.d68(2003)023506_ [ hep - th/0304188 ] ; m. gomez - reino and i. zavala , _ recombination of intersecting d - branes and cosmological inflation , jhep0209(2002)020 _ [ hep - th/0207278 ] . m. majumdar and a. christine - davis , _ cosmological creation of d - branes and anti - d - branes , jhep 0203(2002)056 _ [ hep - th/0202148 ] . g. dvali and a. vilenkin , _ formation and evolution of cosmic d strings _ [ hep - th/0312007 ] . e j. copeland , r. c. myers and j. polchinski , _ cosmic f and d strings _ [ hep - th/0312067 ] . e. halyo , _ cosmic d term strings as wrapped d3 branes _ [ hep - th/0312268 ] . p. binetruy , g. dvali , r. kallosh , a. van proeyen , _ fayet - iliopoulos terms in supergravity and cosmology _ [ hep - th/0402046 ] . t. matsuda , _ string production after angled brane inflation , to appear in prd _ [ hep - ph/0403092 ] . e. halyo , _ d - brane inflation on conifolds _ [ hep - th/0402155 ] ; e. halyo , _ inflation on fractional branes : d - brane inflation as d term inflation _ [ hep - th/0312042 ] ; e. halyo , _ models of inflation on d - branes _ [ hep - th/0307223 ] . dvali , _ infrared hierarchy , thermal brane inflation and superstrings as superheavy dark matter , phys.lett.b459(1999)489_ [ hep - ph/9905204 ] . g. aldazabal , s. franco , l. e. ibanez , r. rabadan , a.m. uranga , _ intersecting brane worlds , jhep 0102(2001)047 _ [ hep - ph/0011132 ] ; r. blumenhagen , d. lust , s. stieberger , _ gauge unification in supersymmetric intersecting brane worlds , jhep 0307(2003)036 _ [ hep - th/0305146 ] ; i. r. klebanov , e. witten , _ proton decay in intersecting d - brane models , nucl.phys.b664(2003)3-20_ [ hep - th/0304079 ] ; m. cvetic , p. langacker , j. wang , _ dynamical supersymmetry breaking in standard - like models with intersecting d6-branes , nucl.phys.b642(2002)139_ [ hep - th/0303208 ] ; m. cvetic , i. papadimitriou , _ more supersymmetric standard - like models from intersecting d6-branes on type iia orientifolds , phys.rev.d67(2003)126006_ [ hep - th/0303197 ] ; s.a . abel , a.w . owen , _ interactions in intersecting brane models , nucl.phys.b663(2003)197_ [ hep - th/0310257 ] ; d. bailin , g.v . kraniotis , a. love , _ standard - like models from intersecting d5-branes , phys.lett.b547(2002)43_ [ hep - th/0210219 ] ; c. f. doran , m. faux , _ intersecting branes in m - theory and chiral matter in four dimensions , jhep 0208(2002)024 _ [ hep - th/0208030 ] ; c. kokorelis , _ exact standard model compactifications from intersecting branes , jhep 0208(2002)036 _ [ hep - th/0206108 ] ; c. kokorelis , _ new standard model vacua from intersecting branes , jhep 0209(2002)029 _ [ hep - th/0205147 ] ; d. cremades , l.e . ibanez , f. marchesano , _ standard model at intersecting d5-branes : lowering the string scale , nucl.phys.b643(2002)93_ [ hep - th/0205074 ] . a. giveon and d. kutasov , _ brane dynamics and gauge theory , rev.mod.phys.71(1999)983_ [ hep - th/9802067 ] and references therein . a. hanany , m. j. strassler and a. zaffaroni , _ confinement and strings in mqcd , nucl.phys.b513(1998)87_ [ hep - th/9707244 ] ; a. hanany , m. j. strassler and a. zaffaroni , _ confinement and strings in mqcd , nucl.phys.b513(1998)87_ [ hep - th/9707244 ] . c. g. callan , jr . and j. m. maldacena , _ brane dynamics from the born - infeld action , nucl.phys.b513(1998)198_ [ hep - th/9708147 ] , k.hashimoto , _ dynamical decay of brane anti - brane and dielectric brane , jhep0207(2002)035_[hep - th/0204203 ] . see for example , mark trodden , _ electroweak baryogenesis , rev.mod.phys.71(1999)1463_ [ hep - ph/9803479 ] and references therein . j. garcia - bellido , d. yu . grigoriev , a. kusenko and m. e. shaposhnikov , _ non - equilibrium electroweak baryogenesis from preheating after inflation , phys.rev.d60(1999)123504_ [ hep - ph/9902449 ] .

this paper is a detailed version of my preprint @xcite , which was written about five years ago . here i do not discuss results that have appeared since then . i plan to survey them soon in another paper . the subject is now evolving into a _ real algebraic knot theory_. this paper is dedicated to the memory of my teacher vladimir abramovich rokhlin . it was v. a. rokhlin , who suggested to me , a long time ago , in 1977 , to develop a theory of real algebraic knots . he suggested this as a topic for my second dissertation ( after phd , like habilitation ) . following this suggestion , i moved then from knot theory and low - dimensional topology to the topology of real algebraic varieties . however , in the topology of real algebraic varieties , problems on spatial surfaces and plane curves were more pressing than problems on spatial curves , and my second dissertation defended in 1983 was devoted to the constructions of real algebraic plane curves and spatial surfaces with prescribed topology . the change in the topic occured mainly because i managed to obtain decent results in another direction , on plane curves . there was also a less respectable reason : i failed to relate the traditional techniques of classical knot theory to real algebraic knots . one of the obstacles was a phenomenon which became the initial point of this paper . a large part of the traditional techniques in knot theory uses plane knot diagrams , i.e. , projections of knots to the plane . the projection of an algebraic curve is algebraic , and one could try to apply results on plane real algebraic curves . however , the projection contains extra real points , which do not correspond to real points of the knot . these points are discussed below . in the seventies they ruined my weak attempts to study real algebraic knots . now they allow us to detect crucial differences between topological and real algebraic knots . i am grateful to alan durfee , tobias ekholm , and v. m. kharlamov for stimulating conversations . the lengthy informal introduction , which follows , is intended to explain the matter prior to going into details . i can not resist the temptation to write in the style of popular mathematics and apologize to the reader whom this style may irritate . in classical knot theory , by a link one means a smooth closed @xmath2-dimensional submanifold of the @xmath0-dimensional sphere @xmath3 , i.e. , the union of several disjoint circles smoothly embedded into @xmath3 . a link may be equipped with various additional structures such as orientation or framing and considered up to various equivalence relations like smooth ( or ambient ) isotopy , pl - isotopy , cobordism or homotopy . see , e.g. , @xcite or @xcite . in algebraic geometry classical links naturally appear as links of singular points of complex plane algebraic curves . given a singular point @xmath4 of a complex plane algebraic curve @xmath5 , the intersection of @xmath5 with the boundary of a sufficiently small ball centered at @xmath4 is called the _ link of the singularity_. it provides a base for a fruitful interaction between topology and algebraic geometry with a long history and lots of important results . another obvious opportunity for interaction between algebraic geometry and knot theory is based on the fact that a classical link may emerge as the set of real points of a real algebraic curve . this opportunity was completely ignored , besides that a number of times it was proved that any classical link is approximated by ( and hence isotopic to ) the set of real points of a real algebraic curve . there are two natural directions in which algebraic geometry and knot theory may interact in the study of real algebraic links : first , the study of relationships between invariants which are provided by link theory and algebraic geometry , second , developing a theory parallel to the classical link theory , but taking into account the algebraic nature of the objects . from the viewpoint of this second direction , it is more natural to consider real algebraic links up to isotopy consisting of real algebraic links , which belong to the same continuous family of algebraic curves , rather than up to smooth isotopy in the class of classical links . i call an isotopy of the former kind a _ rigid isotopy _ , following the terminology established by rokhlin @xcite in a similar study of real algebraic plane projective curves and the likes ( see , e.g. , the survey @xcite ) . of course , there is a forgetting functor : any real algebraic link can be regarded as a classical link and a rigid isotopy as a smooth isotopy . it is interesting to see how much is lost under that transition . in this paper i point out a real algebraic link invariant which is lost . it is unexpectedly simple . in an obvious sense it is a nontrivial vassiliev invariant of degree @xmath2 on the class of real algebraic knots ( recall that a knot is a link consisting of one component ) . in classical knot theory the lowest degree of a nontrivial vassiliev knot invariant is @xmath6 . thus there is an essential difference between classical knot theory and the theory of real algebraic knots . of course this difference has a simple topological explanation : a real algebraic link is more complicated topologically , besides its set of real points contains the set of complex points invariant under the complex conjugation and a rigid isotopy induces an equivariant smooth isotopy of this set . the invariant of real algebraic links which is defined below is very similar to the self - linking number of a framed knot . in @xcite i call it also the _ self - linking number_. its definition looks like a replacement of an elementary definition of the writhe of a knot diagram , but taking into consideration the imaginary part of the knot . here i propose to change this name ( i.e. , self - linking number ) to _ encomplexed writhe _ , and , in general , since many other characteristics can also be enhanced in a similar way , i suggest a new verb _ encomplex _ for similar enhancements by taking into consideration additional imaginary ingredients . this would agree with the general usage of the prefix _ en- _ which is described in the oxford dictionary of current english as follows : `` en- _ prefix _ @xmath7 forming verbs @xmath7 * 1 * from nouns , meaning ` put into or on ' ( _ engulf ; entrust ; embed _ ) , * 2 * from nouns or adjectives , meaning ` bring into the condition of ' ( _ enslave _ ) @xmath7 '' . complexification _ does not seem to be appropriate for what we do here with the writhe . a complexification of the writhe should be a complex counterpart for the writhe , it should be a characteristic of complex objects , while our enhancement of the writhe is defined only for real objects possessing complexification . the linking number is a well - known numerical characteristic of a pair of disjoint oriented circles embedded in three - dimensional euclidean space . roughly speaking , it measures how many times one of the circles runs around the other . it is one of the most classical topological invariants , introduced in the nineteenth century by gauss @xcite . in the classical theory , a self - linking number of a knot is defined if the knot is equipped with an additional structure like a framing or just a vector field nowhere tangent to the knot . the self - linking number is the linking number of the knot oriented somehow and its copy obtained by a small shift in the direction specified by the vector field . it does not depend on the orientation , since reversing the orientation of the knot is compensated by reversing the induced orientation of its shifted copy . of course , the self - linking number depends on the homotopy class of the vector field . a knot has no natural preferable homotopy class of framings , which would allow us to speak about a self - linking number of the knot without a special care on the choice of the framing . some framings appear naturally in geometric situations . for example , if one fixes a generic projection of a knot to a plane , the vector field of directions of the projection appears . the corresponding self - linking number is called the _ writhe _ of the knot . however , it depends on the choice of the projection and changes under isotopy . the linking number is a vassiliev invariant of order @xmath2 of two - component oriented links . this means that it changes by a constant ( in fact , by @xmath6 ) when the link experiences a homotopy with the generic appearance of an intersection point of the components . whether the linking number increases or decreases depends only on the local picture of orientations near the double point : when it passes from @xmath8{fo1.eps}}}$ ] through @xmath9{fo2.eps}}}$ ] to @xmath9{fo3.eps}}}$ ] , the linking number increases by @xmath6 . generalities on vassiliev invariants see , e.g. , in @xcite . in a sense the linking number is the only vassiliev invariant of degree @xmath2 of two - component oriented links : any vassiliev invariant of degree @xmath2 of two - component oriented links is a linear function of the linking number . similarly , the self - linking number is a vassiliev invariant of degree @xmath2 of framed knots ( it changes by @xmath6 when the knot experiences a homotopy with a generic appearance of a self - intersection point ) and it is the only vassiliev of degree @xmath2 of framed knots in the same sense . the necessity of a framing for the definition of self - linking number can now be formulated more rigorously : only constants are vassiliev invariants of degree @xmath2 of ( non - framed ) knots . the diagrammatical definition of the writhe , which is imitated below , runs as follows : for each crossing point of the knot projection , one defines a _ local writhe _ equal to @xmath10 if near the point the knot diagram looks like @xmath9{fo3.eps}}}$ ] and @xmath1 if it looks like @xmath8{fo1.eps}}}$ ] . then one sums the local writhes over all double points of the projection . the sum is the writhe . a continuous change of the projection may cause the vanishing of a crossing point . this happens under the first reidemeister move shown in the left hand half of figure [ f1 ] . this move changes the writhe by @xmath11 . if a link is algebraic , then its projection to a plane is algebraic , too . a generic projection has only ordinary double points and the total number of its complex double points is constant . the number of real double points can vary , but only by an even number . a real double point can not turn alone into an imaginary one , as it seems to happen under the first reidemeister move . under an algebraic version of the first reidemeister move , the double point stays in the real domain , but becomes solitary , like the only real point of the curve @xmath12 . the algebraic version of the first reidemeister move is shown in the right hand half of figure [ f1 ] . it is not difficult to prove that the family of spatial curves that realizes this move can be transformed by a local diffeomorphism to the family of affine curves defined by the following system of equations @xmath13 where @xmath14 is the parameter of the deformation . these are rational curves , admitting a rational parametrization @xmath15 the projection corresponds to the standard projection @xmath16 to the coordinate @xmath17-plane . it maps these curves to the family of affine plane rational cubic curves defined by @xmath18 with @xmath19 . a solitary double point of the projection is not the image of any real point of the link . it is the image of two imaginary complex conjugate points of the complexification of the link . the preimage of the point in the 3-space under the projection is a real line . it is disjoint from the real part of the link , but intersects its complexification in a couple of complex conjugate imaginary points . in the model of the first reidemeister move above , @xmath20 is the double point of the projection for each @xmath21 . if @xmath22 , it is a usual crossing point . its preimage consists of two real points @xmath23 and @xmath24 . if @xmath25 , it is a solitary double point . its preimage consists of two imaginary conjugate points @xmath26 and @xmath27 , which lie on a real line @xmath28 below , in section [ s0.3 ] , with any solitary double point of the projection , a local writhe equal to @xmath29 is associated . this is done in such a way that the local writhe of the crossing point vanishing in the first reidemeister move is equal to the local writhe of the new - born solitary double point . in the case of an algebraic knot , the sum of local writhes of all double points , both solitary and crossings , does not depend on the choice of projection and is invariant under rigid isotopy . this sum is the encomplexed writhe . a construction similar to the construction of the encomplexed writhe number of an algebraic knot can be applied to an algebraic _ link_. however in this case there are two versions of the construction . in the first of these , we define an encomplexed writhe number generalizing the encomplexed writhe number defined above for knots . we consider a link diagram and the sum of local writhes at solitary double points and crossing points where the branches belong the same connected component of the set of real points . at these crossing point , to define a local writhe , we need orientations of the branches . as above , we choose an orientation on each of the components . if we make another choice , at a crossing point for which the branches belong the same component , either both orientations change or none . hence the local writhe numbers at crossing points of this kind do not depend on the choice . in section [ s0.1 ] below , we prove that the whole sum of local writhes over crossing points of this kind and solitary double points does not depend on the projection and is invariant under rigid isotopy . we call this sum the _ encomplexed writhe number _ of the link @xmath30 and denote by @xmath31 . in the second version of the construction , we consider a real algebraic link which is equipped with an orientation of the set of real points , use these orientations to define local writhe numbers at all crossing points and sum the local writhe numbers over all crossing points and all solitary double points . the result is called the _ encomplexed writhe number of an oriented real algebraic link_. this encomplexed writhe number does not change when the orientation reverses . an orientation considered up to reversing is called a _ semi - orientation_. thus the encomplexed writhe number depends only on the semi - orientation of the link . the ( semi-)orientation may be an artificial extra structure , but it may also appear in a natural way , say , as a complex orientation , if the set of real points divides the set of real points , see @xcite . in fact , the complex orientation is defined up to reversing , so it is indeed a semi - orientation . another important class of semi - oriented algebraic links appears as transversal intersections of two real algebraic surfaces of degrees @xmath4 and @xmath32 with @xmath33 . the encomplexed writhe number of ( semi-)oriented real algebraic link differs from the encomplexed writhe number of the same link without orientation by the sum of all pairwise linking numbers of the components multiplied by @xmath6 : let @xmath30 be a real algebraic link , let @xmath34 be the same link equipped with an orientation of its set of real points and @xmath35 the ( oriented ) connected components of this set , then @xmath36 in the case of a knot , the encomplexed writhe number defines a natural class of framings , since homotopy classes of framings are enumerated by their self - linking numbers and we can choose the framing having the self - linking number equal to the algebraic encomplexed writhe number . i do not know any direct construction of this framing . moreover , there seems to be a reason for the absence of such a construction . in the case of links , the construction above gives a single number , while framings are enumerated by sequences of numbers with entries corresponding to components . let @xmath30 be a nonsingular real algebraic curve in @xmath0-dimensional projective space . then the set @xmath37 of its real points is a smooth closed @xmath2-dimensional submanifold of @xmath38 , i.e. , a smooth projective link . the set @xmath39 of its complex points is a smooth complex @xmath2-dimensional submanifold of @xmath40 . let @xmath41 be a point of @xmath38 . consider the projection @xmath42 from @xmath41 . assume that @xmath41 is such that the restriction to @xmath39 of @xmath43 is generic . this means that it is an immersion without triple points and at each double point the images of the branches have distinct tangent lines . it follows from well - known theorems that those @xmath41 s for which this is the case form an open dense subset of @xmath38 ( in fact , it is the complement of a @xmath6-dimensional subvariety ) . the real part @xmath44 of the image consists of the image @xmath45 of the real part and , maybe , several solitary points , which are double points of @xmath46 . there is a purely topological construction which assigns a local writhe equal to @xmath29 to a crossing belonging to the image of only one component of @xmath37 . this construction is well - known in the case of classical knots . here is its projective version . i borrow it from drobotukhina s paper @xcite on the generalization of kauffman brackets to links in projective space . , @xmath47 , @xmath48 . ] let @xmath49 be a smooth connected one - dimensional submanifold of @xmath38 , and @xmath41 be a point of @xmath50 . let @xmath51 be a generic double point of the projection @xmath52 and @xmath53 be the line which is the preimage of @xmath51 under the projection . denote by @xmath54 and @xmath55 the points of @xmath56 . the points @xmath54 and @xmath55 divide the line @xmath57 into two segments . choose one of them and denote it by @xmath58 . choose an orientation of @xmath49 . let @xmath59 and @xmath60 be tangent vectors of @xmath49 at @xmath54 and @xmath55 respectively directed along the selected orientation of @xmath49 . let @xmath47 be a vector tangent to @xmath57 at @xmath54 and directed inside @xmath58 . let @xmath48 be a vector at @xmath54 such that it is tangent to the plane containing @xmath57 and @xmath60 and is directed to the same side of @xmath58 as @xmath60 ( in an affine part of the plane containing @xmath58 and @xmath60 ) . see figure [ f2 ] . the triple @xmath59 , @xmath47 , @xmath48 is a base of the tangent space @xmath61 . define the local writhe of @xmath51 to be the value taken by the orientation of @xmath38 on this frame . the construction of the local writhe of @xmath51 contains several choices . here is a proof that the result does not depend on them . we have chosen an orientation of @xmath49 . had the opposite orientation been selected , then @xmath59 and @xmath48 would be replaced by the opposite vectors @xmath62 and @xmath63 . this would not change the result , since @xmath62 , @xmath47 , @xmath63 defines the same orientation as @xmath59 , @xmath47 , @xmath48 . we have chosen the segment @xmath58 . if the other half of @xmath57 was selected , then @xmath47 and @xmath48 would be replaced by the opposite vectors . but @xmath59 , @xmath64 , @xmath63 defines the same orientation as @xmath59 , @xmath47 , @xmath48 . the construction depends on the order of points @xmath54 and @xmath55 . the other choice ( with the same choice of the orientation of @xmath49 and segment @xmath58 ) gives a triple of vectors at @xmath55 . it can be moved continuously without degeneration along @xmath58 into the triple @xmath48 , @xmath64 , @xmath59 , which defines the same orientation as @xmath59 , @xmath47 , @xmath48 . let @xmath30 , @xmath41 , and @xmath43 be as in the beginning of section [ s0.1 ] and let @xmath65 be a solitary double point of @xmath43 . here is a construction assigning @xmath29 to @xmath66 . i will also call the result a _ local writhe _ of @xmath66 . denote the preimage of @xmath66 under @xmath43 by @xmath57 . this is a real line in @xmath38 connecting @xmath41 and @xmath66 . it intersects @xmath39 in two imaginary complex conjugate points , say , @xmath54 and @xmath55 . since @xmath54 and @xmath55 are conjugate , they belong to different components of @xmath67 . choose one of the common points of @xmath39 and @xmath68 , say , @xmath54 . the natural orientation of the component of @xmath67 defined by the complex structure of @xmath68 induces an orientation on @xmath69 as on the boundary of its closure . the image under @xmath43 of the local branch of @xmath39 passing through @xmath54 intersects the plane of the projection @xmath70 transversally at @xmath66 . take the local orientation of the plane of projection such that the local intersection number of the plane and the image of the branch of @xmath39 is @xmath10 . thus the choice of one of two points of @xmath71 defines an orientation of @xmath69 and a local orientation of the plane of projection @xmath70 ( we can speak only of a local orientation of @xmath70 , since the whole @xmath70 is not orientable ) . the plane of projection intersects . if you would like to think on it as on the set of lines of @xmath38 passing through @xmath41 , please , identify it in a natural way with any real projective plane contained in @xmath38 and disjoint from @xmath41 . all such embeddings @xmath72 are isotopic . ] transversally @xmath69 in @xmath66 . the local orientation of the plane , the orientation of @xmath69 and the orientation of the ambient @xmath38 determine the intersection number . this is the local writhe . it does not depend on the choice of @xmath54 . indeed , if one chooses @xmath55 instead , then both the orientation of @xmath69 and the local orientation of @xmath70 would be reversed . the orientation of @xmath73 would be reversed , because @xmath69 inherits opposite orientations from the different halves of @xmath67 . the local orientation of @xmath70 would be reversed , because the complex conjugation involution @xmath74 preserves the complex orientation of @xmath75 , preserves @xmath70 ( point - wise ) and maps one of the branches of @xmath46 at @xmath66 to the other reversing its complex orientation . now for any real algebraic projective link @xmath30 , choose a point @xmath76 such that the projection of @xmath30 from @xmath41 is generic and sum the writhes of all crossing points of the projection belonging to the image of only one component of @xmath37 and the writhes of all solitary double points . this sum is called the _ encomplexed writhe number of @xmath30_. i have to show that it does not depend on the choice of projection . the proof given below proves more : the sum is invariant under _ rigid isotopy _ of @xmath30 . by rigid isotopy we mean an isotopy consisting of nonsingular real algebraic curves . the effect of a movement of @xmath41 on the projection can be achieved by a rigid isotopy defined by a path in the group of projective transformations of @xmath38 . therefore the following theorem implies both the independence of the encomplexed writhe number from the choice of projection and its invariance under rigid isotopy . [ mainth ] for any two rigidly isotopic real algebraic projective links @xmath77 and @xmath78 whose projections from the same point @xmath76 are generic , the encomplexed writhe numbers of @xmath77 and @xmath78 defined via @xmath41 are equal . this theorem is proved in section [ s0.6 ] . [ cor1 ] the encomplexed writhe number of a real algebraic projective link does not depend on the choice of the projection involved in its definition . a projection depends only on the center from which it is done . the effect on the projection of a movement of the center can be achieved by a rigid isotopy defined by a path in the group of projective transformations of @xmath38 . thus the encomplexed writhe number is a characteristic of a real algebraic link . [ cor2 ] the encomplexed writhe number of a real algebraic projective link is invariant under rigid isotopy . as in the purely topological situation of an isotopy of a classical link , a generic rigid isotopy of a real algebraic link may be decomposed into a composition of rigid isotopies , each of which involves a single local standard move of the projection . there are @xmath79 local standard moves . they are similar to the reidemeister moves . the first of these @xmath79 moves is shown in the right hand half of figure [ f1 ] . the other moves are shown in figure [ o23 ] . the first two of these coincide with the second and third reidemeister moves . the fourth move is similar to the second reidemeister move : also two double points of projection come to each other and disappear . however the double points are solitary . the fifth move is similar to the third reidemeister move : a triple point also appears for a moment . but at this triple point only one branch is real , the other two are imaginary conjugate to each other . in this move a solitary double point traverses a real branch . to prove theorem [ mainth ] , first replace the rigid isotopy by a generic one and then decompose the latter into local moves described above , in section [ s0.5 ] . only in the first , fourth and fifth moves solitary double points are involved . the invariance under the second and the third move follows from the well - known fact of knot theory that the topological writhe is invariant under the second and third reidemeister moves . thus the following three lemmas imply theorem [ mainth ] . [ lem1 ] in the fifth move the writhe of the solitary point does not change . [ lem2 ] in the fourth move the writhes of the vanishing solitary points are opposite . [ lem3 ] in the first move the writhe of vanishing crossing point is equal to the writhe of the new - born solitary point . proof of lemma [ lem3 ] is postponed to section [ s0.9 ] . note that although lemma [ lem3 ] is the most difficult to prove , it is the least significant : here its only role is to justify the choice of sign made in the definition of local writhe in solitary double point of the projection . it is clear that the writhes of vanishing double points involved in the first move are related , and if they were opposite to each other , then the definition of the encomplexed writhe number should be changed , but would not be destroyed irrecoverably . this is obvious . indeed , the real branch of the projection does not interact with the imaginary branches , it just passes through their intersection point . at the moment of the fourth move take a small ball @xmath80 in the complex projective plane centered in the solitary self - tangency point of the projection of the curve . its intersection with the projection of the complex point set of the curve consists of two smoothly embedded disks tangent to each other and to the disk @xmath81 . under the move each of the disks experiences a diffeotopy . before and after the move the intersection the curve with @xmath80 is the union of the two disks meeting each other transversally in two points , but before the move the disks do not intersect @xmath70 , while after the move they intersect @xmath70 in their common points . to calculate the writhe at both vanishing solitary double points , let us select the same imaginary branch of the projection of the curve passing through the points . this means that we select one of the disks described above . the sum of the local intersection numbers of this disk ( equipped with the complex orientation ) and @xmath81 ( equipped with some orientation ) is zero since under the fourth move the intersection disappears , while in the boundary of @xmath80 no intersection happens . therefore the local orientations of the projective plane in the vanishing solitary double points defined by this branch define opposite orientations of @xmath81 . ( recall that the local orientations are distinguished by the condition that the local intersection numbers are positive . ) on the other hand , under the move the preimages of the vanishing solitary double points come to each other up to coincidence at the moment of the move and their orientations defined by the choice of the same imaginary branch are carried to the same orientation of the preimage of the point of solitary self - tangency . indeed , the preimages are real lines and points of intersection of their complexifications with the selected imaginary branch of the curve also come to the same position . therefore the halves of the complexifications containing the points come to coincidence , as well as the orientations defined by the halves on the real lines . it follows that the intersection numbers of @xmath80 with the preimages of the vanishing solitary double points equipped with these orientations are equal . since the local orientations of the projective plane in the vanishing solitary double points define distinct orientations of @xmath81 , the writhes are opposite to each other . it is sufficient to consider the model family of curves described in section [ si3 ] . recall that the curves of this family are defined by the following system of equations @xmath13 where @xmath14 is the parameter of the deformation . these curves admit a rational parametrization @xmath82 the projection corresponds to the standard projection @xmath16 to the coordinate @xmath17-plane . it maps these curves to the family of affine plane rational cubic curves defined by @xmath18 with @xmath19 . we must prove that the local writhe at @xmath20 for @xmath22 coincides with the local writhe at @xmath20 for @xmath25 . let us calculate the local writhe for @xmath22 . denote @xmath83 by @xmath84 . the preimage of @xmath20 consists of points @xmath85 and @xmath86 corresponding to the values @xmath87 and @xmath84 of @xmath88 , respectively , see figure [ f3 ] . the tangent vectors to the curve at these points are @xmath89 and @xmath90 . the vector @xmath47 connecting @xmath54 and @xmath55 is @xmath91 . by definition , the writhe is the value taken by the orientation of @xmath92 on the frame @xmath59 , @xmath47 , @xmath48 . this value is equal to the value of this orientation on the frame @xmath93 , @xmath94 , @xmath95 multiplied by the sign of @xmath96 let us calculate the local writhe for @xmath25 . denote @xmath97 by @xmath84 . the preimage of @xmath20 consists of points @xmath98 and @xmath99 corresponding to the values @xmath100 and @xmath101 of @xmath88 . choose the branch which passes through @xmath102 . it belongs to the upper half of the line @xmath28 , which induces the positive orientation of the real part directed along @xmath95 . at @xmath102 the branch of the curve has tangent vector @xmath103 and the real basis consisting of @xmath59 and @xmath104 positively oriented with respect to the complex orientation of this branch . the projection maps this basis to the positively oriented basis @xmath105 , @xmath106 of the projection of the branch . the intersection number of this projection and @xmath107 in @xmath108 is the sign of @xmath109 hence the orientation of @xmath107 such that its local intersection number with the selected branch of the projection does not coincide with the orientation defined by the standard basis . the intersection number of the line @xmath28 with the standard orientation and the @xmath17-plane with the standard orientation is the value of the orientation of the ambient space @xmath110 taken on the standard basis @xmath93 , @xmath94 , @xmath95 . therefore the local writhe is opposite to this value . [ rem ] there is a more conceptual proof of lemma [ lem3 ] . it is based on a local version of the rokhlin complex orientation formula , see @xcite and @xcite . in fact , the original proof was done in that way . however , the complex orientation formula is more complicated than the calculation above . to speak about vassiliev invariants , we need to fix a connected family of curves , in which links under consideration comprise the complement to a hypersurface . in the case of classical knots one could include all knots in such a family by adjoining knots with self - intersections and other singularities . a singular knot is a right equivalence class of a smooth map of the circle to the space ( recall that two maps from a circle are right equivalent if one of them is a composition of a self - diffeomorphism of the circle with the other one ) . in the case of real algebraic knots , such a family including all real algebraic knots does not exist . even the space of complex curves in the three - dimensional projective space consists of infinitely many components . it is impossible to change the homology class realized by the set of complex points of an algebraic curve in @xmath40 by a continuous deformation . recall that the homology class belongs to the group @xmath111 and is a positive multiple @xmath112 $ ] of the natural generator of @xmath113\in h_2(\c p^3)$ ] realized by a line . the coefficient @xmath114 is called the _ order _ of the curve . the genus is another numerical characteristic of a complex curve which takes the same value for all nonsingular curves in any irreducible family . as is well known , the nonsingular complex curves of given order and genus in three - dimensional projective space are parametrized by a finite union of quasi - projective varieties . for each of these varieties , one can try to build a separate theory of vassiliev invariants on a class of nonsingular real algebraic curves whose complexifications are parametrized by points of this variety . ( a similar phenomenon takes place in topology : links with different numbers of components can not be included into a reasonable connected family , and therefore for each number of components there is a separate theory of vassiliev invariants . ) among the varieties of algebraic curves in three - dimensional projective space , there are two special families : for each natural number @xmath114 there is an irreducible variety of rational curves of order @xmath114 ( recall that a an algebraic curve is called rational if it admits an algebraic parametrization by a line ) , and for each pair of natural numbers @xmath4 and @xmath32 there is an irreducible variety of curves which can be presented as intersection of surfaces of degrees @xmath4 and @xmath32 . in the class of real algebraic rational curves of order @xmath114 , singular curves comprise a _ discriminant hypersurface _ in which a generic point is a rational curve such that it has exactly one singular point and this point is an ordinary double point . an ordinary double point may be of one of the following two types : either it is an intersection point of two real branches , or two imaginary conjugate branches . any two real algebraic rational nonsingular curves of order @xmath114 can be connected by a path in the space of real rational curves of degree @xmath114 that intersects the discriminant hypersurface only transversally at a finite number of generic points . such a path can be regarded as a deformation of a curve to the other one . when it intersects the discriminant hypersurface at a point , which is a curve with singularity on real branches , the set of real points of the curve behaves as in classical knot theory : two pieces of the set of real points come to each other and pass through each other . as in classical knot theory , at the moment of intersection , the generic projection of the curve experiences an isotopy . nothing happens besides that one crossing point becomes for a moment the image of a double point and then changes back into a crossing point , but with the opposite writhe . when the path intersects the discriminant hypersurface at a point , which is a curve with singularity on imaginary branches , two complex conjugate imaginary branches pass through each other . at the moment of passing , they intersect in a real isolated double point . at this moment the set of real points of a generic projection experiences an isotopy . no event happens besides that a solitary double point becomes for a moment the image of a solitary real double point of the curve and then changes back into an ordinary solitary double point of the projection ( which is not the image of a real point of the knot ) , but with the opposite writhe number . it is clear that the encomplexed writhe number of an algebraic curve changes under a modification of each of these kinds by @xmath115 , with the sign depending only on the local structure of the modification near the double point . this means that _ the encomplexed writhe number on the family of real rational curves under consideration is a vassiliev invariant of degree @xmath2_. this is true also for any space of nonsingular real algebraic curves that can be included into a connected family of real algebraic curves by adjoining a hypersurface , penetration through which at a generic point looks as in the family of rational curves described above . there are many families of this kind besides the families of rational knots . however , in many families of algebraic curves a transversal penetration through the discriminant hypersurface in a generic point looks differently . in particular , for intersections of two surfaces it is a morse modification of the real part of the curve . at the moment , the old double points of the projection , both solitary and crossing , do not change . an additional double point appears just for a moment . however the division of crossing points to self - crossing points of a single component and crossing points of different components may change . therefore the encomplexed writhe number changes in a complicated way . if the degrees of the surfaces defining the curve are of the same parity , the real part of the curve has a natural semi - orientation . the morse modification respects this semi - orientation . therefore the encomplexed writhe number of the semi - oriented curve does not change . [ vanishsl ] the encomplexed writhe number of any nonsingular semi - oriented real algebraic link which is a transversal intersection of two real algebraic surfaces whose degrees are of the same parity is zero . any two nonsingular real curves of the type under consideration can be connected by a path as above . hence their self - linking numbers coincide . on the other hand , it is easy to construct , for any pair of natural numbers @xmath4 and @xmath32 of the same parity , a pair of nonsingular real algebraic surfaces of degrees @xmath4 and @xmath32 transversal to each other in three - dimensional projective space such that their intersection has zero self - linking number . in contrast to this vanishing result , one can prove that the _ encomplexed writhe number of a real algebraic rational knots of degree @xmath114 can take any value in the interval between @xmath116 and @xmath117 including these limits and congruent to them modulo @xmath6_. the same construction may be applied to real algebraic curves in @xmath38 having singular imaginary points , but no real singularities . in the construction we can eliminate projections from the points such that some singular point is projected from them to a real point . indeed , for any imaginary point there exists only one real line passing through it ( the line connecting the point with its complex conjugate ) , thus we have to exclude a finite number of real lines . this gives a generalization of encomplexed writhe numbers with the same properties : it is invariant with respect to rigid isotopies ( i.e. , isotopies made of curves from this class ) , and is multiplied by @xmath1 under a mirror reflection . the construction of this paper can be applied to algebraic links in the sphere @xmath3 . although from the viewpoint of knot theory this is the most classical case , from the viewpoint of algebraic geometry the case of curves in the projective space is simpler . the three - dimensional sphere @xmath3 is a real algebraic variety . it is a quadric in four - dimensional real affine space . the stereographic projection is a birational isomorphism of @xmath3 onto @xmath38 . it defines a diffeomorphism between the complement of the center of the projection in @xmath3 and a real affine space . given a real algebraic link in @xmath3 , one may choose a real point of @xmath3 from the complement of the link and project the link from this point to an affine space . then include the affine space into the projective space and apply the construction above . the image has no real singular points , therefore we can use the result of the previous section . this construction blows up the center of projection , making a real projective plane out of it , and maps the complement to the center of the projection in the set of real points of the sphere isomorphically onto the complement of the projective plane . in the imaginary domain , it contracts each generatrix of the cone which is the intersection of the sphere with its tangent plane at the center of projection . the image of the cone is an imaginary quadric curve contained in the projective plane which appeared as the result of blowing up of the central point . first , consider the most straightforward generalization . let @xmath57 be a nonsingular real algebraic @xmath118-dimensional subvariety in the projective space of dimension @xmath119 . its generic projection to @xmath120 has only ordinary double points . at each double point either both branches of the image are real or they are imaginary complex conjugate . if the set of real points is orientable , then one can repeat everything with obvious changes and obtain a definition of a numeric invariant generalizing the encomplexed writhe number defined above . let @xmath121 be a nonsingular three - dimensional real algebraic variety with oriented set of real points equipped with a real algebraic fibration over a real algebraic surface @xmath122 with fiber a projective line . there is a construction which assigns to a real algebraic link ( i.e. , a nonsingular real algebraic curve in @xmath121 ) with a generic projection to @xmath122 an integer , which is invariant under rigid isotopy , is multiplied by @xmath1 under the orientation reversal in @xmath121 and is a vassiliev invariant of degree @xmath2 . this construction is similar to the one presented above , but uses , instead of the projection to @xmath70 , an algebraic version of turaev s shadow descriptions of links @xcite . here we discuss only one example . however it can be easily generalized . consider immersions of the sphere @xmath123 to @xmath124 . up to regular homotopy ( i.e. , a homotopy consisting of immersions whose differentials also comprise a homotopy ) , an immersion @xmath125 is defined by its smale invariant @xcite , which is an element of @xmath126 . for a generic immersion , it can be expressed as the sum of local self - intersection numbers over all double points of the immersion , see @xcite . let us encomplex the smale invariant . for this , first , we have to consider a real algebraic counterpart for the notion of generic immersion @xmath125 . the identification is defined via the universal covering @xmath127 . replace euclidean space @xmath124 by torus @xmath128 , which has the advantage of being compact . the classification of immersions @xmath129 up to regular homotopy coincides with the smale classification of immersions @xmath125 . the sphere @xmath123 is the real part of a quadric projective hypersurface . the torus @xmath128 is the real part of a complex abelian variety . consider real regular maps of the quadric to the abelian variety . a generic map defines an immersion both for the complex and real parts . the only singularities are transversal double points . double points in the real part of the target variety are of two kinds . at a double point of the first kind two sheets of the image of @xmath123 meet . at a double point of the second kind the images two complex conjugate sheets of the complexification of @xmath123 meet . the smale invariant is the sum of the local intersection numbers over the double points of the first kind . one can extend the definition of the local intersection number to the double points of the second kind in such a way that the total sum of the local intersection numbers over double points of both kinds would be invariant under continuous deformations of regular maps . this total sum is the _ encomplexed smale invariant_. notice that it is , in a sense , more invariant than the original smale invariant . the smale invariant may change under homotopy , it is invariant only under regular homotopy . the encomplexed smale invariant does not change under a homotopy in the class of regular maps , which corresponds to the class of all continuous maps . drobotukhina , _ an analogue of the jones polynomial for links in @xmath130 and a generalization of the kauffman murasugi theorem _ , algebra i analiz * 2 * ( 1990 ) , no . 3 , 171191 ; english transl . leningrad math . j. * 2 * ( 1991 ) , no . 3 , 613630 . o. ya . viro , _ progress in the topology of real algebraic varieties over the last six years _ , uspekhi mat . nauk * 41 * ( 1986 ) , no . 3 , 4567 ; english transl russian math . surveys * 41 * ( 1986 ) , no . 3 , 5582 .

gamma - ray bursts are the most powerful explosions in the universe . ( for reviews see piran 1999 and mszros 2002 . ) a crucial advance in understanding gamma - ray bursts began with the discovery of `` afterglows '' , starting with _ bepposax _ observations in the soft x - ray band of grb 970228 ( costa et al . 1997 ; wijers , rees , & mszros 1997 ) . ( for a review of grb afterglows see van paradijs , kouveliotou , & wijers 2000 . ) if grbs were isotropic , then the measured redshifts would imply total explosion energies of @xmath5 ergs ( frail et al . . theoretical work on relativistic jet expansion , however , shows that one expects a steepening in the decay light curve if one is looking down the axis of a jet as the flow decelerates from a bulk lorentz factor @xmath6 to @xmath7 , where @xmath1 is the jet beaming angle ( e.g. , rhoads 1997 , sari , piran , & halpern 1999 , panaitescu & mszros 1999 , panaitescu & kumar 2001ab , 2002 ) . one does in fact see such steepenings in the light curves ( e.g. , stanek et al . 1999 ; harrison et al . prior to the time when @xmath8 the expansion is effectively `` spherical '' from the observer s viewpoint because the relativistic beaming is narrower than the jet itself . in other words , if the grb emission were coming from one spot on a large , relativistically expanding sphere , aimed directly at the observer , the observer would not see any emission from the other parts of the sphere . after the time when @xmath8 the observer can `` see '' the entire jet , and a faster rate of decline in the luminosity is predicted . a separate issue that we will address in this work is the sideways or lateral expansion of the jet as the increasing solid angle of the jet enables a larger fraction of the circumstellar medium ( csm ) surrounding the progenitor star to be intercepted and provide decelerating gas . several groups have claimed that this leads to a faster ( exponential ) decrease in @xmath2 , which acts as an additional agent to diminish the amplitude of the relativistic beaming . the concept of a `` break '' in the afterglow light curves occurring when @xmath8 has been used to infer the presence of strong beaming in grbs ( frail et al . 2001 ; panaitescu & kumar 2001ab , 2002 ; panaitescu , mszros , & rees 1998 ; piran , kumar , panaitescu , & piro 2001 ) . frail et al . ( 2001 ) utilize the theoretical framework of sari , piran , & halpern ( 1999 ) which takes the jet evolution to be spherical adiabatic expansion to show that , after correcting the `` isotropic '' energies to account for the specific beaming factor for each burst , the total burst energy reduces to a narrow range centered on @xmath9 ergs . frail et al . ( 2001 ) give a tabulation of 17 afterglows for which redshifts had been measured , up to 2001 january . the redshifts range from 0.433 to 4.5 . of these , 10 also have break times known to within @xmath10% , ranging from 1 d to 25 d. ( in addition , three grbs have listed lower limits on the break time , and two have listed upper limits . ) the combination of redshifts , fluences , and break times lead to estimates of the jet angles @xmath1 ranging from @xmath11 to @xmath12 , with an aggregation near @xmath13 . ( the number of grb s with redshift determinations is currently @xmath14 . for an update of the work described in frail et al . 2001 , see berger , kulkarni , & frail 2003 . ) prior investigations of grb jet expansion have been largely analytical , or involved computational models with some imposed symmetry , typically spherical or axial ( e.g. , rhoads 1997 , 1999 ; sari 1997 , sari , piran , & halpern 1999 ; panaitescu & mszros 1999 ; granot , piran , & sari 1999a ; kumar & panaitescu 2000ab ; panaitescu & kumar 2001ab , 2002 ; granot et al . these works have tended for the most part to confirm estimates and scalings based on the analytical formalism of blandford & mckee ( 1976=bm76 ) , although for many of the studies the agreement is ( to some extent ) circular , given that they rely at least in part on the bm76 formalism . kobayashi , piran , & sari ( 1999 ) present results based on a spherically symmetric relativistic lagrangian code . they identify three regimes during the evolution of the grb jet : ( i ) an acceleration phase during which @xmath15 and @xmath2 increases rapidly to a large value @xmath16 , ( ii ) a coasting phase during which @xmath2 is relatively constant as the mass accumulated from the csm is small compared to that in the jet , and ( iii ) a deceleration phase during which the accumulated mass forces a rapid decrease in @xmath2 . previous studies using analytical methods divided the evolution into regimes defined by some ordering of distance , velocity , or energy scales ( see table 3 of piran 1999 and figure 7 of mszros 2002 ) . the main findings that seem to be common to all studies are that the break in the decaying light curve @xmath17 occurs roughly when the deceleration has decreased @xmath2 to roughly the reciprocal of the jet beaming angle , and that the subsequent decrease in @xmath2 is roughly exponential with distance . also , panaitescu & mszros ( 1999 ) present axisymmetric calculations to study the combined effects of the transition from @xmath6 to @xmath7 and the lateral jet expansion . in other subdisciplines of astronomy the use of relativistic hydrodynamics codes has been standard for some time . for example , workers studying extragalactic jets have used such codes to continuously inject a collimated supersonic beam into a surrounding medium , usually under the assumption of pressure equilibrium . ( see , e.g. , norman et al . 1982 for a thorough discussion ) . in the context of grbs , work has been done using 2d and 3d relativistic hydro codes to consider the evolution of the grb jet as it propagates through the envelope of the progenitor star , up to the point where it breaks out of the stellar surface and produces the prompt grb emission ( zhang , woosley , & macfadyen 2003 , zhang , woosley , & heger 2003 ) . in this work we consider the evolution covering the afterglow time ( i.e , after the period considered by zhang et al . we utilize a three dimensional relativistic hydrodynamical code to study the propagation of an initially ultrarelativistic blob into a dense csm . we study the spatial spreading of the blob both along the direction of propagation and orthogonal to it , as well as the evolution of @xmath2 in space and time . we also calculate afterglow light curves , taking a simple prescription in which the local emissivity scales with the local pressure . the model used is that of del zanna & bucciantini ( 2002=db02 ) . these authors present a simple and efficient numerical scheme for special relativistic hydrodynamics ( srhd ) that does not rely on computationally expensive spectral decomposition ( and the accompanying matrix inversions ) . the semi - discrete form of the srhd equations is solved , so that time integration can be carried out using a standard runge - kutta method . unlike many previous multidimensional calculations run on supercomputers , characteristic decomposition and riemann solvers are not required . the only local pieces of information required are the highest characteristic speeds . the simplicity and efficiency of the model allow for runs to be carried out on a modern pc with @xmath18 ghz cpu speed and @xmath19 gb ram . we refer the reader to db02 and to londrillo & del zanna ( 2000 ) for the details . ( also , del zanna & bucciantini 2003 add mhd to the srhd formalism of db02 . ) the basic idea is to calculate a vector of `` fluxes '' * f * that is then used within each time step to advance a vector of `` conserved '' variables * u * by computing the spatial divergences of the * f * s . the time integration is carried out with a third order runge - kutta scheme . the vector of `` primitive '' or physical variables @xmath20^t$ ] must be recovered from the conserved variables @xmath21^t$ ] in each runge - kutta step , and at each grid point . primitive variables are reconstructed at the left ( l ) and right ( r ) cell faces using the `` convex essentially non - oscillatory '' interpolation described in db02 . this method gives third order accuracy by adaptively adjusting the `` stencil '' , or spatial interval , used for computing differences in the vicinity of a given cell . db02 present two versions of the fluxes , * f*@xmath22 , where `` hll '' denotes harten , lax , & van leer ( 1983 ) , that are precise but at the same time potentially prone to numerical instabilities , and * f*@xmath23 , where `` llf '' denotes local lax - friedrichs ( cf . lax & liu 1998 ) , that represent a smoother , numerically dissipative flux . for the main application we consider in this work we will utilize the llf fluxes . db02 present a suite of test results for 1 , 2 , and 3 dimensions . we have reproduced these tests , and now show the results of a 3d spherical expansion that results in higher @xmath2 values than the tests discussed in db02 . we utilize a cubical grid of @xmath24 points , with each side having length unity . minutes per runge - kutta integration for a 3d problem with @xmath25 nodes , utilizing the third order reconstruction for the l and r values of the primitives , and using a 1 ghz linux pc ; our run time of @xmath26 s is consistent with the 1.7 ghz linux pc used in this work . ] initial conditions are that everywhere @xmath27 , @xmath28 , and @xmath29 within a radius 0.1 of the center of the cube , and @xmath30 outside . this is the 3d analog of the 1d `` piston '' problem . ( an identical test with less extreme initial conditions is shown and discussed by hughes , miller , & duncan 2002 , see their fig . 3 . ) the enormous over - pressure launches a relativistic blast wave at @xmath31 . figure 1 shows the evolution of @xmath32 , @xmath33 , @xmath34 , and @xmath2 for a slice taken along the @xmath35axis . the density inside the sphere becomes small for late times . to avoid numerical instabilities in this region we utilize the smoother llf fluxes for small densities ( @xmath36 ) , and the hll fluxes elsewhere . we show the initial configuration plus 10 time slices from a run with 360 total time steps , taking a courant number of @xmath37 . after the run begins one can see the development of a strong spherical outflow from the center of the cube . the lorentz @xmath2 factor has increased to @xmath38 in the outer parts of the expanding shell by the end of the run . the run was halted before the expansion reached the edge of the grid . for the last three times steps , a kink develops in the @xmath33 profile at the center of the sphere because of the prolonged spherical outflow from localized point . figure 2 shows the corresponding evolution for pure llf fluxes . all variables now show a smoother evolution and @xmath39 at large radii for late times @xmath40 somewhat less than for the hll fluxes . conservation of rest - mass energy @xmath41 and energy @xmath42 ( where the specific energy @xmath43 ) integrated over the computational grid is as follows : by the end of the run shown in fig . 1 , @xmath44 and @xmath45 have increased by about 1 part in @xmath46 ; by the end of the run shown in fig . 2 , @xmath44 has increased by @xmath473 parts in @xmath48 , and @xmath45 is constant ( to within machine accuracy ) . we now present results for the case of interest , namely a relativistic blob of material that expands roughly axisymmetrically ( orthogonal to the direction of motion ) as it decelerates from ultrarelativistic to relativistic speeds . we set up our initial problem as follows : our blob is made to propagate along the positive @xmath35axis through a 3d rectangular grid of @xmath49 cells . the box measures 5 units along the @xmath50 axis , and 1 unit each along the @xmath51 and @xmath52 axes . therefore the grid spacing along each axis is 0.01 units . we explore a variety of configurations for our initial state . we performed runs in which the blob initially is either a small cone , sphere , or plate , with symmetry about the propagation vector . the dimensions of the initial blob are optimized , according to its shape , so that @xmath26 grid points out of the @xmath53 in the computational domain initially comprise the blob . an initial spread is imparted to each fluid element by setting the @xmath51 and @xmath52 components of velocity such that @xmath54 . the motivation for taking @xmath55 is the study of frail et al . ( 2001 ) that finds a peak in the frequency histogram distribution of inferred spreading angles @xmath1 from observed grbs of @xmath13 or 0.07 radian . we divide this value by two to obtain the half - angle spread @xmath56 radian . the blob is given a lorentz factor @xmath57 so that the relativistic beaming angle @xmath3 lies within the physical spreading angle @xmath1 of the jet . we set @xmath58 ( inside the blob ) , @xmath59 ( outside the blob ) , and @xmath60 . we also perform one trial with @xmath61 , i.e. , @xmath62 at all points . this is the `` null hypothesis '' run . the effective density contrast perceived in the lab ( i.e. , csm ) frame is @xmath63 . soon after the computation starts , the initial profile relaxes to one in which @xmath64 , therefore the effective density contrast for the early run is @xmath65 . this is less than expected astrophysically , but required in our computations in order to see significant deceleration of the blob by the time it reaches the end of the grid @xmath66 . in other words , we must telescope the evolution into the finite dimensions of our grid . from conservation of momentum , the condition for significant deceleration is that the total swept - up mass - energy roughly equals that in the initial blob . experimentation using hll fluxes shows that strong internal shocks almost immediately create large - amplitude sawtooth @xmath32 variations within the blob , leading to noisy results . therefore we utilize the smoother llf fluxes in this work . we follow the evolution of the blob in terms of both its motion in @xmath50 and its spreading in @xmath51 and @xmath52 . taking a courant number of @xmath67 necessitates @xmath68 time steps for the blob to reach the end of the grid , traveling at @xmath69 . this is a simple consequence of the box length and grid spacing @xmath70 , from which it follows that the time to traverse the box is @xmath71 units traveling at @xmath72 . a density contrast of @xmath73 is sufficient to see the desired deceleration from @xmath74 to @xmath75 during a run . in order to avoid undue complexity in these experiments , that currently are purely hydrodynamic and do not yet contain proper prescriptions for emission from bremsstrahlung and synchrotron processes , we calculate a simple measure of the emissivity by taking the local emission to scale as @xmath34 , which would be expected roughly for optically thin synchrotron emission characteristic of frequencies significantly above the self - absorption frequency . we avoid the issues of synchrotron self - absorption ( granot , piran , & sari 1999b , 2000 ; panaitescu & kumar 2000 ; sari & esin 2001 ) , and of whether the evolution is adiabatic or radiative ( panaitescu & mszros 1998b ; sari , piran , & narayan 1998 ) @xmath40 our calculations are adiabatic . we calculate effective `` light curves '' for observers at various viewing angles between @xmath76 and @xmath77 from the center of the jet . the amplification of the photon energy flux is dictated by three factors : ( 1 ) the rate of photon emission is increased by a factor of @xmath2 as one goes from the blob frame to the lab frame . ( 2 ) the individual photons are doppler - shifted so that their lab frame energy is increased by a factor of @xmath78 , where @xmath79 is the angle between the line of sight and blob s direction of motion . ( 3 ) the photons are focused in the direction of motion , so that an angle of emission ( relative to the direction of motion ) in the blob s frame , @xmath80 , is related to the lab frame angle @xmath79 by @xmath81 . the differential solid angle @xmath82 $ ] . thus , the net amplification of the photon energy flux for each fluid element is @xmath83 , similar expressions have been derived previously ( e.g. , granot , piran , & sari 1999a , eqns . [ 3 ] , [ 4 ] ; wang , & loeb 2001 , eqns . [ 13 ] , [ 31 ] ) . the formalisms of granot et al ( 1999a ) and wang & loeb ( 2001 ) also contain a provision for time delays based on the relative positions and velocities of different fluid elements . the `` time '' of interest in summing the relative emission contributions is the retarded time @xmath84 . we compute light curves by keeping running totals of the contributions to the flux from 100 slices through the computational domain ( straddling the position of local disturbance in the fluid caused by the advancing blob ) that are orthogonal to the blob velocity vector ( i.e. , cuts in the @xmath85 plane ) and advance in time toward the observer at @xmath86 . over the course of a run we build up a light curve based on constant retarded times with respect to the observer . this approach is essential insofar as the photons emitted by a given fluid element outpace the blob itself only slightly . figure 3 shows the evolution of physical quantities for the run which begins with a small cone along a cut through the center of the computational box , along the direction of propagation of the blob . the initial conditions quickly disappear and give way to the deceleration phase noted by kobayashi et al . ( 1999 ) . during the entire evolution there is a local maximum in @xmath34 in advance of the propagating blob that accompanies the leading pile - up in density . the region of highest @xmath2 lies within the density minimum that lies just behind this shocked region . although the initial pressure is small , after the bow shock is fully developed we have @xmath87 throughout the subsequent evolution . the evolution proceeds in a roughly self - similar manner and the forward / reverse shock system is stable , in accord with analytic estimates ( wang , loeb , & waxman 2002 ) . figures 4 and 5 show a time series of contour plots in @xmath2 and @xmath32 that follow the blob propagation and radial spreading . the full 500 grid points along the @xmath50 direction are shown , and each frame represents 240 time steps ( i.e. , @xmath88 ) . the leading contours show the density enhancement associated with the shock . the @xmath32 contours indicate the values 0.2 , 0.3 , 0.4 , and 0.5 , while the ( trailing ) @xmath2 contours , indicate values of 1.25 , 2.5 , 5 , 10 , 15 , and 20 . the contours are formed by taking a cut at @xmath89 through the @xmath90 plane . the initial velocity vector of the blob points toward the top of the plot , with @xmath57 . one sees a strong bow shock associated with the maximum in pressure at the point where the blob encounters the csm . the lack of a bow shock in the @xmath2 contours suggests that little or no material in advance of the blob is accelerated to a significant bulk lorentz factor . as the evolution progresses the deceleration causes the higher @xmath2 contours to disappear , and those that remain become increasingly distorted . the bilateral symmetry of the contours evident in figures 4 and 5 attests to the power of the third order differencing scheme given in db02 , insofar as the 3d model has no enforced symmetry . figure 6 shows the evolution of the total rest - mass energy @xmath44 ( dotted ) and total energy @xmath45 ( dashed ) within the grid , over the course of the conical run . the values of @xmath44 and @xmath45 have been normalized to their initial values . the curves that extend significantly below an abscissa value of unity indicate the values summed over the grid . at the six faces of the computational box we continually reset the values of all variables to their initial values , and keep track of the differences between those values and the initial ones . as the run progresses and more high velocity gas reaches the edge of the computational domain and is extracted , there is a net extraction of positive @xmath44 and @xmath45 from the grid . in addition , there is a low density wake that trails the blob , and therefore at the contact points between the wake and the sides of the grid there is a net extraction of negative @xmath44 and @xmath45 . from a practical standpoint , the density inside the wake is so small that the net extraction of negative @xmath44 and @xmath45 does not significantly affect the bookkeeping . the curves in figure 6 that lie close to an abscissa value of unity indicate the summed values for @xmath44 and @xmath45 corrected for both net extracted positive and negative rest - mass energies . the second panel of figure 6 gives an expanded view of the first panel around the abscissa value of unity . after doing the bookkeeping on the @xmath44 and @xmath45 in the extracted gas , conservation of total rest - mass energy and total energy are good to within @xmath91 parts in @xmath92 . figure 7 shows values of total rest - mass energy within selected high @xmath2 cuts through the computational domain for the initially conical run , and the evolution of the @xmath35component of momentum @xmath93 for five different runs @xmath40 the initial blob as a plate ( `` p1 '' and `` p2 '' ) , sphere ( `` s '' ) , cone ( `` c '' ) , and cone with zero spreading ( `` @xmath94 '' ) . the weighting function used in evaluating @xmath93 is @xmath95 . the run p2 uses fewer grid points ( @xmath96 ) than the other runs , thus material shunted aside by the bow shock reaches the edges of the computational domain earlier . in spite of this , the evolution of p1 and p2 is quite similar , i.e. , there is no dynamically significant back reaction on the material inside the computational domain induced by the departing fluid . as we detail in the discussion section , the drag force on the blob increases as the effective area presented by the blob , which for a roughly constant lateral spreading goes as @xmath97 . therefore the integral of the relativistic impulse equation @xmath98 provides a solution of the form @xmath99 , where our fitting for the conical run ( @xmath100 , @xmath101 , and @xmath102 ) is shown by the curve labeled `` @xmath103 '' . the curves in figure 7 showing the variation of the rest - mass energy contained within different @xmath2 cuts follow the values @xmath104 1.001 , 1.01 , 1.1 , 1.5 , 2 , 4 , 6 , and 8 . the higher @xmath2 cuts @xmath105 reveal that a negligible fraction of csm matter gets accelerated to significant values . this is because the curved bow shock shunts material laterally in front of the advancing blob , rather than accelerating it up to a significant fraction of the blob s bulk lorentz factor @xmath106 . the rest - mass energy curves are relatively constant in time up to @xmath107 . at roughly this point the total accumulated csm gas becomes comparable to that in the initial blob . if the blob did not spread laterally , this time would occur when the blob had plowed through a density @xmath108 , where @xmath109 and @xmath110 . this means that the blob would have shed of order its initial momentum by the time it had traveled @xmath65 times its initial length @xmath111 , or @xmath112 units . in practice , the initially imposed tangential velocity @xmath113 leads to the lateral expansion that effectively increases the cross section for interaction of the blob . the negligible decrease in @xmath93 for the @xmath94 run demonstrates the importance of this effect . figure 8 shows the lateral expansion of the blob for the conical run . the blob edge is computed in each time step by first finding the local maximum in either @xmath32 , @xmath2 , or @xmath114 along the @xmath35axis , and then stepping laterally to the position at which the background ( csm ) value of the relevant quantity has been increased by 10% due to expansion of the blob . the transient lateral expansion @xmath115 lasting until @xmath116 is unphysical insofar as it occurs during the period of adjustment to the initial density and velocity profiles . the value of the later spreading rate @xmath117 is basically dictated by the initial @xmath113 value given to the gas . figure 9 shows the evolution of @xmath93 and @xmath1 as the blob propagates . the solid curve in the top panel indicates the same weighted value of @xmath93 as shown in fig . after the transient physical conditions associated with the initial state have vanished , one sees a period of deceleration associated with the increasing drag force . the dashed curves indicate the reciprocal of the full width spreading angle determined by summing and averaging the local values of @xmath118 . due to the complicated pattern of spreading and the mismatch between contours of constant @xmath32 and constant @xmath2 , the specific value for @xmath119 depends on where one cuts off the outer edge of the blob . the three dashed curves use limiting lower values of 4 , 5 and 6 , respectively , for the @xmath2 value , in determining the volume averages that enter into @xmath119 . the spherical / jetlike transition occurs when @xmath120 drops below the @xmath119 value computed using the lower limit @xmath121 = 6 . the bottom panel shows the number of cells used in computing the @xmath119 values . at late times the disappearance of higher velocity matter makes problematic a calculation of @xmath119 based on cuts in @xmath2 . figure 10 shows the evolution of our canonical `` luminosity '' measure , namely @xmath122 , for various viewing angles @xmath79 . as noted previously , the light curves are calculated by summing the emission from all points with the computational domain that lie , in a given time step , on 100 @xmath85 planes which straddle the position of the blob and advance toward the observer at @xmath86 . one sees a slight break in the `` light curve '' for the @xmath123 observer at @xmath124 due to the transition from spherical to jetlike expansion . although our model is too simple to allow any detailed quantitative comparison with observations at this point , it is worth noting that many of the observed afterglow transitions are also smooth ( e.g. , fig . 2 of stanek et al . 1999 ; fig . 1 of harrison et al . 1999 ) , rather than the abrupt breaks evident in some of the semi - analytical models . the transition is not so visible for off - axis viewing angles . the general pattern in which increasingly off - axis observers see significantly lower emission until well past maximum light curve can be seen in the light curves shown in fig . 4b of panaitescu & mszros ( 1999 ) . afterglows coming from those grbs for which the observer does not lie within the initial cone of grb emission are referred to `` orphan''afterglows . future surveys of orphan afterglows might provide constraints on the beaming that are different from those of frail et al . ( 2001 , e.g. , totani & panaitescu 2002 ; levinson , ofek , waxman , & gal - yam 2002 ) . utilizing 3d relativistic hydrodynamical calculations , we have examined the evolution of an expanding relativistic blob of gas intended to be representative of a jet associated with ejecta from an extremely energetic event such as a hypernova , that produces a gamma - ray burst ( aloy et al . 2000 ; tan , matzner , & mckee 2001 ; macfadyen , woosley , & heger 2001 , zhang , woosley , & heger 2003 , zhang , woosley , & macfadyen 2003 ) . since these are the first such calculations applied to the blob during the time in which the afterglow radiation is produced , we have purposely kept them simple in an effort to concentrate on the most fundamental aspects of the physics . we restrict our attention to the transition from spherical to jetlike expansion that occurs during the time that the lorentz factor becomes less than the reciprocal of the jet spreading angle . we have not yet attached specific numbers to our results . from the srhd equations , one sees that the relevant quantities are the ratios of pressure to density , and of distance to time . if we specify either one of these two sets of numbers , the other one is also determined . the column giving the observed afterglow break time @xmath125 in table 1 of frail et al . ( 2001 ) indicates @xmath126 d as being representative . for an observer directly on the velocity vector of the blob , the time @xmath127 between the grb and afterglow @xmath128 \simeq { 1\over 2c } \int \delta x \ \gamma(x)^{-2},\ ] ] where the dominant contribution to the integral comes from later times . thus the light travel time of 1 day is multiplied by @xmath129 , assuming the spherical - to - jetlike transition giving the break in the afterglow light curve happens at @xmath130 . for the conical run , the break in the light curve occurs at at @xmath131 . if we designate this point as corresponding to a time 2 d , then @xmath132 translates to @xmath133 cm @xmath134 cm . since we have 100 grid points per unit length , this means each grid point spacing corresponds to @xmath135 cm . zhang , woosley , & macfadyen ( 2003 ) utilize a 2d relativistic hydrodynamics code to follow the evolution of the blob through the envelope of the progenitor star and past the point of `` break - out '' through the stellar surface . they find that a blob launched at @xmath136 km with an initial spreading angle of @xmath137 and a bulk lorentz factor of 50 experiences strong shock heating and also lateral confinement , so that when it emerges from the progenitor it has a large internal energy @xmath138 , a bulk lorentz factor @xmath139 , and is more confined than initially ( @xmath140 ) . the subsequent expansion due to the large @xmath141 value leads to an effective @xmath142 for a distant observer . in the standard model for grbs , the gamma radiation is produced by strong internal shocks in the expanding fireball at the point where it becomes optically thin to its own radiation , at @xmath143 s , where @xmath144 is the total energy carried by the jet , corrected for beaming , in units of @xmath145 ergs , and @xmath146 is the csm density in units of @xmath147 @xmath148 . during the time of gamma ray emission , the bulk lorentz factor of the ejecta @xmath149 . zhang , woosley , & heger ( 2003 ) compute the evolution after break - out covering the time of the prompt grb emission . they estimate it should occur at about @xmath150 to @xmath151 cm and have a duration @xmath152 s. in this work we focus on the time after this event , i.e. , subsequent to the evolution considered by zhang et al . ( 2003ab ) , when the `` afterglow '' is produced . this emission is thought to arise primarily in the bow shock where the strong heating leads to @xmath153 ( wang & loeb 2001 ) . in essence , this energy production arises from `` external shocks '' , as opposed to the `` internal shocks '' that generate the gamma radiation ( rees & mszros 1994 ; panaitescu , & mszros 1998a ; spada , panaitescu , & mszros 2000 , kobayashi & sari 2001 ) . after the blob is ejected from the envelope of the star that was the hypernova precursor and continues to propagate , the medium through which it travels should be dominated by the density profile left from the precursor s stellar wind ( chevalier & li 1999 , 2000 , li & chevalier 2001 ) . if the density varies as @xmath154 away from the star , the mass loss of the wind @xmath155 , and the wind velocity @xmath156 km s@xmath157 , the circumstellar density at the point where the afterglow radiation is emitted @xmath158 cm will be @xmath159 @xmath148 . in the calculation of zhang et al ( 2003 ) , the density inside the jet is @xmath160 @xmath148 at @xmath161 cm , and @xmath162 . if the spreading angle of the jet remains roughly constant from @xmath161 cm to @xmath158 cm ( lithwick & sari 2001 ) , then the density inside the jet at @xmath158 cm should be lower by @xmath163 , or about @xmath164 @xmath148 . at this point @xmath34 inside the jet should be negligible compared to @xmath32 . the density contrast between the blob and csm is greater than what we have assumed in the results shown previously . a separate run taking an initial density contrast of @xmath165 shows the same basic effects as the previous runs , however , namely a lateral expansion rate of @xmath166 and the non - accretion of csm gas . our computations lie within the `` deceleration phase '' discussed by kobayashi et al . ( 1999 , see also kobayashi & sari 2000 , 2001 ) . we find a change in the form of the luminosity decrease corresponding to the transition between spherical and jetlike expansion . the determination of the average spreading angle @xmath119 is nontrivial because it depends on how one does the averaging , and how much of the diffuse , sideways - expanding jet material is included in the computation . in figure 9 we presented cuts for gas possessing @xmath167 4 , 5 , and 6 as representative of material in the flow that partakes most strongly in producing the observed radiation . we do not see a dramatic increase in @xmath119 during the deceleration phase ; the @xmath119 value basically reflects the ballistic motion of material following its initial @xmath113 value . also , because the deceleration in our problem is forced not by the accumulation of gas from the csm but rather the drag force of the csm on the blob , we do not see an exponential decrease in @xmath2 with distance during the deceleration phase , but rather a decrease well - approximated by a cubic dependence that is expected for a roughly constant lateral spreading rate . many of the previous studies , utilizing the formalism of bm76 in which matter accreted from the csm is added onto the grb shell and facilitates the deceleration , have discussed the sideways expansion of the jet that supposedly occurs near the time that most of the initial kinetic energy of the blob has been shed . ( a qualitative depiction of this effect is shown in fig . 1 of piran 2002 . ) our results do not appear to support this viewpoint . we find that the csm gas hitting the bow shock does not add significantly to the mass of the jet , but rather is swept back into its wake as the blob passes a given location . in this sense , the relevant factor in determining the deceleration is simply the projected csm surface density relative to that in the blob . as a result , there is no runaway phase of rapid lateral expansion @xmath40 a physical process that has been `` built in '' to some previous theories . we do not see an abrupt jump in the lateral spreading coincident with the deceleration phase ; the lateral spreading rate is mandated basically by the initial @xmath113 value imparted to the gas . the pressure - driven lateral expansion is negligible until such time as @xmath168 , which for @xmath169 is so late as to be uninteresting . in the aforementioned works there is a close connection between the transition @xmath170 and the change from spherical to jetlike emission , whereas in our study there is not . the break in the light curve produced by emission from the bow shock is due solely to the lessening of relativistic beaming brought about by a slow , uniformly increasing drag from the csm , and the fact that once @xmath3 becomes less than @xmath1 , the observer perceives the finite width of the jet . rhoads ( 1999 ) estimates that the lateral spreading speed of the jet should be @xmath171 . sari , piran , & halpern ( 1999 ) argue for a faster spreading rate @xmath172 . rhoads ( 1999 ) divides the dynamical evolution of the blob into two regimes , characterized by a power law decay ( in time ) of the bulk lorentz factor , followed by an exponential decay . during the latter stage , the swept - up mass increases exponentially in time . panaitescu & mszros ( 1999 ) calculate light curves for observers at varying angles from the jet axis , and calculate separately the effects of including and excluding the lateral jet expansion . they find that the maxima in the light curves occur substantially later in runs which do not take into account the jet broadening ( see their fig . 4 ) . in the analytical models of both rhoads ( 1999 ) and panaitescu & mszros ( 1999 ) , the physics of mass accumulation from the csm is an integral component of the formalism ; all mass within the solid angle of the expanding shell is assumed to accrete . workers have applied the results of rhoads ( 1999 ) and sari , piran , & halpern ( 1999 ) to the afterglow evolution , however the results of zhang et al . ( 2003 ) cast doubt on the validity of this exercise , because at the time corresponding to the afterglow emission one anticipates that @xmath173 and the lateral spreading rate would not be governed by the internal sound speed but rather the ballistic motions of the ejecta comprising the blob as they leave the vicinity of the progenitor star . the blob has a large internal thermal energy @xmath174 as it emerges from the progenitor star , and even by the time the expanding ejecta have become optically thin to their own emission from internal shocks ( producing the grb ) , one still expects @xmath175 . by the ( much later ) time of the afterglow emission , however , the blob would have cooled to the point that @xmath176 , which provides the impetus for our initial condition @xmath177 . following the arguments of rhoads ( 1999 ) and sari , piran , & halpern ( 1999 ) , if the lateral spreading rate were mandated by the internal sound speed , then in our calculations it should be @xmath178 , whereas we find it to be @xmath139 times larger . the spreading rate basically reflects our initial @xmath113 value . thus the physics of the lateral expansion is different than in rhoads ( 1999 ) and sari , piran , & halpern ( 1999 ) . how might this result be influenced by systematic effects present in our model ? one obvious potential shortcoming is the absence of cooling . in this work we have assumed an adiabatic gas , whereas in reality one might envision the presence of cooling within the shock . this might then reduce the ability of the shock to deflect the gas as efficiently as it now does , which in our model prevents the acceleration of csm material to lorentz factors approaching that of the blob . we adopt a relatively small density contrast between material inside the blob and the exterior csm in order to see significant evolution of the blob during the course of the simulation within our eulerian grid . we have run additional models using @xmath179 , and although these could not be run for sufficient time to see the deceleration phase begin , we find in these runs also a lateral spreading rate @xmath180 , for the same initial @xmath113 values , and negligible accretion of gas from the csm . we may understand the deceleration of the blob with a simple analytical model . if we assume that the only significant contribution to @xmath181 is through forward motion , that the drag force varies as @xmath182 , and that @xmath183 , then @xmath184 implies @xmath185 where @xmath186 , @xmath187 is the cross sectional area for interaction @xmath188 , @xmath189 is the ( constant ) lateral spreading speed ( @xmath180 in our calculations ) , and @xmath190 is a dimensionless number ( typically of order 0.2 in laboratory applications ) . taking @xmath191 and @xmath192 gives @xmath193 where the blob mass @xmath194 . the blob volume is given by the number of cells comprising the blob initially ( @xmath26 ) times the volume of an element @xmath195 . therefore @xmath196 . integrating gives @xmath197 where @xmath198 is the initial value of @xmath199 and @xmath200 . from the fitting to @xmath201 presented in figure 7 , we infer that @xmath202 . the specific numerical value for @xmath190 is probably influenced by our numerical resolution . a comparison of the curves labelled `` c '' and `` @xmath103 '' in fig . 7 shows that the functional decrease in @xmath201 with @xmath50 is reasonably described by a cubic , as expected if the cross sectional area increases quadratically with @xmath50 , or equivalently @xmath203 . note that we assumed @xmath204 in this exercise , which is a good approximation for the evolution of interest . the calculations we present are the first 3d relativistic hydrodynamical calculations of grb jet evolution pertinent to the afterglow phase that do not enforce any special symmetry ( e.g. , spherical or axial ) . we find that ( i ) the csm gas does not accrete onto the advancing blob , but rather is shunted aside by the bow shock , ( ii ) the decay light curve steepens roughly when one first `` sees '' the edge of the jet @xmath4 , with this effect being strongest for `` face - on '' observers ( confirming previous studies ) , and ( iii ) the rate of decrease of the @xmath35component of momentum @xmath93 is well - characterized by a simple model in which the cross sectional area of the blob increases quadratically with laboratory time ( or distance ) . the primary impetus for the built - in assumption of accretion of matter in previous studies was the influential work of bm76 in which spherical relativistic expansion was considered . accretion of gas onto the relativistically expanding shell is obviously justified for spherical expansion , but subsequent grb workers applied the results to the case of the grb jet , in which a thin wedge of material propagates through a low density medium . in such a situation the natural tendency of material in front of the jet is to be pushed aside and to form a `` channel flow '' around the jet , rather than to accrete . a separate issue is that workers used the results of rhoads ( 1999 ) and sari et al . ( 1999 ) that give lateral expansion rates of order @xmath86 to explain the phase of afterglow evolution , while it now turns out from the work of zhang et al . ( 2003ab ) that the much smaller internal energy in the afterglow - producing blob would be expected to give a much lower spreading rate , were that the only relevant spreading mechanism . the evolution of the blob prior to the afterglow epoch , however , in particular its emergence from the progenitor envelope , did encompass a period of much greater @xmath141 during which a large @xmath189 was imparted to the gas . in summary , there is nothing in our results to suggest that bm76 , rhoads ( 1999 ) , and sari et al . ( 1999 ) are not internally consistent , rather it appears that the subsequent application of their results to afterglow evolution may have been inappropriate . recent semi - analytical work has centered on `` structured '' relativistic jet modeling , wherein one replaces the `` top - hat '' assumption of uniform jet properties , e.g. , lorentz factor and @xmath141 constant within the initial cone angle , with more physically motivated forms in which the physical parameters vary with angle away from the jet symmetry axis @xmath79 ( e.g. , kumar & granot 2003 , granot & kumar 2003 ) . structured jets represent a refinement in the semi - analytical and analytical work in the sense that unphysical artifacts of the top - hat models are avoided . in this work we allow naturally occurring gradients within the flow determine the evolution . in comparing the runs with different starting conditions , for instance , we see that only the gross initial characteristics ( bulk lorentz factor and spreading rate ) are of relevance . one could impose a structured initial state , in terms of having an angle - dependent relation between , for instance lorentz factor and @xmath79 , and/or @xmath141 and @xmath79 , but one suspects that the initial intricacies would be washed out , just as the initial blob shape is . one might in fact imagine inverting the problem and using our results to determine physically motivated functional forms for @xmath205 and @xmath206 as input to a structured jet formalism , but in our models the detailed evolution of the physical parameters and the interaction of the blob with the bow shock reveal gross changes along the jet axis . therefore the motivation for trying to characterize the jet properties as simple functions of @xmath79 , averaged along the axis of the jet , seems questionable . in addition , one would still be missing important physical effects , such as the non - accretion of the forward csm material . two obvious refinements , currently being carried out , are to ( i ) treat the problem on a lagrangian grid in which the mesh points follow the blob and are adaptively inserted in regions with strong gradients , so as to be able to explore regimes in which the density contrast between the blob and csm is much larger , and ( ii ) include provisions for realistic bremsstrahlung and synchrotron physics , in order to produce light curves that can be compared directly with observations so as to test different aspects of the theory and thereby constrain the allowed parameter space . our sincerest thanks go to luca del zanna who provided key insights into the workings of db02 . we also thank john baker , david band , tom cline , chris fragile , chris fryer , markos georganopoulos , peter goldreich , demos kazanas , pawan kumar , zhi - yun li , andrew macfadyen , peter mszros , ewald mueller , jay norris , alin panaitescu , sterl phinney , steve reynolds , james rhoads , steve ruden , sabrina savage , craig wheeler , and weiqun zhang for helpful comments . the evolution of a spherical relativistic expansion due to a large over - pressure inside a sphere of radius 0.1 , adopting hll fluxes . the evolution encompasses 360 time steps , taking a courant number of 0.05 . shown are the initial conditions plus ten equally spaced time steps taken from a slice along the @xmath35axis depicting the evolution of ( i ) pressure @xmath34 ( _ top panel _ ) , ( ii ) density @xmath32 ( _ second panel _ ) , ( iii ) @xmath207 ( _ third panel _ ) , and ( iv ) lorentz factor @xmath2 ( _ bottom panel _ ) . the small numbers beside each curve indicate the evolution . one sees an expansion of the central over - pressurized sphere into the surrounding medium . the strong relativistic outflow peaks at @xmath208 near the end of the evolution . the general trend in which @xmath209 within the expansion can be derived from fundamental principles and is well known from classical solutions . at late times the strong decrease in @xmath32 at the center of the sphere , which is a singularity in this test , begins to cause numerical instability . we utilize the llf fluxes for @xmath36 and the hll fluxes for @xmath210 . the evolution of a spherical relativistic expansion due to a large over - pressure inside a sphere of radius 0.1 , adopting pure llf fluxes . quantities shown are the same as in fig . the evolution is nearly identical to that in fig . 1 , except the variations in the physical variables are smoother , and @xmath211 at larger radii by the end of the run @xmath40 somewhat smaller than in fig . 1 . the evolution of a relativistic blob launched as a small cone traveling along the @xmath212-axis , initially confined to between @xmath213 and @xmath214 and a maximum radius 0.025 . the small numbers beside each curve indicate the time step . the panels are the same as shown in fig . 1 . snapshots represent conditions along a slice through the center of the rectangular grid , taken every 120 time steps ( @xmath215 ) . for this trial and all that follow we utilize a courant number of 0.25 and pure lff fluxes . the initial bulb lorentz factor @xmath2 was set to 25 , and the initial full width flaring angle to 0.07 radian as measured by @xmath216 . the spreading in @xmath32 in the @xmath217 plane . the eight panels depict the initial state , and seven subsequent snapshots taken every 240 time steps ( i.e. , @xmath88 ) . the evolution represents a slice through the midplane of the grid , and the full length and width of the box are shown . contours indicate values of 0.2 , 0.3 , 0.4 , and 0.5 ( in the same dimensionless units as shown in fig . 3 ) . figure 5 . the spreading in @xmath2 in the @xmath217 plane . the conventions are the same as in the previous figure . contours indicate values of 1.25 , 2.5 , 5 , 10 , 15 , and 20 . at later times the higher @xmath2 contours disappear due to deceleration . the absence of a bow shock in this depiction shows that material in front of the advancing blob is not accelerated to @xmath2 values much above unity . conservation of total rest - mass energy @xmath44 and total energy @xmath45 . at @xmath218 material starts to reach the edges of the grid and flow off , therefore one sees the start of a linear decrease in @xmath44 and @xmath45 . the curves which closely follow @xmath219 show the values for @xmath44 and @xmath45 after correcting for material extracted from the grid . the bottom panel shows an expanded view of the region near @xmath219 in the first panel . the evolution of rest mass energy contained within various mass cuts , for the conical run ( solid lines ) , and the evolution of the weighted @xmath35component of the momentum @xmath93 , divided by 5 so as to be on a common scale with the rest mass energy curves ( dashed lines ) . the numbers associated with the solid lines indicate the @xmath2 value used in each rest mass energy cut . we show the @xmath93 evolution for the conical run ( c ) , the spherical run ( s ) and the two plate runs ( p1 : @xmath220 , p2 : @xmath221 ) . for these four runs the initial spread in velocities is such that @xmath222 , whereas for the run labelled @xmath94 there is no tangential component initially to the velocities . consequently the reduced lateral spreading rate leads to a slower @xmath93 decrease . the function @xmath103 represents a fitting to the conical run given by @xmath223 . although the rest mass energy curves for @xmath224 and 1.01 show a sharp increase ( up until @xmath225 when material starts to leave the edges of the grid ) , the near constancy of the @xmath226 curve and the slight decline for higher @xmath227 curves indicate that there is negligible acceleration of high@xmath228 material . also , the functional form of the decrease in @xmath93 is similar for all runs in which @xmath222 initially . the location of the edge of the jet , determined by first finding the location along the direction of propagation of the maximum in either @xmath2 ( _ top curve _ ) , @xmath114 ( _ middle curve _ ) , or @xmath32 ( _ bottom curve _ ) , then stepping laterally to the point at which the background value has increased by 10% of its original value . ( the precise value of this constant does not affect the results . ) the initial maximum radius of the blob is 0.05 . after @xmath229 the two lower curves coincide . the evolution of the weighted mean of the @xmath35 component of momentum and a measure of the spreading angle @xmath119 , taken to be the ratio of tangential to axial speeds , in the observer s frame . shown are ( i ) @xmath93 ( solid line ) and the reciprocal of @xmath119 , measured in radians ( dashed lines ) , where the lower limiting @xmath2 value used in the averaging for @xmath119 is taken to be either 4 , 5 , or 6 ( _ top panel _ ) . ( ii ) the @xmath119 values whose reciprocal values are indicated in the first panel ( _ middle panel _ ) , and ( iii ) the number of cells entering into the averages for the three @xmath121 values in the first two panels ( _ bottom panel _ ) . at late times the deceleration makes the @xmath119 calculations problematic because the number of high @xmath2 cells drops . this progression in the loss of high @xmath2 cells can be seen at @xmath230 in the bottom panel . figure 10 . light curves constructed by summing the quantity @xmath231 over the grid . we take this global sum as being a measure of the luminosity seen by an observer looking down the jet . the light curves are built up during the course of the run by summing the emission from 100 slices moving toward the observer at @xmath86 that straddle the location of the disturbance induced by the blob . thus the light curve consists of a maximum of 100 points , where @xmath232 is measured with respect to the arrival time of the first comoving slice . the five values of the angle @xmath79 between the observer and the local velocity vector of a fluid element are ( going to smaller emission values ) @xmath76 , @xmath233 , @xmath234 , @xmath235 , and @xmath77 . for the `` face - on '' viewer ( inclination @xmath236 ) , there is a mild change in the decay slope , i.e. , break , between the regimes of `` spherical '' and `` jetlike '' expansion .

a hundred years and more of fundamental research in atomic and nuclear physics has shown that all matter is corpuscular , with the atoms that comprise us , themselves containing a dense nuclear core . this core is composed of protons and neutrons , referred to collectively as nucleons , which are members of a broader class of femtometre - scale particles , called hadrons . in working toward an understanding of hadrons , we have discovered that they are complicated bound - states of quarks and gluons . these quarks and gluons are elementary , pointlike excitations , whose interactions are described by a poincar invariant quantum non - abelian gauge field theory ; namely , quantum chromodynamics ( qcd ) . the goal of hadron physics is the provision of a quantitative explanation of the properties of hadrons through a solution of qcd . quantum chromodynamics is the strong - interaction part of the standard model of particle physics and solving qcd presents a fundamental problem that is unique in the history of science . never before have we been confronted by a theory whose elementary excitations are not those degrees - of - freedom readily accessible via experiment ; i.e. , whose elementary excitations are _ confined_. moreover , there are numerous reasons to believe that qcd generates forces which are so strong that less - than 2% of a nucleon s mass can be attributed to the so - called current - quark masses that appear in qcd s lagrangian ; viz . , forces capable of generating mass from nothing , a phenomenon known as dynamical chiral symmetry breaking ( dcsb ) . neither confinement nor dcsb is apparent in qcd s lagrangian and yet they play the dominant role in determining the observable characteristics of real - world qcd . the physics of hadrons is ruled by _ emergent phenomena _ such as these , which can only be elucidated through the use of nonperturbative methods in quantum field theory . this is both the greatest novelty and the greatest challenge within the standard model . we must find essentially new ways and means to explain precisely via mathematics the observable content of qcd . the complex of dyson - schwinger equations ( dses ) is a powerful tool , which has been employed with marked success to study confinement and dcsb , and their impact on hadron observables . this will be emphasised and exemplified in this concise review , which describes selected recent progress in the study of mesons . it complements and extends earlier and other efforts @xcite . the basic problem of hadron physics is to solve qcd . this inspiring goal will only be achieved through a joint effort from experiment and theory because it is the feedback between them that leads most rapidly to improvements in understanding . the hadron physics community now has a range of major facilities that are accumulating data , of unprecedented accuracy and precision , which pose important challenges for theory . the opportunities for researchers in hadron physics promise to expand with the use of extant accelerators , and upgraded and new machines and detectors that will appear on a five - to - ten - year time - scale , in china , germany , japan , switzerland and the usa . a short list of facilities may readily be compiled : beijing s electron - positron collider ; in germany cosy ( jlich cooler synchrotron ) , elsa ( bonn electron stretcher and accelerator ) , mami ( mainz microtron ) , and fair ( facility for antiproton and ion research ) under construction near darmstadt ; in japan j - parc ( japan proton accelerator research complex ) under construction in tokai - mura , 150 km ne of tokyo , and kek , tsukuba ; in switzerland , the alice and compass detectors at cern ; and in the usa , both the thomas jefferson national accelerator facility ( jlab ) , currently being upgraded , with new generation experiments expected in 2016 , and rhic ( relativistic heavy ion collider ) at brookhaven national laboratory . asymptotic coloured states have not been observed , but is it a cardinal fact that they can not ? no solution to qcd will be complete if it does not explain confinement . this means confinement in the real world , which contains quarks with light current - quark masses . this is distinct from the artificial universe of pure - gauge qcd without dynamical quarks , studies of which tend merely to focus on achieving an area law for a wilson loop and hence are irrelevant to the question of light - quark confinement . in stepping toward an answer to the question of confinement , it will likely be necessary to map out the long - range behaviour of the interaction between light - quarks ; namely , qcd s @xmath4-function at infrared momenta . in this connection it is noteworthy that the spectrum of meson and baryon excited states , and hadron elastic and transition form factors provide unique information about the long - range interaction between light - quarks and , in addition , the distribution of a hadron s characterising properties such as mass and momentum , linear and angular amongst its qcd constituents . the upgraded and promised future facilities will provide data that should guide the charting process . however , to make full use of that data , it will be necessary to have poincar covariant theoretical tools that enable the reliable study of hadrons in the mass range @xmath5-@xmath6gev . crucially , on this domain both confinement and dcsb are germane . it is known that dcsb ; namely , the generation of mass _ from nothing _ , does take place in qcd . it arises primarily because a dense cloud of gluons comes to clothe a low - momentum quark @xcite . this is readily seen by solving the dse for the dressed - quark propagator ; i.e. , the gap equation , which yields the result illustrated in fig.[gluoncloud ] . ( in our euclidean metric : @xmath7 ; @xmath8 ; @xmath9 , tr@xmath10=-4 \epsilon_{\mu\nu\rho\sigma}$ ] ; @xmath11 $ ] ; @xmath12 ; and @xmath13 timelike @xmath14 @xmath15 . more information is available in app.a of ref.@xcite . ) however , the origin of the interaction strength at infrared momenta , which guarantees dcsb through the gap equation , is currently unknown . this relationship ties confinement to dcsb . the reality of dcsb means the higgs mechanism is largely irrelevant to the bulk of normal matter in the universe . instead the most important mass generating mechanism for light - quark hadrons is the strong interaction effect of dcsb ; e.g. , one may identify it as being responsible for 98% of a proton s mass @xcite . there is a caveat ; namely , as so often , the pion is exceptional . its mass is given by the simple product of two terms , one of which is the ratio of two order parameters for dcsb whilst the other is determined by the current - quark mass ( sec.[psmassformula ] ) . hence the pion would be massless in the absence of a mechanism that can generate a current - mass for at least one light - quark . the impact of a massless , strongly - interacting particle on the physics of the universe would be dramatic . dressed - quark mass function , @xmath16 in eq.([sgeneraln ] ) : _ solid curves _ dse results , explained in @xcite , `` data '' numerical simulations of lattice - qcd @xcite . ( nb . @xmath17mev is the uppermost curve and current - quark mass decreases from top to bottom . ) one observes the current - quark of perturbative qcd evolving into a constituent - quark as its momentum becomes smaller . the constituent - quark mass arises from a cloud of low - momentum gluons attaching themselves to the current - quark . this is dynamical chiral symmetry breaking ( dcsb ) : an essentially nonperturbative effect that generates a quark mass _ from nothing _ ; namely , it occurs even in the chiral limit . ( figure adapted from ref.@xcite.),scaledwidth=66.0% ] it is natural to ask whether the connection between confinement and dcsb is accidental or causal . there are models with dcsb but not confinement , however , a model with confinement but lacking dcsb has not yet been identified ( see , e.g. , secs.2.1 and 2.2 of ref.@xcite ) . this leads to a conjecture that dcsb is a necessary consequence of confinement . it is interesting that there are numerous models and theories which exhibit both confinement and dcsb , and possess an external control parameter such that deconfinement and chiral symmetry restoration occur simultaneously at some critical value of this parameter ; e.g. , quantum electrodynamics in three dimensions with @xmath18 electrons @xcite , and models of qcd at nonzero temperature and chemical potential @xcite . whether this simultaneity is a property possessed by qcd , and/or some broader class of theories , in response to changes in : the number of light - quark flavours ; temperature ; or chemical potential , is a longstanding question . the momentum - dependence of the quark mass , illustrated in fig.[gluoncloud ] , is an essentially quantum field theoretic effect , unrealisable in quantum mechanics , and a fundamental feature of qcd . this single curve connects the infrared and ultraviolet regimes of the theory , and establishes that the constituent - quark and current - quark masses are simply two connected points separated by a large momentum interval . the curve shows that qcd s dressed - quark behaves as a constituent - quark , a current - quark , or something in between , depending on the momentum of the probe which explores the bound - state containing the dressed - quark . it follows that calculations addressing momentum transfers @xmath19 , where @xmath20 is the mass of the hadron involved , require a poincar - covariant approach that can veraciously realise quantum field theoretical effects @xcite . owing to the vector - exchange character of qcd , covariance also guarantees the existence of nonzero quark orbital angular momentum in a hadron s rest - frame @xcite . the dressed - quark mass function has a remarkable capacity to correlate and to contribute significantly in explaining a wide range of diverse phenomena . this brings urgency to the need to understand the relationship between parton properties in the light - front frame , whose peculiar properties simplify some theoretical analyses , and the structure of hadrons as measured in the rest frame or other smoothly related frames . this is a problem because , e.g. , dcsb , an established keystone of low - energy qcd , has not explicitly been realised in the light - front formulation . the obstacle is the constraint @xmath21 for massive quanta on the light front @xcite . it is therefore impossible to make zero momentum fock states that contain particles and hence the vacuum is `` trivial '' . on the other hand , it is conceivable that dcsb is inextricably tied with the formation and structure of goldstone modes and not otherwise a measurable property of the vacuum . this conjecture is being explored @xcite and is something about which more will be written herein ( sec.[inhadroncondensate ] ) . in addition , parton distribution functions , which have a probability interpretation in the infinite momentum frame , must be calculated in order to comprehend their content : parametrisation is insufficient . it would be very interesting to know , e.g. , how , if at all , the distribution functions of a goldstone mode differ from those of other hadrons @xcite . it is worth stating plainly that the potential between infinitely - heavy quarks measured in numerical simulations of quenched lattice - regularised qcd the so - called static potential is simply _ irrelevant _ to the question of confinement in the real world , in which light quarks are ubiquitous . in fact , it is a basic feature of qcd that light - particle creation and annihilation effects are essentially nonperturbative and therefore it is impossible in principle to compute a potential between two light quarks @xcite . drawing on a long list of sources ; e.g. , refs.@xcite , a perspective on confinement was laid out in ref.@xcite . confinement can be related to the analytic properties of qcd s schwinger functions , which are often called euclidean - space green functions . for example , it can be read from the reconstruction theorem @xcite that the only schwinger functions which can be associated with expectation values in the hilbert space of observables ; namely , the set of measurable expectation values , are those that satisfy the axiom of reflection positivity . this is an extremely tight constraint . it can be shown to require as a necessary condition that the fourier transform of the momentum - space schwinger function is a positive - definite function of its arguments . this condition suggests a practical confinement test , which can be used with numerical solutions of the dses ( see , e.g. , sec.iii.c of ref.@xcite and sec.iv of ref.@xcite ) . the implications and use of reflection positivity are discussed and illustrated in sec . 2 of ref.@xcite . it is noteworthy that any 2-point schwinger function with an inflexion point at @xmath22 must breach the axiom of reflection positivity , so that a violation of positivity can be determined by inspection of the pointwise behaviour of the schwinger function in momentum space ( sec.iv.b of ref.@xcite ) . consider then @xmath23 , which is the single scalar function that describes the dressing of a landau - gauge gluon propagator . a large body of work has focused on exposing the behaviour of @xmath23 in the pure yang - mills sector of qcd . these studies are reviewed in ref.@xcite . a connection with the expression and nature of confinement in the yang - mills sector is indicated in fig.[fig : gluonrp ] . the appearance of an inflexion point in the two - point function generated by the gluon s momentum - dependent mass - function is impossible to overlook . hence this gluon can not appear in the hilbert space of observable states . the inflexion point possessed by @xmath24 , visible in fig.[gluoncloud ] , conveys the same properties on the dressed - quark propagator . numerical simulations of lattice - qcd confirm the appearance of an inflexion point in both the dressed - gluon and -quark propagators ; e.g. , see fig.[gluoncloud ] and ref.@xcite . the signal is clearest for the gluon owing to the greater simplicity of simulations in the pure yang - mills sector @xcite . we emphasise that this sense of confinement is essentially quantum field theoretical in nature . amongst its many consequences is that confinement in qcd can not veraciously be represented in potential models . @xmath23 , the function that describes dressing of a landau - gauge gluon propagator , plotted for three distinct cases . a bare gluon is described by @xmath25 ( the dashed line ) , which is plainly convex on @xmath26 . such a propagator has a representation in terms of a non - negative spectral density . in some theories , interactions generate a mass in the transverse part of the gauge - boson propagator , so that @xmath27 , which can also be represented in terms of a non - negative spectral density . in qcd , however , self - interactions generate a momentum - dependent mass for the gluon , which is large at infrared momenta but vanishes in the ultraviolet @xcite . this is illustrated by the curve labelled `` ir - massive but uv - massless . '' with the generation of a mass-_function _ , @xmath23 exhibits an inflexion point and hence can not be expressed in terms of a non - negative spectral density . , scaledwidth=67.0% ] from the perspective that confinement can be related to the analytic properties of qcd s schwinger functions , the question of light - quark confinement can be translated into the challenge of charting the infrared behavior of qcd s _ universal _ @xmath4-function . ( although this function may depend on the scheme chosen to renormalise the theory , it is unique within a given scheme @xcite . of course , the behaviour of the @xmath4-function on the perturbative domain is well known . ) this is a well - posed problem whose solution is an elemental goal of modern hadron physics and which can be addressed in any framework enabling the nonperturbative evaluation of renormalisation constants . it is the @xmath4-function that is responsible for the behaviour evident in figs.[gluoncloud ] and [ fig : gluonrp ] , and one of the more interesting of contemporary questions is whether it is possible to reconstruct the @xmath4-function , or at least constrain it tightly , given empirical information on the gluon and quark mass functions . in order to proceed it is necessary to describe explicitly the best known and simplest dse . the dyson or gap equation describes how quark propagation is influenced by interactions ; viz . , for a quark of flavour @xmath28 , @xmath29 where : @xmath30 is the gluon propagator ; @xmath31 , the quark - gluon vertex ; @xmath32 , a symbol that represents a poincar invariant regularization of the four - dimensional euclidean integral , with @xmath33 the regularization mass - scale ; @xmath34 , the current - quark bare mass ; and @xmath35 , respectively , the vertex and quark wave function renormalisation constants , with @xmath36 the renormalisation point dependence upon which is not usually made explicit . the gap equation s solution is the dressed - quark propagator , @xmath37 which is obtained from eq.([gendsen ] ) augmented by a renormalisation condition . a mass - independent scheme is a useful choice and can be implemented by fixing all renormalisation constants in the chiral limit . ( see , e.g. , ref.@xcite and references therein ; or ref.@xcite for a detailed discussion of renormalisation . ) the mass function , @xmath38 , is independent of the renormalisation point , @xmath36 ; and the renormalised current - quark mass , @xmath39 wherein @xmath40 is the renormalisation constant associated with the lagrangian s mass - term . like the running coupling constant , this `` running mass '' is familiar from textbooks . however , it is not commonly appreciated that @xmath41 is simply the dressed - quark mass function evaluated at one particular deep spacelike point ; viz , @xmath42 the renormalisation - group invariant current - quark mass may be inferred via @xmath43^{\gamma_m } m_f(p^2)\,,\ ] ] where @xmath44 : @xmath18 is the number of quark flavours employed in computing the running coupling ; and @xmath45 is qcd s dynamically - generated renormalisation - group - invariant mass - scale . the chiral limit is expressed by @xmath46 moreover , @xmath47 however , we would like to emphasise that in the presence of dcsb the ratio @xmath48 is not independent of @xmath49 : in the infrared ; i.e. , @xmath50 , it then expresses a ratio of constituent - like quark masses , which , for light quarks , are two orders - of - magnitude larger than their current - masses and nonlinearly related to them @xcite . the gap equation illustrates the features and flaws of each dse . it is a nonlinear integral equation for the dressed - quark propagator and hence can yield much - needed nonperturbative information . however , the kernel involves the two - point function @xmath30 and the three - point function @xmath31 . the gap equation is therefore coupled to the dses satisfied by these functions , which in turn involve higher @xmath51-point functions . hence the dses are a tower of coupled integral equations , with a tractable problem obtained only once a truncation scheme is specified . it is unsurprising that the best known truncation scheme is the weak coupling expansion , which reproduces every diagram in perturbation theory . this scheme is systematic and valuable in the analysis of large momentum transfer phenomena because qcd is asymptotically free but it precludes any possibility of obtaining nonperturbative information . given the importance of dcsb in qcd , it is significant that the dressed - quark propagator features in the axial - vector ward - takahashi identity , which expresses chiral symmetry and its breaking pattern : @xmath52 \,\gamma_5^{fg}(k;p ) = s_f^{-1}(k_+ ) i \gamma_5 + i \gamma_5 s_g^{-1}(k_- ) \ , , \label{avwtimn}\ ] ] where @xmath53 is the total - momentum entering the vertex and @xmath54 is the relative - momentum between the amputated quark legs . to be explicit , @xmath55 , with @xmath56 $ ] , and hence @xmath57 , @xmath58 . in a poincar covariant approach , such as presented by a proper use of dses , no observable can depend on @xmath0 ; i.e. , the definition of the relative momentum . sec.[flavourless ] discusses the important differences encountered in treating flavourless pseudoscalar mesons . in eq.([avwtimn ] ) , @xmath59 and @xmath60 are , respectively , the amputated axial - vector and pseudoscalar vertices . they are both obtained from an inhomogeneous bethe - salpeter equation ( bse ) , which is exemplified here using a textbook expression @xcite : @xmath61_{tu } = z_2 [ \gamma_5 \gamma_\mu]_{tu}+ \int_q^\lambda [ s(q_+ ) \gamma_{5\mu}(q;p ) s(q_- ) ] _ { sr } k_{tu}^{rs}(q , k;p ) , \label{bsetextbook}\ ] ] in which @xmath62 is the fully - amputated quark - antiquark scattering kernel , and the colour- , dirac- and flavour - matrix structure of the elements in the equation is denoted by the indices @xmath63 . nb . by definition , @xmath62 does not contain quark - antiquark to single gauge - boson annihilation diagrams , nor diagrams that become disconnected by cutting one quark and one antiquark line . the ward - takahashi identity , eq.([avwtimn ] ) , entails that an intimate relation exists between the kernel in the gap equation and that in the bse . ( this is another example of the coupling between dses . ) therefore an understanding of chiral symmetry and its dynamical breaking can only be obtained with a truncation scheme that preserves this relation , and hence guarantees eq.([avwtimn ] ) without a fine - tuning of model - dependent parameters . through the gap and bethe - salpeter equations the pointwise behaviour of the @xmath4-function determines the pattern of chiral symmetry breaking ; e.g. , the behaviour in fig.[gluoncloud ] . moreover , the fact that these and other dses connect the @xmath4-function to experimental observables entails that comparison between computations and observations of the hadron mass spectrum , and hadron elastic and transition form factors , can be used to constrain the @xmath4-function s long - range behaviour . in order to realise this goal , a nonperturbative symmetry - preserving dse truncation is necessary . steady quantitative progress can be made with a scheme that is systematically improvable @xcite . in fact , the mere existence of such a scheme has enabled the proof of exact nonperturbative results in qcd . before describing a number of these in some detail , it is worth explicating the range of applications . for example , there are : veracious statements about the pion @xmath64-term @xcite ; radially - excited and hybrid pseudoscalar mesons @xcite ; heavy - light @xcite and heavy - heavy mesons @xcite ; novel results for the pion susceptibility obtained via analysis of the isovector - pseudoscalar vacuum polarisation @xcite , which bear upon the essential content of the so - called `` mexican hat '' potential that is used in building models for qcd ; and a derivation @xcite of the weinberg sum rule @xcite . turning now to a fuller illustration , the first of the results was introduced in ref.@xcite ; namely , a mass formula that is exact for flavour non - diagonal pseudoscalar mesons : @xmath65 where : @xmath66 are the current - masses of the quarks constituting the meson ; and @xmath67 where @xmath68 is the pseudoscalar meson s bound - state bethe - salpeter amplitude : @xmath69 , \label{gengpi}\end{aligned}\ ] ] which is determined from the associated homogeneous bse . it is worth emphasising that the quark wavefunction and lagrangian mass renormalisation constants , @xmath70 , respectively , depend on the gauge parameter in precisely the manner needed to ensure that the right - hand sides of eqs.([fpigen ] ) , ( [ rhogen ] ) are gauge - invariant . moreover , @xmath71 ensures that the right - hand side of eq.([fpigen ] ) is independent of both @xmath36 and @xmath33 , so that @xmath72 is truly an observable ; and @xmath73 ensures that @xmath74 is independent of @xmath33 and evolves with @xmath36 in just the way necessary to guarantee that the product @xmath75 is renormalisation - point - independent . in addition , it should be noted that eq.([mrtrelation ] ) is valid for every pseudoscalar meson and for any value of the current - quark masses ; viz . , @xmath76 , @xmath77 . this includes arbitrarily large values and also the chiral limit , in whose neighbourhood eq.([mrtrelation ] ) can be shown @xcite to reproduce the familiar gell - mann oakes renner relation . the axial - vector ward - takahashi identity , eq.([avwtimn ] ) , is a crucial bridge to eqs.([mrtrelation ] ) ( [ rhogen ] ) ; and on the way one can also prove the following goldberger - treiman - like relations @xcite : @xmath78 wherein the superscript indicates that the associated quantity is evaluated in the chiral limit , and @xmath79 , @xmath80 , @xmath81 are analogues in the inhomogeneous axial - vector vertex of the scalar functions in the @xmath82-meson s bethe - salpeter amplitude . these identities are of critical importance in qcd . the first , eq.([gtlrele ] ) , can be used to prove that a massless pseudoscalar meson appears in the chiral - limit spectrum if , and only if , chiral symmetry is dynamically broken . moreover , it exposes the fascinating consequence that the solution of the two - body pseudoscalar bound - state problem is almost completely known once the one - body problem is solved for the dressed - quark propagator , with the relative momentum within the bound - state identified unambiguously with the momentum of the dressed - quark . this latter emphasises that goldstone s theorem has a pointwise expression in qcd . the remaining three identities are also important because they show that a pseudoscalar meson _ must _ contain components of pseudovector origin . this result overturned a misapprehension of twenty - years standing ; namely , that only @xmath83 is nonzero @xcite . these pseudovector components materially influence the observable properties of pseudoscalar mesons @xcite , as do their analogues in other mesons @xcite . it is natural to reiterate here a prediction for the properties of non - ground - state pseudoscalar mesons , which follows from the exact results described above ; namely , in the chiral limit @xcite @xmath84 this is the statement that goldstone modes are the only pseudoscalar mesons to possess a nonzero leptonic decay constant in the chiral limit when chiral symmetry is dynamically broken . the decay constants of all other pseudoscalar mesons on this trajectory , e.g. , radial excitations , vanish . on the other hand , in the absence of dcsb the leptonic decay constant of each such pseudoscalar meson vanishes in the chiral limit ; i.e , eq.([fpin ] ) is true @xmath85 . lattice - qcd results for the ratio of the decay constants for the first - excited- and ground - state pseudoscalar mesons as a function of the pion mass squared . ( lattice parameters : volume@xmath86 ; @xmath87 , spacing @xmath88fm , two flavours of degenerate sea quarks ; wilson gauge action and clover fermions . ) the `` not improved '' results were obtained from a fermion action with poor chiral symmetry properties . in this case @xmath89 , consistent with expectations based on quantum mechanics . the `` improved '' results were obtained through implementation of the full alpha method for the nonperturbative improvement of the fermion action , which greatly improves the simulation s chiral symmetry properties . in this case , the sign of the ratio was not determined in the simulation but is discussed in ref.@xcite . figure adapted from ref.@xcite.),scaledwidth=66.0% ] from the perspective of quantum mechanics , eq.([fpin ] ) is a surprising fact . the leptonic decay constant for @xmath91-wave states is typically proportional to the wave function at the origin . compared with the ground state , this is smaller for an excited state because the wave function is broader in configuration space and wave functions are normalised . however , it is a modest effect ; e.g. , consider the @xmath92 decay of vector mesons , for which a calculation in relativistic quantum mechanics based on light - front dynamics @xcite yields @xmath93 , consistent with the value inferred from experiment . thus , it is not uncommon for eq.([fpin ] ) to be perceived as `` remarkable '' or `` unbelievable . '' notwithstanding this , in connection with the pion s first radial excitation , the value of @xmath94mev predicted in ref.@xcite is consistent with experiment @xcite and simulations of lattice - qcd @xcite , as illustrated in fig.[fig : fradial ] . it is now recognised that the suppression of @xmath95 is a useful benchmark , which can be used to tune and validate lattice qcd techniques that try to determine the properties of excited states mesons . for the last thirty years , _ condensates _ ; i.e. , nonzero vacuum expectation values of local operators , have been used as parameters in order to correlate and estimate essentially nonperturbative strong - interaction matrix elements @xcite . they are also basic to current algebra analyses . it is conventionally held that such quark and gluon condensates have a physical existence , which is independent of the hadrons that express qcd s asymptotically realisable degrees - of - freedom ; namely , that these condensates are not merely mass - dimensioned parameters in a theoretical truncation scheme , but in fact describe measurable spacetime - independent configurations of qcd s elementary degrees - of - freedom in a hadronless ground state . owing to confinement , however , this view is erroneous @xcite . one may readily recapitulate upon the argument . to begin , note that eq.([fpigen ] ) is the exact expression in qcd for the leptonic decay constant of a pseudoscalar meson . it is a property of the pion and , as consideration of the integral expression reveals , it can be described as the pseudovector projection of the pion s bethe - salpeter wavefunction onto the origin in configuration space . note that the product @xmath96 is called the bethe - salpeter wavefunction because , when a nonrelativistic limit can validly be performed , the quantity @xmath97 at fixed time becomes the quantum mechanical wavefunction for the system under consideration . ( nb . in the neighborhood of the chiral limit , a value for @xmath72 can be estimated via either of two approximation formulae @xcite . these formulae both illustrate and emphasize the role of @xmath72 as an order parameter for dcsb . ) if chiral symmetry were not dynamically broken , then in the neighborhood of the chiral limit @xmath98 @xcite . of course , chiral symmetry is dynamically broken in qcd @xcite and for the ground - state pseudoscalar @xmath99 taken together , these last two observations express the fact that @xmath72 , which is an intrinsic property of the pseudoscalar meson , is a _ bona fide _ order parameter for dcsb . an analysis within chiral perturbation theory @xcite suggests that the chiral limit value , @xmath100 , is @xmath101% below the measured value of 92.4mev ; and efficacious dse studies give a 3% chiral - limit reduction @xcite . now , eq.([rhogen ] ) is kindred to eq.([fpigen ] ) ; it is the expression in quantum field theory which describes the _ pseudoscalar _ projection of the pseudoscalar meson s bethe - salpeter wavefunction onto the origin in configuration space . it is thus truly just another type of pseudoscalar meson decay constant . in this connection it is therefore notable that one may rigorously define an `` in - meson '' condensate ; viz.@xcite : @xmath102 now , using eq.([gtlrele ] ) , one finds @xcite @xmath103 hence the so - called vacuum quark condensate is , in fact , the chiral - limit value of the in - meson condensate ; i.e. , it describes a property of the chiral - limit pseudoscalar meson . one can therefore argue that this condensate is no more a property of the `` vacuum '' than the pseudoscalar meson s chiral - limit leptonic decay constant . moreover , ref.@xcite establishes the equivalence of all three definitions of the so - called vacuum quark condensate : a constant in the operator product expansion @xcite ; via the banks - casher formula @xcite ; and the trace of the chiral - limit dressed - quark propagator . ( nb . in the presence of confinement it is impossible to write a valid nonperturbative definition of a single quark or gluon annihilation operator and therefore impossible to rigorously define a second quantised vacuum for qcd . to do so would be to answer the question : what is the state that is annihilated by an operator which is unknowable ? ) the chiral - limit vacuum quark condensate is therefore qualitatively equivalent to the pseudoscalar - meson leptonic decay constant in the sense that both are obtained as the chiral - limit value of well - defined gauge - invariant hadron - to - vacuum transition amplitudes that possess a spectral representation in terms of the current - quark - mass . thus , whereas it might sometimes be convenient to imagine otherwise , neither is essentially a constant mass - scale that fills all spacetime . hence , in particular , the quark condensate can be understood as a property of hadrons themselves a perspective also advocated in ref.@xcite and now established for all hadrons @xcite which is expressed , for example , in their bethe - salpeter or light - front wave functions . in the latter instance , the light - front - instantaneous quark propagator appears to play a crucial role @xcite . this has enormous implications for the cosmological constant . the universe is expanding at an ever - increasing rate and theoretical physics has tried to explain this in terms of the energy of quantum processes in vacuum ; viz . , condensates carry energy and so , if they exist , must contribute to the universe s dark energy , which is expressed in the computed value of the cosmological constant . the problem is that hitherto all potential sources of this so - called vacuum energy give magnitudes that far exceed the value of the cosmological constant that is empirically determined . this has been described as `` the biggest embarrassment in theoretical physics '' @xcite . however , given that , in the presence of confinement , condensates do not leak from within hadrons , then there are no space - time - independent condensates permeating the universe @xcite . this nullifies completely their contribution to the cosmological constant and reduces the mismatch between theory and observation by a factor of @xmath104 @xcite , and possibly by far more , if technicolour - like theories are the correct paradigm for extending the standard model . in connection with electric - charge - neutral pseudoscalar mesons , eq.([mrtrelation ] ) is strongly modified owing to the non - abelian anomaly . this entails that whilst the classical action associated with qcd is invariant under @xmath105 ( non - abelian axial transformations generated by @xmath106 , where @xmath107 $ ] ) , the quantum field theory is not . the modification is particularly important to properties of @xmath0 and @xmath1 mesons . the latter is obviously a peculiar pseudoscalar meson because its mass is far greater than that of any other light - quark pseudoscalar meson ; e.g. , @xmath108 . we note that the diagram depicted in fig.[fig : nonanomaly ] is often cited as central to a solution of the @xmath0-@xmath1 puzzle . however , as will become clear below , whilst it does contribute to flavour - mixing , the process is immaterial in resolving the @xmath0-@xmath1 conundrum . this simple flavour - mixing diagram is immaterial to the resolution of the @xmath0-@xmath1 conundrum , as is any collection of processes for which the figure may serve as a skeleton diagram . ( straight lines denote quarks and springs denote gluons . ) , scaledwidth=50.0% ] the correct mass formula for flavourless pseudoscalars follows from consideration of the complete @xmath105 ward - takahashi identity : @xmath109 which generalises eq.([avwtimn ] ) . in eq.([avwtig ] ) , @xmath110 are the generators of @xmath111 in the fundamental representation , orthonormalised according to tr@xmath112 ; the dressed - quark propagator @xmath113diag@xmath114 $ ] is matrix - valued ; and @xmath115,\ ] ] where @xmath116 is a matrix of renormalised current - quark masses and the trace is over flavour indices . the final term in the last line of eq.([avwtig ] ) expresses the non - abelian axial anomaly . it can be written @xmath117 and since @xmath118 is a pseudoscalar , it has the general form @xmath119.\end{aligned}\ ] ] the matrix element in eq.([au ] ) represents an operator expectation value in full qcd ; the operation in eq.([amputate ] ) amputates the external quark lines ; and @xmath120 \label{topq } = \partial_\mu k_\mu(x)\ ] ] is the topological charge density operator , where the trace is over colour indices and @xmath121 is the matrix - valued gluon field strength tensor . it is plain and important that only @xmath122 is nonzero . nb . while @xmath123 is gauge invariant , the associated chern - simons current , @xmath124 , is not . thus in qcd no physical state can couple to @xmath124 and hence no state which appears in the observable spectrum can contribute to a resolution of the so - called @xmath125-problem ; i.e. , physical states can not play any role in ensuring that the @xmath1 is not a goldstone mode . as described in sec.[psmassformula ] , if one imagines there are @xmath18 massless quarks , then dcsb is a necessary and sufficient condition for the @xmath126 components of eq.([avwtig ] ) to guarantee the existence of @xmath127 massless bound - states of a dressed - quark and -antiquark . owing to eq.([amputate ] ) , @xmath128 in eq.([avwtig ] ) requires special consideration . one case is easily covered ; viz . , it is clear that if @xmath129 , then the @xmath128 component of eq.([avwtig ] ) is no different to the others and there is an additional massless bound - state in the chiral limit . on the other hand , the large disparity between the mass of the @xmath1-meson and the octet pseudoscalars suggests that @xmath130 in real - world qcd . if one carefully considers that possibility , then the goldberger - treiman relations in eqs.([gtlrele ] ) ( [ gtlrelh ] ) become @xcite @xmath131 it follows that the relationship @xmath132 is necessary and sufficient to guarantee that @xmath133 , the flavourless pseudoscalar vertex , does not possess a massless pole in the chiral limit ; i.e. , that there are only @xmath127 massless goldstone bosons . now , in the chiral limit , @xmath134 if , and only if , chiral symmetry is dynamically broken . hence , the absence of an additional massless bound - state is only assured through the existence of an intimate connection between dcsb and an expectation value involving the topological charge density . this critical connection is further highlighted by the following result , obtained through a few straightforward manipulations of eqs.([avwtig ] ) , ( [ amputate ] ) and ( [ au ] ) : @xmath135 the absence of a goldstone boson in the @xmath128 channel is only guaranteed if this explicit identity between the chiral - limit in - meson condensate and a mixed vacuum polarisation involving the topological charge density is satisfied . mass formulae valid for all pseudoscalar mesons have also been obtained @xcite @xmath136 where @xmath137 for charged pseudoscalar mesons , eq.([newmass ] ) is equivalent to eq.([mrtrelation ] ) , but the novelty of eq.([newmass ] ) is what it expresses for flavourless pseudoscalars . to illustrate , consider the case of a @xmath138-symmetric mass matrix , in which all @xmath139 current - quark masses assume the single value @xmath41 , then this formula yields : @xmath140 plainly , the @xmath1 is split from the goldstone modes so long as @xmath141 . numerical simulations of lattice - qcd have confirmed this identity @xcite . it is important to elucidate the physical content of @xmath142 . returning to the definition : @xmath143 it is readily seen to be another type of in - meson condensate . it is analogous to those discussed in sec.[inhadroncondensate ] but in this case the hadron - to - vacuum transition amplitude measures the topological content of the @xmath1 . one may therefore state that the @xmath1 is split from the goldstone modes so long as its wavefunction possesses nonzero topological content . this is plainly very different to requiring that the qcd vacuum is topologically nontrivial . an illustration of the nature of the contribution to the bethe - salpeter kernel associated with the non - abelian anomaly . all terms have the `` hairpin '' structure illustrated in the lower panel . no finite sum of such intermediate states is sufficient . ( straight lines denote quarks , with @xmath144 and @xmath145 independent , and springs denote gluons . ) ( figure adapted from ref.@xcite.),scaledwidth=70.0% ] within qcd the properties of the @xmath1 can be computed via the bse , just like other mesons . a nonzero value of @xmath142 can be achieved with a bethe - salpeter kernel that contains the contribution depicted in fig.[etaglue ] because one may argue from eqs.([au ] ) and ( [ topq ] ) that an anomaly - related contribution to a meson s bethe - salpeter kernel can not contain external quark or antiquark lines that are connected to the incoming lines : purely gluonic configurations must mediate , as illustrated in fig.[etaglue ] . furthermore , it is straightforward to see that no finite sum of gluon exchanges can serve this purpose . indeed , consider any one such single contribution in the chiral limit . it will be proportional to the total momentum and hence vanish for @xmath146 , in conflict with eq.([etapchiral ] ) . this lies behind the need for something like the kogut - susskind _ ghost _ ; i.e. , the coupling of a massless axial - vector gauge - like field to the chern - simons current , which does not appear in the particle spectrum of qcd because the current is not gauge invariant . ( see ref.@xcite and sec.5.1 of ref.@xcite . ) it is argued @xcite that in qcd with @xmath147 colours , @xmath148 and it can be seen to follow from the gap equation , the homogeneous bse and eqs.([fpia ] ) , ( [ cpres ] ) that @xmath149 one thus obtains @xmath150 the first term vanishes in the limit @xmath151 while the second remains finite . subsequently taking the chiral limit , the @xmath1 mass approaches zero in the manner characteristic of all goldstone modes . ( nb . one must take the limit @xmath151 before the chiral limit because the procedures do not commute @xcite . ) these results are realised in the effective lagrangian of ref.@xcite in a fashion that is consistent with all the constraints of the anomalous ward identity . this is not true of the so - called thooft determinant @xcite . implications of the mass formula in eq.([newmass ] ) were exemplified in ref.@xcite using an elementary dynamical model that includes a one - parameter _ ansatz _ for that part of the bethe - salpeter kernel related to the non - abelian anomaly , an illustration of which is provided in fig.[etaglue ] . the study compares ground - state pseudoscalar- and vector - mesons constituted from all known quarks , excluding the @xmath152-quark . amongst the notable results is a prediction for the mixing angles between neutral mesons ; e.g. , @xmath153 nb . there are necessarily two mixing angles , with each determined at the appropriate pole position in the inhomogeneous vertex . it is interesting that the angles are approximately equal and compare well with the value inferred from a single mixing angle analysis @xcite @xmath154 . it is worth explicating the nature of the flavour - induced difference between the @xmath155 and @xmath156 masses . if one ignores mixing with mesons containing other than @xmath157-quarks ; viz . , works solely within @xmath158 , then @xmath159mev . on the other hand , the full calculation yields @xmath160mev , a factor of ten greater , and one obtains a @xmath155-@xmath0 mixing angle , whose value at the neutral pion mass shell is @xmath161 for comparison , ref.@xcite infers a mixing angle of @xmath162 from a @xmath62-matrix analysis of the process @xmath163he@xmath164 . plainly , mixing with the @xmath0-meson is the dominant non - electromagnetic contribution to the @xmath156-@xmath155 mass splitting . the analogous angle at the @xmath0 mass - shell is @xmath165 the angles in eq.([valmixing ] ) correspond to latexmath:[\ ] ] wherein the bethe - salpeter amplitudes and dressed - quark propagators are discussed above and @xmath404 is a generalization of the dressed - quark - photon vertex , describing a dressed - quark scattering from a zero momentum photon . it satisfies a bse ( here with a rainbow - ladder kernel ) with the inhomogeneous term @xmath405 . here @xmath406 is a light - like vector satisfying . in choosing rainbow - ladder truncation one implements a precise parallel to the symmetry - preserving treatment of the pion charge form factor at , wherein the vector current is conserved by use of ladder dynamics at all three vertices and rainbow dynamics for all three quark propagators @xcite . equation ( [ eucl_pdf_lr_ward ] ) ensures automatically that @xmath407 since @xmath408 gives the ward - identity vertex and the bethe - salpeter amplitudes are canonically normalised . . _ solid curve _ full dse calculation @xcite ; _ dot - dashed curve _ semi - phenomenological dse - based calculation in ref.@xcite ; _ filled circles _ experimental data from ref.@xcite , at scale ( 4.05gev)@xmath319 ; _ dashed curve _ nlo reanalysis of the experimental data @xcite ; and _ dot - dot - dashed curve _ nlo reanalysis of experimental data with inclusion of soft - gluon resummation @xcite . ( adapted from ref.@xcite . ) [ fig : pi_dse],scaledwidth=66.0% ] figure [ fig : pi_dse ] displays the dse results for the pion s valence @xmath179-quark distribution , evolved from a resolving scale @xmath409gev)@xmath319 to @xmath410 using leading - order evolution , and a comparison with @xmath411 drell - yan data @xcite at a scale @xmath412 , inferred via a leading - order analysis . the computation s resolving scale , @xmath413 , was fixed by matching the @xmath414 moments for @xmath415 to an experimental analysis at ( 2gev)@xmath319 @xcite . it is notable that at @xmath413 the dse results yield @xmath416 ( for comparison , the parametrised valence - like pion parton distributions of ref.@xcite yield a gluon momentum fraction of @xmath417 . ) in each case the remainder of the hadron s momentum is carried by gluons , which effect binding of the meson bound state and are invisible to the electromagnetic probe . some fraction of the hadron s momentum is carried by gluons at all resolving scales unless the hadron is a point particle @xcite . indeed , it is a simple algebraic exercise to demonstrate that the only non - increasing , convex function which can produce @xmath418 and @xmath419 , is the distribution @xmath420 , which is uniquely connected with a pointlike meson ; viz . , a meson whose bethe - salpeter amplitude is momentum - independent . thus eqs.([momcons ] ) are an essential consequence of momentum conservation . whilst the dse results in fig.[fig : pi_dse ] are both consistent with eq.([pqcduvx ] ) ; i.e. , they produce algebraically the precise behaviour predicted by perturbative qcd , on the valence - quark domain it is evident that they disagree markedly with the drell - yan data reported in ref.@xcite . this tension was long seen as a crucial mystery for a qcd description of the lightest and subtlest hadron @xcite . its re - emergence with ref.@xcite motivated a nlo reanalysis of the drell - yan data @xcite , the result of which is also displayed in fig.[fig : pi_dse ] . at nlo the extracted pdf is softer at high-@xmath395 but the discrepancy nevertheless remains . to be precise , ref.@xcite determined a high-@xmath395 exponent of @xmath421 whereas the exponents produced by the dse studies @xcite are , respectively , @xmath422 and @xmath423 at the common model scale . they do not allow much room for a harder pdf at high-@xmath395 . following the highlighting of this discrepancy in ref.@xcite , a resolution of the conflict between data and well - constrained theory was proposed . in ref.@xcite a long - overlooked effect was incorporated ; namely , `` soft - gluon resummation . '' with the inclusion of this next - to - leading - logarithmic threshold resummation effect in the calculation of the drell - yan cross section , a considerably softer valence - quark distribution was obtained at high-@xmath395 . this is readily understood . the drell - yan cross - section factorises into two pieces : one hard and the other soft . the soft piece involves the pdf and the hard piece is calculable in perturbation theory . adding additional interactions to the latter , which are important at large-@xmath395 ; viz . , soft gluons , provides greater strength in the hard piece on the valence - quark domain . hence a description of the data is obtained with a softer pdf . indeed , the distribution obtained thereby matches precisely the expectations based on perturbative - qcd and obtained using dses . this is evident in a comparison between the _ dash - dot _ and _ dash - dot - dot _ curves in fig.[fig : pi_dse ] . dse prediction for the ratio of @xmath179-quark distributions in the kaon and pion @xcite . the full bethe - salpeter amplitude produces the _ solid _ curve ; a reduced bse vertex produces the _ dashed _ curve . the reduced amplitude retains only the invariants and amplitudes involving pseudoscalar and axial vector dirac matrices , and ignores dependence on the variable @xmath424 . these are part of the reductions that occur in a pointlike treatment of pseudoscalar mesons . the experimental data is from @xcite . ( adapted from ref.@xcite.),scaledwidth=66.0% ] the ratio @xmath425 measures the effect of the local hadronic environment . in the kaon , the @xmath179-quark is bound with a heavier partner than in the pion and this should cause @xmath426 to peak at lower-@xmath395 than @xmath427 . the dse prediction @xcite is shown in fig.[fig : pi_dse_ratio ] along with available drell - yan data @xcite . the parameter - free dse result agrees well with the data . we note that @xmath428 viz , the ratio approaches one under evolution to larger resolving scales owing to the increasingly large population of sea - quarks produced thereby @xcite . on the other hand , the value at @xmath429 is a fixed - point under evolution : @xmath430 i.e. , it is the same at every value of the resolving scale @xmath323 , and is therefore a persistent probe of nonperturbative dynamics @xcite . with ref.@xcite a significant milestone was achieved ; , unification of the computation of distribution functions that arise in analyses of deep inelastic scattering with that of numerous other properties of pseudoscalar mesons , including meson - meson scattering @xcite and the successful prediction of electromagnetic elastic and transition form factors . the results confirm the large-@xmath395 behavior of distribution functions predicted by the qcd parton model ; provide a good account of the @xmath431-@xmath432 drell - yan data for @xmath427 ; and a parameter - free prediction for the ratio @xmath402 that agrees with extant data , showing a strong environment - dependence of the @xmath179-quark distribution . the new drell - yan experiment running at fnal is capable of validating this comparison , as is the compass ii experiment at cern . such an experiment should be done so that complete understanding of qcd s goldstone modes can be claimed . while a symmetry - preserving description of mesons is essential , it is only part of the physics that nonperturbative qcd must describe because nature also presents us with baryons : light - quarks in three - particle composites . an explanation of the spectrum of baryons and the nature of interactions between them is basic to understanding the standard model . the present and planned experimental programmes at jlab , and other facilities worldwide , are critical elements in this effort . no approach to qcd is comprehensive if it can not provide a unified explanation of both mesons and baryons . we have explained that dcsb is a keystone of the standard model , which is evident in the momentum - dependence of the dressed - quark mass function fig.[gluoncloud ] : it is just as important to baryons as it is to mesons . the dses furnish the only extant framework that can simultaneously connect both meson and baryon observables with this basic feature of qcd , having provided , e.g. , a direct correlation of meson and baryon properties via a single interaction kernel , which preserves qcd s one - loop renormalisation group behaviour and can systematically be improved @xcite . in quantum field theory a baryon appears as a pole in a six - point quark green function . the residue is proportional to the baryon s faddeev amplitude , which is obtained from a poincar covariant faddeev equation that sums all possible exchanges and interactions that can take place between three dressed - quarks . a tractable faddeev equation for baryons @xcite is founded on the observation that an interaction which describes colour - singlet mesons also generates nonpointlike quark - quark ( diquark ) correlations in the colour-@xmath433 ( antitriplet ) channel @xcite . the dominant correlations for ground state octet and decuplet baryons are scalar ( @xmath434 ) and axial - vector ( @xmath435 ) diquarks because , for example , the associated mass - scales are smaller than the baryons masses @xcite and their parity matches that of these baryons . it follows that only they need be retained in approximating the quark - quark scattering matrix which appears as part of the faddeev equation @xcite . we note that diquarks do not appear in the strong interaction spectrum @xcite but the attraction between quarks in this channel justifies a picture of baryons in which two quarks within a baryon are always correlated as a colour-@xmath433 diquark pseudoparticle , and binding is effected by the iterated exchange of roles between the bystander and diquark - participant quarks . here it is important to emphasise strongly that qcd supports _ nonpointlike _ diquark correlations @xcite . hence models that employ pointlike diquark degrees of freedom have little connection with qcd . numerous properties of the nucleon have recently been described using the faddeev equation just outlined , e.g. : a survey of the form factors @xcite , with additional results explicated in refs.@xcite ; indications of the role of quark orbital angular momentum in forming the nucleon s spin @xcite ; and computation of the nucleon s @xmath64-term @xcite . this body of work will be reviewed elsewhere . however , much more can and should be done , as described , for example , in sec.iii of ref.@xcite ; and ref.@xcite is a first step . it is expected that feedback between dse results and extant and forthcoming precision data on nucleon elastic and transition form factors can serve as a practical means by which to empirically chart the momentum evolution of the dressed - quark mass function , and therefrom the infrared behavior of qcd s @xmath4-function . in particular , it should enable the unambiguous location of the transition boundary between the constituent- and current - quark domains , signalled by the sharp drop with increasing momentum that is apparent in fig.[gluoncloud ] , which can likely be related to an infrared inflexion point in qcd s running coupling , whose properties are determined by the @xmath4-function . dynamical chiral symmetry breaking ( dcsb ) is a fact in qcd . it is manifest in dressed - propagators and vertices , and , amongst other things , it is responsible for : the transformation of the light current - quarks in qcd s lagrangian into heavy constituent - like quark s , in terms of which order was first brought to the hadron spectrum ; the unnaturally small values of the masses of light - quark pseudoscalar mesons ; the unnaturally strong coupling of pseudoscalar mesons to light - quarks @xmath436 ; and the unnaturally strong coupling of pseudoscalar mesons to the lightest baryons @xmath437 . herein we have used a diverse range of phenomena to illustrate the dramatic impact that dcsb has upon meson observables , amongst them : the spectrum ; form factors ; and parton distribution functions . unavoidably , many valuable contributions during the last few years were overlooked , some of which may be pursued through refs.@xcite . we have also completely neglected an extensive body of work that focuses on elucidating the nature of the primordial state of matter that has been recreated with rhic : refs.@xcite serve as a starting point for exploring this area of dse research . ( as noted in ref.@xcite , the rainbow truncation of qcd s gap equation excludes what in a point - meson field theory would appear as @xmath438-suppressed meson - loop dressing of the quark propagator , and this is why chiral symmetry restoration is a mean field transition in all models within the rainbow - truncation class @xcite . such meson - loop corrections are capable of correcting the critical exponents ; i.e. , of introducing that infrared divergence in the dressed - quark self - energy which can force a deviation from mean field critical exponents in the chiral symmetry restoring transition . this is illustrated in ref.@xcite . ) in addition , there are attempts to derive an equation of state that may be used in modern astrophysics in order , e.g. , to assist in the identification of a neutron star with a quark - matter core @xcite . a `` smoking gun '' for dcsb is the behaviour of the dressed - quark mass function . the momentum dependence manifest in fig.[gluoncloud ] is an essentially quantum field theoretical effect . exposing and elucidating its consequences therefore requires a nonperturbative and symmetry - preserving approach , where the latter means preserving poincar covariance , chiral and electromagnetic current - conservation , etc . the dyson - schwinger equations ( dses ) provide such a framework . experimental and theoretical studies are underway that will identify observable signals of @xmath24 and thereby confirm and explain the mechanism responsible for the vast bulk of visible mass in the universe . there are many reasons why this is an exciting time in hadron physics . we have focused on one . namely , through the dses , we are positioned to unify phenomena as apparently diverse as : the hadron spectrum ; hadron elastic and transition form factors , from small- to large-@xmath323 ; and parton distribution functions . the key is an understanding of both the fundamental origin of nuclear mass and the far - reaching consequences of the mechanism responsible ; namely , dcsb . these things might lead us to an explanation of confinement , the phenomenon that makes nonperturbative qcd the most interesting piece of the standard model . we acknowledge valuable input from a. bashir , i.c . clot , b. el - bennich , l. x. gutirrez - guerrero , a. kizilers , y .- x . liu , s .- x . qin , j. rodriguez - quintero , s.m . schmidt and d. j. wilson . this work was supported by : the u.s . department of energy , office of nuclear physics , contract no . de - ac02 - 06ch11357 ; forschungszentrum jlich gmbh ; and the u.s . national science foundation , under grant no . phy-0903991 , in conjunction with a conacyt mexico - usa collaboration grant . c. d. roberts and a. g. williams , prog . . phys . * 33 * , pp . 477 - 575 ( 1994 ) . c. d. roberts and s. m. schmidt , prog . * 45 * , pp . s1-s103 ( 2000 ) . maris and c. d. roberts , int . j. mod . e * 12 * , pp . 297 - 365 ( 2003 ) . m. r. pennington , j. phys . ser . * 18 * , pp . 1 - 73 ( 2005 ) . a. hll , c. d. roberts and s. v. wright , `` hadron physics and dyson - schwinger equations , '' nucl - th/0601071 , contribution to the _ proceedings of the 20th annual hampton university graduate studies program ( hugs 2005 ) _ , newport news , virginia , 31 may - 17 jun 2005 , 103 pages . c. s. fischer , j. phys . g * 32 * , pp . r253-r291 ( 2006 ) . c. d. roberts , m. s. bhagwat , a. hll and s. v. wright , eur . j. st * 140 * , pp . 53 - 116 ( 2007 ) . c. d. roberts , prog . . phys . * 61 * , pp . 50 - 65 ( 2008 ) . r. j. holt and c. d. roberts , rev . phys . * 82 * , pp . 2991 - 3044 ( 2010 ) . e. s. swanson , aip conf . * 1296 * , pp . 75 - 121 ( 2010 ) . m. s. bhagwat , i. c. clot and c. d. roberts , `` covariance , dynamics and symmetries , and hadron form factors , '' in _ exclusive reactions at high momentum transfer _ , eds . a. radyushkin and p. stoler ( world scientific , singapore , 2008 ) pp . 112 - 120 ; arxiv:0710.2059 [ nucl - th ] . m. s. bhagwat , m. a. pichowsky , c. d. roberts and p. c. tandy , phys . c * 68 * , 015203 ( 2003 ) [ 9 pages ] . m. s. bhagwat and p. c. tandy , aip conf . proc . * 842 * , pp . 225 - 227 ( 2006 ) . o. bowman _ et al_. , phys . d * 71 * , 054507 ( 2005 ) [ 7 pages ] . v. v. flambaum _ et al_. , few body syst . * 38 * pp . 31 - 51 ( 2006 ) . a. hll , p. maris , c. d. roberts and s. v. wright , nucl . suppl . * 161 * , pp . 87 - 94 ( 2006 ) . a. bashir , a. raya , i. c. clot and c. d. roberts , phys . c * 78 * , 055201 ( 2008 ) [ 7 pages ] . a. bashir , a. raya , s. sanchez - madrigal and c. d. roberts , few body syst . * 46 * , pp . 229 - 237 ( 2009 ) . c. p. hofmann , a. raya and s. s. madrigal , phys . d * 82 * , 096011 ( 2010 ) [ 8 pages ] . a. bender , d. blaschke , yu.l . kalinovsky and c. d. roberts , phys . lett . * 77 * , pp . 3724 - 3727 ( 1996 ) . d. blaschke , c. d. roberts and s. m. schmidt , phys . b * 425 * , pp . 232 - 238 ( 1998 ) a. bender _ et al_. , phys . b * 431 * , pp . 263 - 269 ( 1998 ) . h. chen _ et al_. , phys . d * 78 * , 116015 ( 2008 ) [ 11 pages ] . c. s. fischer and j. a. mueller , phys . d * 80 * , 074029 ( 2009 ) [ 14 pages ] . s .- x . qin _ et al_. , phys . lett . * 106 * , 172301 ( 2011 ) [ 4 pages ] . s .- x . qin , l. chang , y .- x . liu and c. d. roberts , phys . d**84 * * , 014017 ( 2011 ) [ 6 pages ] . y .- x . liu , s .- x . qin , l. chang , c. d. roberts , aip conf . proc . * 1354 * , pp . 91 - 96 ( 2011 ) . i. c. clot _ et al_. , few body syst . * 46 * , pp . 1 - 36 ( 2009 ) . m. s. bhagwat , a. krassnigg , p. maris and c. d. roberts , eur . phys . j. a * 31 * , pp . 630 - 637 ( 2007 ) . m. s. bhagwat and p. maris , phys . c * 77 * , 025203 ( 2008 ) [ 12 pages ] . i. c. clot , a. krassnigg and c. d. roberts , in _ proceedings of 11th international conference on meson - nucleon physics and the structure of the nucleon ( menu 2007 ) _ , jlich , germany , 10 - 14 sep 2007 , 125/1 - 11 , arxiv:0710.5746 [ nucl - th ] . s. j. brodsky and h. c. pauli , lect . notes phys . * 396 * , pp . 51 - 121 ( 1991 ) . s. j. brodsky , c. d. roberts , r. shrock and p. c. tandy , phys . rev . c * 82 * , 022201 ( 2010 ) [ 5 pages ] . s. j. brodsky and r. shrock , proc . sci . * 108 * , pp . 45 - 50 ( 2011 ) . l. chang , c. d. roberts and p. c. tandy , `` expanding the concept of in - hadron condensates , '' arxiv:1109.2903 [ nucl - th ] . g. s. bali _ et al_. , phys . d * 71 * , 114513 ( 2005 ) [ 26 pages ] . l. chang _ et al_. , chin . c * 33 * , pp . 1189 - 1196 ( 2009 ) . articles cited in : v. n. gribov , eur . j. c * 10 * , pp . 91 - 105 ( 1999 ) . h. j. munczek and a. m. nemirovsky , phys . rev . d * 28 * , pp . 181 - 186 ( 1983 ) . m. stingl , phys . d * 29 * , pp . 2105 - 2117 ( 1984 ) . r. t. cahill , austral . j. phys . * 42 * , pp . 171 - 186 ( 1989 ) . g. krein , c. d. roberts and a. g. williams , int . j. mod . phys . a * 7 * , pp . 5607 - 5624 ( 1992 ) . r. f. streater and a. s. wightman , _ pct , spin and statistics , 3rd edition _ ( addison - wesley , reading , massachussets , 1980 ) . f. t. hawes , c. d. roberts and a. g. williams , phys . rev . d * 49 * , pp . 4683 - 4693 ( 1994 ) . p. maris , phys . d * 52 * , pp . 6087 - 6097 ( 1995 ) . boucaud , j.p . leroy , a. le yaouanc , j. micheli , o. pne and j. rodriguez - quintero , `` the infrared behaviour of the pure yang - mills green functions , '' to appear in few body syst . , arxiv:1109.1936 [ hep - ph ] . f. d. r. bonnet , p. o. bowman , d. b. leinweber and a. g. williams , phys . rev . d * 62 * , 051501(r ) ( 2000 ) [ 4 pages ] . j. i. skullerud and a. g. williams , phys . d * 63 * , 054508 ( 2001 ) [ 13 pages ] . w. kamleh , p. o. bowman , d. b. leinweber , a. g. williams and j. zhang , phys . d * 76 * , 094501 ( 2007 ) [ 9 pages ] . w. celmaster and r. j. gonsalves , phys . d * 20 * , pp . 1420 - 1434 ( 1979 ) . l. chang _ et al_. , phys . c * 79 * , 035209 ( 2009 ) [ 9 pages ] . p. pascual and r. tarrach , _ qcd : renormalization for the practitioner _ , lecture notes in physics * 194 * ( springer - verlag , berlin , 1984 ) . e. e. salpeter and h. a. bethe , phys . rev . * 84 * , pp . 1232 - 1252 ( 1951 ) . h. j. munczek , phys . d * 52 * , pp . 4736 - 4740 ( 1995 ) . a. bender , c. d. roberts and l. von smekal , phys . lett . b * 380 * , pp . 7 - 12 ( 1996 ) . a. hll , a. krassnigg and c. d. roberts , phys . c * 70 * , 042203(r ) ( 2004 ) [ 5 pages ] . a. hll _ et al_. , phys . c * 71 * , 065204 ( 2005 ) [ 12 pages ] . m. a. ivanov , yu . l. kalinovsky and c. d. roberts , phys . d * 60 * , 034018 ( 1999 ) [ 17 pages ] . l. chang _ et al_. , phys . c * 81 * , 032201(r ) ( 2010 ) [ 4 pages ] l. chang , y. x. liu , w. m. sun and h. s. zong , phys . b * 669 * , pp . 327 - 330 ( 2008 ) . s. weinberg , phys . lett . * 18 * , pp . 507 - 509 ( 1967 ) . p. maris , c. d. roberts and p. c. tandy , phys . b * 420 * , pp . 267 - 273 ( 1998 ) . r. delbourgo and m. d. scadron , j. phys . g * 5 * , pp . 1621 - 1632 ( 1979 ) . maris and c. d. roberts , phys . c * 58 * pp . 3659 - 3665 ( 1998 ) . m. b. hecht , c. d. roberts and s. m. schmidt , phys . c * 63 * , 025213 ( 2001 ) [ 8 pages ] . l. x. gutirrez - guerrero , a. bashir , i. c. clot and c. d. roberts , phys . c * 81 * , 065202 ( 2010 ) [ 5 pages ] . h. l. l. roberts _ et al_. , phys . c * 82 * , 065202 ( 2010 ) [ 10 pages ] . h. l. l. roberts _ et al_. , phys . c * 83 * , 065206 ( 2011 ) [ 12 pages ] . t. nguyen , a. bashir , c. d. roberts and p. c. tandy , phys . c * 83 * , 062201(r ) ( 2011 ) [ 5 pages ] . p. maris and p. c. tandy , phys . c * 60 * , 055214 ( 1999 ) [ 15 pages ] . p. maris and p. c. tandy , phys . c * 61 * , 045202 ( 2000 ) [ 11 pages ] . p. maris and p. c. tandy , phys . c * 62 * , 055204 ( 2000 ) [ 8 pages ] . j. p. b. de melo , t. frederico , e. pace and g. salme , phys . d * 73 * , 074013 ( 2006 ) [ 35 pages ] . m. diehl and g. hiller , jhep * 0106 * , 067 ( 2001)/1 - 36 . s. x. qin _ et al_. , `` commentary on rainbow - ladder truncation for excited states and exotics , '' arxiv:1109.3459 [ nucl - th ] . c. mcneile and c. michael [ ukqcd collaboration and ukqcd collaboration ] , phys . b * 642 * , pp . 244 - 247 ( 2006 ) p. colangelo and a. khodjamirian , `` qcd sum rules , a modern perspective , '' arxiv : hep - ph/0010175 . h. pagels and s. stokar , phys . d * 20 * , pp . 2947 - 2952 ( 1979 ) . r. t. cahill and c. d. roberts , phys . d * 32 * , pp . 2419 - 2428 ( 1985 ) . l. chang and c. d. roberts , phys . * 103 * , 081601 ( 2009 ) [ 4 pages ] . j. bijnens , prog . part . * 58 * , pp . 521 - 586 ( 2007 ) . maris and c. d. roberts , phys . c * 56 * , pp . 3369 - 3383 ( 1997 ) . k. langfeld _ et al_. , phys . c * 67 * , 065206 ( 2003 ) [ 7 pages ] . k. d. lane , phys . d * 10 * , pp . 2605 - 2618 ( 1974 ) . h. d. politzer , nucl . b * 117 * , pp . 397 - 406 ( 1976 ) . t. banks and a. casher , nucl . b * 169 * , pp . 103 - 125 ( 1980 ) . a. casher and l. susskind , phys . d * 9 * , pp . 436 - 460 ( 1974 ) . m. burkardt , phys . d * 58 * , 096015 ( 1998 ) [ 11 pages ] . m. s. turner , `` dark energy and the new cosmology , '' astro - ph/0108103 . m. s. bhagwat _ et al_. , phys . c * 76 * , 045203 ( 2007 ) [ 10 pages ] . w. a. bardeen , a. duncan , e. eichten and h. thacker , phys . d * 62 * , 114505 ( 2000 ) [ 16 pages ] . s. ahmad , j. t. lenaghan and h. b. thacker , phys . d * 72 * , 114511 ( 2005 ) [ 12 pages ] . j. b. kogut and l. susskind , phys . d * 11 * , pp . 3594 - 3610 ( 1975 ) . g. a. christos , phys . rept . * 116 * , pp . 251 - 336 ( 1984 ) . e. witten , nucl . b * 156 * , pp . 269 - 283 ( 1979 ) . g. veneziano , nucl . b * 159 * , pp . 213 - 224 ( 1979 ) . r. narayanan and h. neuberger , nucl . b * 696 * , pp . 107 - 140 ( 2004 ) . p. di vecchia , phys . b * 85 * , pp . 357 - 360 ( 1979 ) . r. j. crewther , phys . b * 70 * , pp . 349 - 354 ( 1977 ) . r. j. crewther , acta phys . austriaca suppl . * 19 * , pp . 47 - 153 ( 1978 ) . c. bini , eur . phys . j. a * 31 * , pp . 446 - 450 ( 2007 ) . a. m. green and s. wycech , phys . rev . c * 68 * , 061601(r ) ( 2003 ) [ 5 pages ] . c. d. roberts , `` confinement , diquarks and goldstone s theorem , '' in _ como 1996 , quark confinement and the hadron spectrum ii _ , eds . n. brambilla and g.m . prosperi ( world scientific , new jersey , 1997 ) , pp . 224 - 230 , proceedings of the `` 2nd international conference on quark confinement and the hadron spectrum , '' como , italy , 26 - 29 jun 1996 , arxiv : nucl - th/9609039 . c. d. roberts , phys . nucl . * 30 * , pp . 223 - 257 ( 1999 ) [ fiz . chast . atom . yadra * 30 * , pp . 537 - 612 ( 1999 ) ] . a. bender , w. detmold , a.w . thomas and c.d . roberts , phys . c * 65 * , 065203 ( 2002 ) [ 16 pages ] . m. s. bhagwat _ et al_. , phys . c * 70 * , 035205 ( 2004 ) [ 15 pages ] . c. j. burden _ et al_. , phys . c * 55 * , pp . 2649 - 2664 ( 1997 ) . p. watson , w. cassing and p. c. tandy , few body syst . * 35 * , pp . 129 - 153 ( 2004 ) . p. maris , aip conf . proc . * 892 * , pp . 65 - 71 ( 2007 ) . c. s. fischer and r. williams , phys . lett . * 103 * , 122001 ( 2009 ) [ 4 pages ] . a. krassnigg , phys . d * 80 * , 114010 ( 2009 ) [ 8 pages ] . a. krassnigg and m. blank , phys . d * 83 * , 096006 ( 2011 ) [ 7 pages ] . a. hll , a. krassnigg , c. d. roberts and s. v. wright , int . j. mod . a * 20 * , pp . 1778 - 1784 ( 2005 ) j. s. ball and t. w. chiu , phys . d * 22 * , pp . 2542 - 2549 ( 1980 ) . a. kizilers and m. r. pennington , phys . d * 79 * , 125020 ( 2009 ) [ 25 pages ] . j. r. pelaez and g. rios , phys . * 97 * ( 2006 ) 242002 [ 4 pages ] . j. ruiz de elvira , j. r. pelaez , m. r. pennington and d. j. wilson , `` chiral perturbation theory , the @xmath439 expansion and regge behaviour determine the structure of the lightest scalar meson , '' arxiv:1009.6204 [ hep - ph ] . h. m. foley and p. kusch , phys . rev . * 73 * , p. 412 j. s. schwinger , phys . rev . * 73 * , pp . 416 - 417 ( 1948 ) . p. j. mohr , b. n. taylor and d. b. newell , rev . phys . * 80 * , pp . 633 - 730 ( 2008 ) . l. chang , y. x. liu and c. d. roberts , phys . lett . * 106 * , 072001 ( 2011 ) [ 4 pages ] . a. i. davydychev , p. osland and l. saks , phys . d * 63 * , 014022 ( 2001 ) [ 31 pages ] . j. i. skullerud _ et al_. , jhep * 0304 * , 047 ( 2003 ) [ 16 pages ] . m. s. bhagwat and p. c. tandy , phys . d * 70 * , 094039 ( 2004 ) [ 6 pages ] . n. i. kochelev , phys . b * 426 * , pp . 149 - 153 ( 1998 ) . d. diakonov , prog . phys . * 51 * , pp . 173 - 222 ( 2003 ) . d. ebert , r. n. faustov and v. o. galkin , eur . j. c * 47 * , pp . 745 - 755 ( 2006 ) . l. chang and c. d. roberts , `` tracing masses of ground - state light - quark mesons , '' arxiv:1104.4821 [ nucl - th ] . j. a. bicudo , j. e. f. ribeiro and r. fernandes , phys . c * 59 * , pp . 1107 - 1112 ( 1999 ) . f. jegerlehner and a. nyffeler , phys . rept . * 477 * , pp . 1 - 110 ( 2009 ) . t. goecke , c. s. fischer and r. williams , phys . d * 83 * , 094006 ( 2011 ) [ 14 pages ] . m. s. bhagwat _ et al_. , few body syst . * 40 * , pp . 209 - 235 ( 2007 ) k. nakamura _ et al . _ , j. phys . g * 37 * 075021 ( 2010 ) [ 1422 pages ] . r. alkofer , p. watson and h. weigel , phys . d * 65 * , 094026 ( 2002 ) [ 10 pages ] . i. c. clot and c. d. roberts , pos * lc2008 * , 047 ( 2008 ) [ 10 pages ] . h. l. l. roberts , l. chang , i. c. clot and c. d. roberts , few body syst . * 51 * , pp . 1 - 12 ( 2011 ) . m. a. pichowsky , s. walawalkar and s. capstick , phys . d * 60 * ( 1999 ) 054030 [ 21 pages ] . g. eichmann _ et al_. , phys . c * 77 * , 042202(r ) ( 2008 ) [ 5 pages ] . j. j. dudek , `` the lightest hybrid meson supermultiplet in qcd , '' arxiv:1106.5515 [ hep - ph ] . j. volmer _ et al . _ , phys . lett . * 86 * , pp . 1713 - 1716 ( 2001 ) . t. horn _ et al . _ , phys . lett . * 97 * , 192001 ( 2006 ) [ 4 pages ] . v. tadevosyan _ et al . _ , phys . c * 75 * , 055205 ( 2007 ) [ 9 pages ] . k. wijesooriya , p. e. reimer and r. j. holt , phys . c * 72 * , 065203 ( 2005 ) [ 5 pages ] . g. r. farrar and d. r. jackson , phys . lett . * 43 * , pp . 246 - 249 ( 1979 ) . a. v. efremov and a. v. radyushkin , phys . b * 94 * , pp . 245 - 250 ( 1980 ) . g. p. lepage and s. j. brodsky , phys . d * 22 * , pp . 2157 - 2198 ( 1980 ) . o. bowman _ et al_. , phys . d * 70 * , 034509 ( 2004 ) [ 4 pages ] . d. ebert , t. feldmann and h. reinhardt , phys . b * 388 * , pp . 154 - 160 ( 1996 ) . c. d. roberts , nucl . a * 605 * , pp . 475 - 495 ( 1996 ) . z. f. ezawa , nuovo cim . a * 23 * , pp . 271 - 290 ( 1974 ) . g. r. farrar and d. r. jackson , phys . lett . * 35 * , pp . 1416 - 1419 ( 1975 ) . b. aubert _ et al . _ , phys . d * 80 * , 052002 ( 2009 ) [ 16 pages ] . h. j. behrend _ et al . _ , z. phys . c * 49 * , pp . 401 - 410 ( 1991 ) . j. gronberg _ et al . _ , phys . d * 57 * , pp . 33 - 54 ( 1998 ) . p. maris and p. c. tandy , phys . c * 65 * , 045211 ( 2002 ) [ 13 pages ] . a. v. radyushkin , phys . d * 80 * , 094009 ( 2009 ) [ 9 pages ] . m. v. polyakov , jetp lett . * 90 * , pp . 228 - 231 ( 2009 ) . s. v. mikhailov and n. g. stefanis , mod . a * 24 * , pp . 2858 - 2867 ( 2009 ) . s. j. brodsky , f. g. cao and g. f. de teramond , phys . rev . * d84 * , 033001 ( 2011 ) [ 14 pages ] . a. p. bakulev , s. v. mikhailov , a. v. pimikov and n. g. stefanis , phys . rev . * d84 * , 034014 ( 2011 ) [ 11 pages ] . s. j. brodsky , f. g. cao and g. f. de teramond , `` meson transition form factors in light - front holographic qcd , '' arxiv:1105.3999 [ hep - ph ] . b. aubert _ et al . _ , phys . d * 74 * , 012002 ( 2006 ) [ 14 pages ] . j. badier _ et al . b * 93 * , pp . 354 - 356 ( 1980 ) . j. badier _ et al . _ [ na3 collaboration ] , z. phys . c * 18 * , pp . 281 - 287 ( 1983 ) . b. betev _ et al . _ [ na10 collaboration ] , z. phys . c * 28 * , pp . 15 - 22 ( 1985 ) . j. s. conway _ et al . _ , phys . d * 39 * , pp . 92 - 122 ( 1989 ) . c. j. burden , c. d. roberts and m. j. thomson , phys . b * 371 * , pp . 163 - 168 ( 1996 ) . b. el - bennich , m. a. ivanov and c. d. roberts , phys . c * 83 * , 025205 ( 2011 ) [ 5 pages ] . m. aicher , a. schafer and w. vogelsang , phys . lett . * 105 * , 252003 ( 2010 ) [ 4 pages ] . p. j. sutton , a. d. martin , r. g. roberts and w. j. stirling , phys . d * 45 * , pp . 2349 - 2359 ( 1992 ) . m. gluck , e. reya and a. vogt , eur . j. c * 5 * , pp . 461 - 470 ( 1998 ) . l. chang and c. d. roberts , aip conf . proc . * 1261 * , pp . 25 - 30 ( 2010 ) . p. bicudo , `` proof of the weinberg theorem for chiral quark models , '' arxiv : nucl - th/0110052 . p. bicudo _ et al_. , phys . d * 65 * , 076008 ( 2002 ) [ 8 pages ] . g. eichmann _ et al_. , phys . c * 79 * , 012202(r ) ( 2009 ) [ 5 pages ] . g. eichmann , phys . rev . * d84 * , 014014 ( 2011 ) . v. mader , g. eichmann , m. blank and a. krassnigg , phys . rev . * d84 * , 034012 ( 2011 ) . r. t. cahill , c. d. roberts and j. praschifka , austral . j. phys . * 42 * , pp . 129 - 145 ( 1989 ) . r. t. cahill , c. d. roberts and j. praschifka , phys . d * 36 * , pp . 2804 - 2812 ( 1987 ) . p. maris , few body syst . * 32 * , pp . 41 - 52 ( 2002 ) . p. maris , few body syst . * 35 * , pp . 117 - 127 ( 2004 ) . i. c. clot _ et al_. , few body syst . * 42 * , pp . 91 - 113 ( 2008 ) . l. chang and c. d. roberts , `` hadron physics : the essence of matter , '' arxiv:1003.5006 [ nucl - th ] . h. l. l. roberts , l. chang , i. c. clot and c. d. roberts , `` exposing the dressed quark s mass , '' in _ exclusive reactions at high momentum transfer iv _ , ed . a. radyushkin ( world scientific , singapore , 2011 ) pp . 201 - 211 , proceedings of the `` 4th workshop on exclusive reactions at high momentum transfer , '' thomas jefferson national accelerator facility newport news , virginia , 18 - 21 may 2010 , arxiv:1007.3566 [ nucl - th ] l. chang , i. c. clot , c. d. roberts , h. l. l. roberts , aip conf . proc . * 1354 * , pp . 110 - 117 ( 2011 ) . i. c. clot , c. d. roberts and d. j. wilson , `` baryon properties from continuum - qcd , '' arxiv:1103.2432 [ nucl - th ] . i. aznauryan _ et al . _ , `` theory support for the excited baryon program at the jlab 12- gev upgrade , '' arxiv:0907.1901 [ nucl - th ] . m. r. pennington , aip conf . proc . * 1343 * , pp . 63 - 68 ( 2011 ) . m. blank , `` properties of quarks and mesons in the dyson - schwinger / bethe - salpeter approach , '' arxiv:1106.4843 [ hep - ph ] . m. he , j. f. li , w. m. sun and h. s. zong , phys . d * 79 * , 036001 ( 2009 ) [ 7 pages ] . t. klhn _ et al_. , phys . c * 82 * , 035801 ( 2010 ) [ 5 pages ] . h. li , x. l. luo , y. jiang and h. s. zong , phys . d * 83 * , 025012 ( 2011 ) [ 7 pages ] . p. maris , c. d. roberts , s. m. schmidt and p. c. tandy , phys . c * 63 * , 025202 ( 2001 ) [ 12 pages ] . a. hll , p. maris and c. d. roberts , phys . c * 59 * , pp . 1751 - 1755 ( 1999 ) c. s. fischer and j. a. mueller , `` on critical scaling at the qcd @xmath440 chiral phase transition , '' arxiv:1106.2700 [ hep - ph ] .

the leading model of gamma - ray bursts ( grbs ) involves a relativistic ( @xmath4 , where @xmath5 is the jet lorentz factor ) fireball jet , where the observed gamma rays are produced by radiation from fermi - accelerated electrons in optically thin shocks ( for reviews see e.g. @xcite ) . in the most popular model for the more common long - duration grbs , the so - called collapsar model , the core of a massive star collapses to a black hole or neutron star , driving a highly relativistic jet which breaks out of the star @xcite . within this scenario , high - energy neutrinos from photomeson interactions of accelerated protons have been studied @xcite . with @xmath6 size neutrino detectors such as icecube @xcite and km3net @xcite being constructed , we are entering an age when we can test various predictions for high - energy neutrinos . waxman and bahcall predicted @xmath7 neutrinos from external reverse shocks @xcite , and similar predictions have been made in refs . @xcite for the forward shock . in addition , internal shocks can occur while the relativistic jet is still in the star . mszros and waxman predicted a @xmath8 neutrino precursor burst from such a set up @xcite . it is expected that a precursor neutrino burst is present even if the jet does not successfully break out of the star , producing a choked jet that is dark in gamma rays and bright in neutrinos . inferring from the observed rate of grbs , highly relativistic ( @xmath9 ) jets occur in perhaps @xmath10 of core - collapse supernovae @xcite . however , the jet signature may be more common , and a significantly higher portion of supernovae may be endowed with slower , mildly relativistic jets ; according to late - time radio observations , perhaps as high as a few @xmath11 @xcite . the recent detection of low - luminosity grb 060218 also suggests from detection rates that low - luminosity grbs , with mildly relativistic ejecta , are more common than conventional high - luminosity grbs by factors as large as @xmath12 @xcite . the occurrence rate is high enough that despite their lower luminosity , their contribution to the neutrino background is comparable to or larger than conventional grbs @xcite . in addition to being more frequent , mildly relativistic jets are more baryon rich , which work positively for neutrino emission . indeed , the detectability of neutrinos from internal shocks of mildly relativistic jets has been studied by razzaque , mszros , and waxman @xcite and extended by ando and beacom @xcite ( see also @xcite ) . in these scenarios , internal shocks occur at @xmath13@xmath14 , smaller than the typical radius of type ib progenitors . accelerated protons interact via proton - photon ( @xmath15 ) and proton - proton ( @xmath16 ) interactions to produce mesons , which consequently produce neutrinos on decay . in this paper we investigate neutrinos arising from protons that are accelerated in the reverse shock , while the relativistic jet is still propagating inside the star . we adopt a type ib progenitor as our model star , but we discuss how our results apply to type ic and ii progenitors also . the reverse shock is strongly motivated because it is a generic prediction of the initial stages of any jet moving through the star , regardless of its lorentz factor , and whether it breaks out of the star or not . therefore , we consider two types of supernovae : one is the perhaps more numerous supernova containing a mildly relativistic ( @xmath17 ) jet , and the other is a supernova containing an ultrarelativistic ( @xmath18 ) jet that can ultimately cause the observed grbs . in both cases we consider choked as well as successful jets , depending on its duration . we show that protons can be accelerated to @xmath0@xmath1 gev in the reverse shock . the low and high - energy protons produce mesons efficiently by @xmath16 and @xmath15 interactions respectively . for energies in between , cooling mechanisms are faster , and result in a characteristic suppression feature in the meson spectrum . mesons that are produced experience significant cooling due to the high proton and photon densities in the shocked jet . however , we still expect a reasonable number of neutrino events . for instance , for a choked mildly relativistic jet with isotropic energy @xmath19 erg at 10 mpc , we expect @xmath20 neutrino events at icecube from the reverse shock . interestingly , assuming the same jet lorentz factor , we find that a lower isotropic energy of @xmath21 erg produces more neutrino events ( @xmath22 events ) . this is due to meson cooling . specifically , we find that neutrinos are favored by high - lorentz factor , low - luminosity , and long - duration jets , as meson cooling is less significant for such jets . the rate of core - collapse supernova within 10 mpc is expected to be @xmath313 yr@xmath23 @xcite , implying particle acceleration in reverse shocks can potentially be tested in the next decades . neutrino emission from the reverse shock is potentially important for short - duration jets that are completely shocked and do not break out of the star , i.e. , choked jets . we argue that while it is possible that internal shocks occur prior to jet shocking ( in particular for low @xmath5 jets ) , the allowed range of radii for internal shocks to occur is constrained from above by the reverse shock , and below by the fact that the stellar material is opaque to high - energy neutrinos . this suggests that neutrino emission from choked jets are of similar magnitude to our estimates . this paper is organized as follows . in sec . ii we discuss the evolution of the jet as it propagates through the star . in sec . iii we treat proton acceleration , and suppression of neutrino emission due to proton and meson cooling . in sec . iv we utilize the results of the previous sections to investigate the detectability of high - energy neutrinos from reverse shocks . finally , we finish with discussions in sec . v and conclusions in sec . we first discuss the evolution of a relativistic @xmath24 jet as it strikes the stellar matter . the aim is to obtain the evolution of the jet head and reverse shock velocities . there are several numerical simulations on the propagation of an initially hot jet in the literature @xcite , as well as analytic treatments of initially cold jets @xcite . we follow and extend the latter . generally , two shocks form when a jet strikes external matter : a reverse shock that decelerates the head of the jet to a lorentz factor @xmath25 , and a forward shock that accelerates the external material to @xmath25 . the jet and stellar material moving at @xmath25 are separated by a contact discontinuity , and are in pressure balance . equations governing the evolution of the shocks are ( for @xmath24 ) @xcite @xmath26 where @xmath27 is the proton mass and @xmath28 is the speed of light . thermodynamic quantities @xmath29 and @xmath30 ( particle number density and internal energy ) are measured in the fluids rest frames , and we have used the following notations : @xmath31 ( jet , unshocked ) , @xmath32 ( jet head , shocked ) , @xmath33 ( stellar , shocked ) , and @xmath34 ( stellar , unshocked ) . lorentz factors are measured in the lab frame , except for @xmath35 which is the lorentz factor of the jet head measured in the frame of the unshocked jet , @xmath36 where @xmath37 is the velocity as a fraction of @xmath28 . the jet particle density is given by ( for a jet with constant opening angle ) @xmath38 where @xmath39 is the isotropic jet luminosity . we obtain @xmath25 by equating the pressure of the shocked jet head @xmath40 and the pressure of the shocked stellar matter @xmath41 @xcite . from eq . ( [ shockequations ] ) , this requires us to solve @xmath42 by substitution of eqs . ( [ gamma3bar ] ) and ( [ n_j ] ) . in what follows we use @xmath43 from the progenitor model e20 from ref . @xcite , an initially @xmath44 rotating star that is reduced to @xmath45 by mass - loss . it has a @xmath46 fe core of radius @xmath47 , a @xmath48 he core that extends to @xmath49 , and a surrounding h envelope . however , the results that we present here apply qualitatively to other models of different masses , with or without h envelopes . as an illustration of this point we include other progenitor models in fig . [ velocity ] . jet head velocity @xmath50 as a function of radius . e15 and e20 are two progenitor models of mass 15@xmath51 and 20@xmath51 respectively , as detailed in the text . l51l53 denotes the isotropic equivalent luminosity of the jet , @xmath52@xmath53 . for the l53 case , two jet lorentz factors are plotted : mildly relativistic ( @xmath54 , solid ) and ultrarelativistic ( @xmath55 , dashed ) . for l52 and l51 , mildly relativistic jets are shown.,width=326 ] due to rotation , the mass along the rotational axis becomes substantially lower than the equatorial , and the rotational axis is an easy escape route for a jet . we inject jets at a radius @xmath56 along the rotational axis , which corresponds to an enclosed mass of @xmath57 . since the explosion and jet mechanisms are still under debate , they are treated parametrically with the following parameters : the isotropic equivalent energy @xmath58 and duration @xmath59 ( which yield the isotropic equivalent luminosity through @xmath60 ) , lorentz factor @xmath5 , and variability @xmath61 . we consider two types of jets , ( i ) mildly relativistic jets with @xmath54 , and ( ii ) ultrarelativistic jets with @xmath55 . in fig . [ velocity ] we show the jet head velocity @xmath62 multiplied by @xmath25 as a function of radius , obtained from solving eq . ( [ v_h ] ) . for @xmath63 , we show our two jet types @xmath54 and @xmath55 , for two progenitor models . the first progenitor model is e20 as described above , and the second is e15 , a similarly rotating star but with initial mass @xmath64 @xcite . we see that the evolution of @xmath62 is independent of @xmath5 , and that the initial jet head velocity is always subrelativistic . the latter is because the external density is very much larger than the jet density , @xmath65 , for conceivable set of jet parameters . as the jet continues to propagate , the stellar density falls faster with radius ( @xmath66 ) than the jet density ( @xmath67 ) , and the jet head effectively accelerates , reaching mildly relativistic values at @xmath68 a few @xmath69 . although the initial value of @xmath62 scales somewhat with @xmath39 , the radius at which the velocity reaches @xmath28 is only weakly dependent on jet parameters . this is because the jet head accelerates to relativistic velocities when the external stellar density suddenly drops , at the edge of the he core . in fact , rewriting eq . ( [ v_h ] ) in the relativistic limit ( @xmath70 ) yields @xmath71 @xcite . note that the kinks in fig . [ velocity ] corresponding to composition changes in the progenitor model are model dependent , e.g. , an abundance of oxygen resides at @xmath72 . however , the general features of @xmath62 that concern us , including its initially @xmath73 value and its reaching @xmath74 at @xmath75 , are model independent . we therefore consider only the e20 progenitor model in the rest of our paper . [ t ] .[table : times]summary of choked jet characteristic times . shown are the core crossing time @xmath76 , and the jet duration time @xmath77 required for reverse shock crossing to occur at @xmath78 cm . these values apply to both mildly relativistic and ultrarelativistic jets . [ cols="^,^,^,^,^",options="header " , ] to estimate @xmath79 for a choked jet , we note that @xmath80 during propagation in the core . the reverse shock thus crosses the entire length of the jet in a time @xmath81 , where @xmath82 is the jet length in the lab frame . we obtain @xmath79 as the distance the jet head has traveled in time @xmath59 . as an example , adopting a jet luminosity @xmath83 , we obtain @xmath84 s ; a jet with duration @xmath85 is therefore choked , and we find that @xmath86 for both the mildly relativistic as well as ultrarelativistic jet cases . for calculation purposes we set @xmath87 . this requires a jet duration of @xmath88@xmath89 , depending on the value of @xmath39 . we summarize duration times in table [ table : times ] . we note that we could have in principle chosen a smaller @xmath79 due to smaller @xmath77 . however , as we discuss in sec . v , the stellar matter becomes opaque to high - energy neutrinos for radii @xmath90 cm , and small radii are uninteresting . we also note here that rotational effects decrease the matter density along the rotational axis , increasing the jet head velocity at small radii and thereby shortening the core crossing time @xmath76 . a decrease in density also decreases neutrino opacity . however , the effect is not expected to be large , because density modifications typically occur at radii smaller than or equal to @xmath91 @xcite . next we discuss jets that successfully break out of the star , with @xmath92 . these jets are progressively shocked as they move through the progenitor , but the reverse shock does not cross the entire jet . therefore , for successful jets , only the front portion of the jet is shocked . we assume @xmath79 just inside of the he core radius . we stress that for successful jets , @xmath79 is _ not _ the point of reverse shock crossing , but the last point of shocking by the reverse shock . a few words about the jet models must be noted . first , we have assumed a conical jet geometry in eq . ( [ n_j ] ) . this is a good approximation for jets propagating in a sufficiently sparse medium , such as the outer regions of stars . from analytic studies , the jet opening angle @xmath93 for @xmath94 and @xmath95 constant for @xmath96 @xcite . we have therefore injected jets at a radius @xmath97 with profile eq . ( [ n_j ] ) . on the other hand , under conditions such as propagation in the dense iron core and/or when strong magnetic fields are associated with the launch of the jet , the high external matter density and/or magnetic hoop stress work to collimate the jet . numerical studies find that jets remain tightly collimated in the iron core @xcite , and further beyond the stronger the magnetic field @xcite . for jets that remain strongly collimated , the reverse shock becomes increasingly less important due to the more penetrating nature of the jet . further , it is less likely for these jets to form choked jets . for illustration , consider an extreme case where a jet of luminosity @xmath98 erg @xmath99 remains tightly cylindrical all the way until break out , with constant cylindrical radius @xmath100 cm @xcite . for such a jet , we find that @xmath101 6 s and @xmath102 s. second , we have assumed a type ib progenitor . a jet with @xmath103 erg s@xmath23 propagating through a type ic progenitor would clear the core in @xmath104 seconds , and jets are more likely to break out . third , we have defined successful jets as those which break out of the star . however , these do not necessarily always produce observable grbs for small @xmath5 and @xmath61 , internal shocks will occur while the jet is optically thick to photons . for example , @xmath105 cm ( we define @xmath106 for a quantity @xmath107 in cgs units ) , while the jet typically becomes optically thin to photons at @xmath108^{1/2 } \sim 10^{14}l_{\rm{iso},52}^{1/2}\gamma_{j,1}^{-1/2}$ ] cm , where @xmath109 is the thomson scattering cross section . the jet is still transparent to high - energy neutrinos . we summarize our various jet definitions in table [ table : jets ] . to conclude , we use the following crossing radii : ( i ) @xmath110 for choked jets and ( ii ) @xmath111 for successful jets . in this section we discuss proton acceleration and neutrino production through decays of pion and kaon mesons . the reverse shock is likely to be collisionless , as we discuss in section iiia . if first - order fermi acceleration is realized , a fraction of the jet protons are accelerated , with a spectrum @xmath112 where the power - law index @xmath113 in the ultrarelativistic shock limit @xcite , and @xmath114 for a nonrelativistic shock . in this paper , we optimistically adopt @xmath114 for all of our figures , but we present numerical results for @xmath115 2.1 , 2.3 , and 2.5 also . the minimum proton energy would be @xmath116 , but otherwise it is not well known ; we assume @xmath117 , but its exact value does not affect our results . the maximum proton acceleration energy is determined by comparing the proton s acceleration time scale to its energy loss time scales . ando and beacom have shown that the photopion process is the most competitive cooling mechanism in the proton energy range of interest @xcite . we therefore focus on photopion production , and refer the reader to appendix a for treatments of other cooling mechanisms such as synchrotron cooling . we will however for completeness display all cooling mechanisms in fig . [ cooling ] . note that the bethe - heitler process turns out to be an important cooling process for @xmath16 interactions , as we will discuss in the next subsection . the energy density in the shocked jet head is @xmath118 . we have specifically used primes ( @xmath119 ) to remind that these quantities are in the jet head comoving frame . we assume that a fraction @xmath120 of the shocked plasma internal energy is converted to relativistic electrons and magnetic fields , in analogy to grbs . this gives the comoving magnetic field strength as @xmath121^{1/2 } ( \bar{\gamma}_h/\gamma_j ) \nonumber \\ & \approx & 1 \times 10^{10 } \ , \left ( \frac{\varepsilon_{b,-1}^{1/2}l_{\rm{iso},52}^{1/2 } } { r_{\times,10.7 } } \right ) \left ( \frac{\bar{\gamma}_{h,1}}{\gamma_{j,1 } } \right ) \ , \mathrm{g},\end{aligned}\ ] ] where we have assumed @xmath80 since @xmath122 for the jet scenarios studied here . one can obtain from eq . ( [ gamma3bar ] ) that for example , @xmath123 yields @xmath124 . the relativistic electrons immediately lose energy by synchrotron radiation , and the photons thermalize due to the large opacity @xmath125 , where @xmath126 and @xmath127 is the proper thickness of the shocked jet plasma . here we adopt @xmath128 , where @xmath129 @xcite . the black - body temperature is then @xmath130^{1/4 } ( \bar{\gamma}_h/\gamma_j)^{1/2 } \nonumber \\ & \approx & 13 \ , \left ( \frac{\varepsilon_{e,-1}^{1/4}l_{\rm{iso},52}^{1/4 } } { r_{\times,10.7}^{1/2 } } \right ) \left ( \frac{\bar{\gamma}_{h,1}}{\gamma_{j,1 } } \right)^{1/2 } \,\mathrm{kev},\end{aligned}\ ] ] and the average photon energy is @xmath131 . the average density is @xmath132^{3/4 } ( \bar{\gamma}_h/\gamma_j)^{3/2 } \nonumber \\ & \approx & 7 \times 10^{25 } \ , \left ( \frac{\varepsilon_{e,-1}^{3/4}l_{\rm{iso},52}^{3/4 } } { r_{\times,10.7}^{3/2 } } \right ) \left(\frac{\bar{\gamma}_{h,1}}{\gamma_{j,1 } } \right)^{3/2 } \,\mathrm{cm^{-3}}.\end{aligned}\ ] ] the total proton cooling time is determined by the inverse sum @xmath133 , where @xmath134 runs through the various cooling mechanisms . as discussed in the opening paragraph , we focus on cooling by photopion production . protons interact with the dense photon field eq . ( [ photons2 ] ) , and for sufficient proton - photon centre of inertia energies mesons are produced . we determine the cooling time scale by @xmath135 . we assume the conventional inelasticity @xmath136/2 $ ] where @xmath33 is the invariant mass of the system . we fit @xmath137 to @xcite , and use a blackbody photon spectrum of temperature eq . ( [ photons ] ) . the time scale is plotted in fig . [ cooling ] . the acceleration time scale in the jet head frame is given by @xmath138 , where @xmath139 is a constant if we assume the diffusion coefficient is proportional to the bohm diffusion coefficient , and @xmath140 is the larmor radius . @xmath141 is a fairly conservative value , and @xmath142 can be achieved for mildly relativistic shocks @xcite . we use @xmath143 which gives agreement with previous studies @xcite . using the magnetic field in eq . ( [ bfield ] ) , we obtain @xmath144 where @xmath145 is measured in gev . inverse of proton cooling and acceleration time scales , in the shocked jet head frame , as functions of the proton energy . the case of reverse shock crossing at @xmath146 is plotted for supernova jet model b , with other parameters @xmath54 , @xmath147 , and @xmath148 . cooling mechanisms shown are : synchrotron ( blue , thin dashed ) , inverse - compton ( green , dot - dashed ) , bethe - heitler ( green , dotted ) , proton - proton ( black , long dashed ) , and photopion ( black , thick dashed ) . also labeled are threshold energies @xmath149 and @xmath150 ( see sec . photopion limits the proton energy to @xmath151.,width=326 ] in fig . [ cooling ] , we plot the proton acceleration and cooling time scales as functions of the proton energy , both in the jet head comoving frame . acceleration is stronger than cooling at low proton energies , but photopion cooling becomes stronger at higher energies . it is evident that photopion cooling is the strongest cooling mechanism . equating @xmath152 , we obtain the maximum proton energy as @xmath153 . the parameter dependency is @xmath154 in the flat region of @xmath155 . we note that the reverse shock is likely to be collisionless , because the proton plasma frequency @xmath156 is larger than the typical radiation and particle collision frequencies . for example , the frequency of proton - photon collision is @xmath157 , while the frequency of proton - proton collision is @xmath158 . depending on its energy , protons produce mesons via @xmath16 and/or @xmath15 interactions . we consider the pion and kaon mesons . the multiplicity in each @xmath16 and @xmath15 interaction is taken to be @xmath159 for pions and @xmath160 for kaons ; this matches the required ratio in the relevant energies @xcite . generally , kaons produce higher energy neutrinos @xcite , but we additionally find that they can dominate those from pions because of less efficient kaon cooling . for low energy protons , @xmath16 interactions dominate . meson production by @xmath16 interaction is efficient since the opacity is very high , @xmath161 , where @xmath162 and @xmath163 @xcite . proton cooling by pair production ( bethe - heitler , hereafter bh , see appendix for treatment ) overtakes @xmath16 interactions at higher energies ( see fig . [ cooling ] ) . we define this transition as @xmath164 . at even higher energies @xmath15 dominates over bh cooling , and we define this transition energy as @xmath165 . the @xmath15 interaction is efficient at producing mesons since the opacity is @xmath166 , where we use @xmath167 @xcite . for model b in table i , @xmath168 gev and @xmath169 gev . we assume mesons are produced with 20% of the parent proton energy . in the absence of competing cooling mechanisms , the meson spectrum follows the initial proton spectrum . on the other hand , when other cooling processes are faster than meson - yielding @xmath16 and @xmath15 interactions , the meson spectrum is suppressed by an amount equal to the ratio of time scales . we find that the bethe - heitler process is a competitive cooling mechanism for @xmath16 interactions , and the resulting meson spectrum will be suppressed by a factor given as @xmath170 we here approximate @xmath171 as a power - law @xmath172 for @xmath173 , where @xmath174 is a constant . we find that the value of @xmath174 is fairly model independent , with values @xmath175 ( 1.2 ) for @xmath5 = 10 ( 100 ) for model b. in this section we discuss meson cooling and the neutrino spectrum . mesons cool similarly to protons , by radiative and hadronic ( collisions with protons ) processes . the meson cooling time scales are analogous to those for the proton and can be summarized as @xmath176 $ ] and @xmath177 , where @xmath178 and @xmath179 are the meson mass and energy , and @xmath180 and @xmath181 are the photon and magnetic field energy densities ( we assume inverse - compton process is in the thompson regime ; synchrotron dominates at high energies ) . here , @xmath182 @xcite is the cross section for meson - proton collisions and @xmath183 @xcite is the energy lost by the meson per collision . the total cooling time is @xmath184 , and the time scales are @xmath185 and @xmath186 for pions . for kaons , the radiative cooling time scale is longer , because @xmath187 . hadronic cooling dominates at low proton energies , and radiative cooling becomes important at higher energies . charged pions and kaons decay into neutrinos through @xmath188 , with the muon neutrino taking 1/4 of the meson energy . we neglect secondary neutrinos from muon decays since muons immediately lose energy by radiative cooling . we define the break energy @xmath189 by equating @xmath190 , where @xmath191 and @xmath192 are the meson lorentz factor and proper lifetime . below @xmath189 , neutrinos have a typical energy @xmath193 and a flat power per decade neutrino spectrum . above @xmath189 , the spectrum is suppressed by a factor @xmath194 . next , we define @xmath195 when radiative cooling begins to dominate , by equating @xmath196 . due to relativistic effects , the neutrino energy in the observer frame is related to the parent meson energy in the jet head frame as @xmath197 and @xmath198 . we define the suppression function @xmath199 as @xmath200 for some supernova jet parameters , it is possible that meson goes from decay dominated straight to radiation cooling dominated . for these cases we define the break energy @xmath201 by equating @xmath202 , and the suppression function @xmath199 as @xmath203 the break energies for neutrinos from pion decay , in a choked mildly relativistic jet , are @xmath204 the parameter dependencies are @xmath205 and @xmath206 for the first and second break energies , respectively . on the other hand , break energies for neutrinos from kaon decay are higher due to the higher kaon mass and faster decay time ; @xmath207 the parameter dependencies are the same as for pions . the break energies are smaller than those of internal shocks @xcite with the same jet parameters , by as much as @xmath32 orders of magnitude . this is a combination of the higher proton and photon densities in the shocked jet causing more effective cooling , and the fact that the neutrino energies in the shocked jet head receive no lorentz boosting . next , the break energies for a choked ultrarelativistic jet are @xmath208 which can also be obtained from scaling eqs . ( [ break12])([break34 ] ) by their dependencies on @xmath5 importantly for neutrino emission , suppression due to radiative cooling is independent of @xmath5 , since the product of @xmath209 is unchanged . also , note that @xmath210 , indicating that radiative cooling dominates the cooling process for pions . in this case , @xmath211 we show in fig . [ suppression1 ] the resulting suppression factors for neutrinos from kaons , for our mildly relativistic and ultrarelativistic jets . suppression is less intense for relativistic jets . we also shade in for reference the energy range where the meson spectrum is suppressed by bethe - heitler cooling , from 5140 gev . note that for pions , radiative cooling dominates for energies of interest . suppression factor as a function of neutrino energy , for neutrinos from kaon decays . shown are the choked mildly relativistic jet ( solid lines ) and choked ultrarelativistic jet ( dashed lines ) . the kaon experiences less cooling for an ultrarelativistic jet . note that pions are always in the radiation cooling regime for neutrino energies shown . we also shade in for reference the energies where bethe - heitler cooling is competitive and suppresses the meson spectrum.,width=326 ] cumulative muon event number from choked jets with model b parameters . thin - black ( thick - green ) lines denote pion ( kaon ) contributions . the majority originates from kaons . solid lines denote a mildly relativistic jet and a supernova distance of @xmath212 . dashed lines denote an ultrarelativistic jet and two supernova distances - @xmath212 for illustrative purposes and @xmath213 for event calculations , as labeled . due to weaker kaon cooling , an ultrarelativistic jet yields more neutrinos than a mildly relativistic one . the spectrum becomes almost flat around @xmath3 100 gev , due to bethe - heitler suppression of the meson spectrum . the total number of events expected above the detector threshold energy is about 6 for the mildly relativistic case . for the ultrarelativistic case with distance @xmath213 we expect 0.02 events . also shown is the atmospheric neutrino background over 1 day in a @xmath214 circle.,width=326 ] in this section we estimate the emission and detectability of high - energy neutrinos from reverse shock accelerated protons . we consider a supernovae at a distance of @xmath215 , and use the code anis ( all neutrino interaction generator ) @xcite to calculate the neutrino induced muon spectrum at a neutrino detector with an effective area of @xmath216 , i.e. an icecube class detector utilizing upgoing muons . first , we estimate the fluence of neutrinos as @xcite @xmath217 where @xmath218 is the meson multiplicity ( @xmath159 for pions and @xmath160 for kaons ) , @xmath219 is the branching ratio of meson decay into neutrinos ( @xmath159 for pions and @xmath220 for kaons ) , and @xmath221 is the fraction of proton energy carried by neutrinos in the absence of energy loses ; @xmath222 for pions and @xmath223 for kaons , since neutral and charged mesons are produced with roughly equal probability , and muon neutrinos carry roughly @xmath223 ( @xmath224 ) of the pion ( kaon ) energy in meson decay . the functions @xmath225 and @xmath226 are the suppression factors due to proton and meson coolings , respectively , and the @xmath227 factor normalizes the proton spectrum to the jet energy ( for @xmath228 ) . in fig . [ muon ] we show the cumulative spectrum of neutrino induced muons , from a supernova with a choked jet . we show three cases - a supernova distance of @xmath229 possessing a mildly relativistic @xmath54 jet ( solid lines ) , a supernova distance of @xmath229 possessing an ultrarelativistic @xmath55 jet ( dashed lines ) , and a similarly ultrarelativistic jet but supernova distance @xmath230 ( dashed lines labeled 400 mpc ) . we use model b with parameters @xmath147 and @xmath231 . we take into account the muon range , which effectively increases the detector volume . we evaluate the muon energy @xmath232 when it enters the detector if it is produced outside , or at the production point if it is inside . considering a detection threshold muon energy of @xmath233 , we expect about 6 muon events from a 10 mpc supernova possessing a choked mildly relativistic jet . similarly , we expect about 28 from an ultrarelativistic jet . we have plotted the ultrarelativistic jet case at an optimistic distance of 10 mpc to illustrate the effect of meson cooling . although the pion contributions are similar between the two jets , the kaon contribution is larger in the ultrarelativistic case due to the weaker suppression . in the remainder we assume the distance 400 mpc for ultrarelativistic jet calculations . for model b , this yields 0.02 events . note that since the muon spectrum is steep , a lower detector energy threshold will yield much higher events . additionally , one may observe the suppression due to bethe - heitler cooling , which is expected to manifest below @xmath234 gev . for a steeper initial proton spectrum , motivated from studies of ultrarelativistic shocks , the expected event is 4 ( 2 , 1 ) for @xmath235 ( 2.3 , 2.5 ) . in all jets , the majority of neutrinos come from kaons , because kaons are more massive than pions and suffer less cooling ( the radiative cooling time scale is @xmath236 ) . also , the kaon decay time is slightly shorter than the pion . the events discussed here arrive in a @xmath237 34 second time bin and @xmath3 3@xmath238 angular bin , allowing very strong rejection of atmospheric neutrino backgrounds . we show in the figure the atmospheric neutrino background within @xmath159 day and a @xmath214 bin for comparison . the expected neutrino signal from a choked mildly relativistic jet easily exceeds the atmospheric neutrino background . same as fig . [ muon ] but for model c parameters . approximately 17 ( 0.04 ) events are expected above the detector threshold for a single mildly relativistic ( ultrarelativistic ) jet and supernova distance @xmath212 ( @xmath213).,width=326 ] same as fig . [ muon ] but for jets with successful jets . the supernova and jet parameters are @xmath239 , @xmath240 , and @xmath241 . about 36 ( 0.1 ) events are expected above the detector threshold muon energy from a mildly relativistic ( ultrarelativistic ) jet supernova at a distance @xmath212 ( @xmath213 ) away.,width=326 ] next we discuss a scan of supernova and jet parameters . first , as shown above , a larger @xmath5 reduces kaon suppression and increases neutrino emission . next , we consider jet model a of table [ table : times ] , i.e. , with larger @xmath39 but smaller @xmath59 , and an overall larger @xmath58 . since the neutrino fluence scales with @xmath58 , one naively expects more neutrinos . however , the expected muon event rate is in fact smaller , @xmath242 ( @xmath243 ) for a mildly relativistic ( ultrarelativistic ) jet . this is because of stronger meson cooling : the proton and photon densities in the shocked jet scale as @xmath244 and @xmath245 , so a larger @xmath39 results in stronger meson cooling . we conclude that neutrino emission from choked jets are more favorable from low - luminosity and long - duration jets , even if @xmath58 is smaller ( assuming the same jet lorentz factor ) . indeed , model c predicts 17 ( 0.04 ) events at icecube . we show the cumulative muon event spectrum for jet model c in fig . [ muonc ] . in fig . [ muon2 ] we show the same as fig . [ muon ] but for successful jets . we show a mildly relativistic @xmath54 jet at @xmath212 ( solid lines ) , and an ultrarelativistic @xmath55 jet at @xmath213 ( dashed lines ) . the other parameters used are @xmath246 , @xmath240 , @xmath241 , @xmath147 , and @xmath247 . with these parameters , @xmath84 s , and the cold jet does indeed break out of the star . we introduce an additional factor of 1/2 to the neutrino fluence for successful jets , because only about @xmath248 of the jet length is shocked by the reverse shock . note that our chosen parameters have the equivalent @xmath58 to jet model a. however , we expect @xmath249 times more neutrino events than jet model a , because @xmath39 is a factor 10 smaller and @xmath79 is a factor 2 larger . indeed , evaluating the muon event rate yields about 36 ( 0.1 ) events per mildly relativistic ( ultrarelativistic ) jet . as in the case of choked jets , most of the contribution comes from kaons . these events will show temporal and angular clustering , and in addition , these will appear as precursor signals to grb photons , allowing very strong rejection of background . our result is that neutrinos from the reverse shock are strongly suppressed due to proton and meson cooling . however , we still expect a reasonable number of events at neutrino detectors . one reason for this is that proton cooling affects only energies near and below current detector threshold energies . this produces a suppression feature in the spectrum but does not decrease event counts at icecube . for our models a , b , and c in table [ table : times ] , we expect 2 , 6 , and 17 muon events respectively at a icecube class detector . here , we have calculated for a supernova @xmath212 away , possessing a @xmath54 mildly relativistic jet , and a @xmath228 initial proton spectrum . if we assume a steeper index , @xmath115 2.1 ( 2.3 , 2.5 ) , we expect 4 ( 2 , 1 ) events for jet model b. in all cases , the neutrino flux is dominated by decay of kaons , since cooling is less efficient for kaons than pions . since neutrino emission from the reverse shock is suppressed by meson cooling , we find that neutrino emission is favored by parameter sets for which meson cooling is weaker high - lorentz factor , low - luminosity , and long - duration jets . this fact is readily observable from the neutrino event predictions of models a , b , and c. as further illustration , if we assume a @xmath55 ultrarelativistic jet we expect 11 , 28 , and 55 events from jet models a , b , and c respectively . note that supernova and jet parameters are expected to show considerable scatter . hence , given neutrino emission is cooling suppressed , the event count need not necessarily correlate with the total burst energy . neutrino emission from the reverse shock is particularly important for choked jets , which are completely shock decelerated by the reverse shock . is it possible for choked jets to produce more neutrino emission than our above estimates ? if the duration is long enough such that @xmath250 , it is possible that internal shocks accelerate protons in the unshocked part of the jet , while the jet head is being shocked . the internal shock accelerated protons do not travel far , producing mesons efficiently via @xmath16 interactions with the large target jet proton density @xcite . for high - lorentz factor jets , the protons can travel and interact with termination shock photons @xcite . in both cases meson cooling is less severe than in the reverse shock , and we can expect more neutrino emission . however , internal shocks must occur at large enough radii that the stellar material is not opaque to high - energy neutrinos . the neutrino opacity is @xmath251 , where @xmath252 is the column number density of the star and @xmath253 is the sum of neutrino charged - current and neutral - current cross sections , which is approximately @xmath254 in the energy range in question . for a @xmath255 neutrino emitted at @xmath256 , @xmath257 and @xmath258 @xcite , and we find that @xmath259 . for a @xmath260 neutrino , @xmath261 . thus , only neutrinos with energies @xmath262 can leave the star from @xmath263 . clearly , these results depend on the stellar density profile assumed , and rotational effects have been neglected in our simple estimate . however , it seems clear that the stellar matter is opaque to high - energy neutrinos for small radii , constraining the range of allowed radii for internal shocks . the case for the reverse shock in successful jets , which break out of the star , seem better . this however is not due to any intrinsic property of successful jets . it is merely because for a given total jet energy , successful jets have longer durations and lower luminosities compared to choked jets . this results in less meson cooling ; for example , we expect 2 events from a choked jet ( of energy @xmath264 ) , while we expect 36 from a successful jet of the same total energy . however , it should be noted that we expect internal shocks to be important for successful jets . considering our successful jet with @xmath246 , @xmath240 , and @xmath265 ( so that internal shocks occur at @xmath266 ) , we find that the total muon event number from the internal shock is @xmath3100 times larger than our estimates from the reverse shock @xcite . this is due to less meson cooling in the cold jet , and we compare in fig . [ suppression2 ] the suppression factor for mesons in the cold jet and hot jet . it is clear that suppression in the cold jet is much weaker . suppression factor for neutrinos from the reverse shock ( solid lines ) and the internal shock ( dot - dashed lines ) . we adopt our successful mildly relativisitc jet ( @xmath246 , @xmath240 , @xmath265 ) , and show the suppression factors for pions ( thin - black lines ) and kaons ( thick - green lines ) . the vertical dotted lines indicate break energies.,width=326 ] despite being overwhelmed in terms of numbers , reverse shock neutrinos will arrive prior to neutrinos from internal shocks . mszros and waxman @xcite studied precursor signals arising from internal shock accelerated protons interacting with photons at @xmath267 , while razzaque , mszros , and waxman @xcite extended this to include @xmath268 . neutrinos from reverse shocks will precede even these precursor signals , provided the radius is large enough that the stellar material is thin to neutrinos . the temporal profile of neutrinos therefore offers the possibility to distinguish locations of particle acceleration for exceptionally close supernovae . the expected time difference between components is of order a few to tens seconds , which is within the time resolution of neutrino detectors . we classified jets by whether they break out of the star or not , depending on their durations . choked jets are bright in neutrinos and dark in gamma rays , while successful jets are bright in neutrinos but may or may not be bright in gamma rays , depending on whether internal shocks are optically thick or not . a successful yet gamma dark jet may be typical for mildly relativistic jets , where the smaller @xmath5 implies internal shocks typically occur at smaller radii @xcite . these could be distinguished from choked jets by the presence of an afterglow like radio signature , for example . in the current work we have assumed a type ib progenitor . however , if the progenitor is a type ic , the situation is slightly different . having no he core , the core crossing time is shorter , @xmath269 s , for @xmath103 erg s@xmath23 , and choked jets will become rarer . this implies the relative importance of internal shocks over reverse shocks . for successful ultrarelativistic jets in type ic progenitors , the neutrino precursor signal can be as strong as those discussed previously @xcite . our results will however apply well to type ii progenitors . the core - collapse supernova rate within 10 mpc is @xmath313 yr@xmath23 @xcite , and therefore , one could test the possibility of particle acceleration in the reverse shocks in the next decade . in particular , correlating with the optical signal of the supernova explosions ( not necessarily the jet - like signature ) @xcite or mev neutrino detections @xcite would greatly help enhance the significance of the signal . examples of recent close supernovae include sn 2007gr in ngc 1058 ( distance @xmath310 mpc ) @xcite , and a further list can be found in ref . the actual rate could in fact be higher , as a systematic survey of all nearby galaxies has not been performed to date . we showed that neutrino emission from reverse shock accelerated protons are strongly suppressed due to meson cooling in the hot jet and hot stellar material surrounding the reverse shock . despite this suppression , we still expect a fairly promising number of events at icecube , ranging from @xmath242 to @xmath270 for a close supernova of @xmath212 hosting a choked mildly relativistic jet of total isotropic energy @xmath21@xmath271 . we find that for a supernova of given energy @xmath58 , neutrino emission is favored by high - lorentz factor , low - luminosity , and long - duration jets , for which meson cooling is less severe . our result sets the scale for neutrino emission from choked jets . for choked jets to provide more neutrinos , internal shocks are required to occur in the unshocked region of the jet . however , this is geometrically constrained from above by the reverse shock , and from below by the fact that the stellar matter is opaque to high - energy neutrinos for small radii . regarding neutrino precursor signals which precede grb photons , we conclude that the reverse shock neutrinos likely plays a small role . however , they will precede those neutrinos from internal shocks , allowing the temporal profile of precursor neutrinos from a close supernova to provide an identification of proton acceleration at the reverse shock . we are grateful to john beacom for discussions and comments , and peter mszros for reading the script . we also thank ehud nakar and yudai suwa for early discussions , and hylke koers for email discussions after initial submission . sa was supported by the sherman fairchild foundation . we discuss cooling mechanisms in the shocked jet head , which act on the proton ( section iiia , shown in fig . [ cooling ] ) and mesons ( section iiic ) . we explicitly denote by primes ( @xmath119 ) the jet head comoving frame . all numerical values are derived for the proton . first , the same magnetic field responsible for electron synchrotron radiation also causes energy losses in protons and mesons , given by @xmath272 where @xmath273 is the electron mass , and @xmath274 and @xmath275 are the particle mass and energy . the numerical value is derived for the proton . due to the strong dependency on the mass , @xmath236 , synchrotron cooling is most effective for the pion meson and least effective for the proton . second , protons and mesons lose energy by inverse compton ( ic ) scattering . the most natural target photons are due to radiation cooling of ultrarelativistic electrons . although a full treatment using a black - body spectrum is simple , a simplistic approximation using eq . ( [ photons ] ) and eq . ( [ photons2 ] ) is sufficient to demonstrate that ic cooling is subdominant . below a transition energy @xmath276 ( for the proton ) , inverse compton scattering proceeds in the thomson regime , and the cooling time scale is given @xmath277 again , note the strong dependency on the particle mass . above the transition energy , inverse compton scattering proceeds in the klein - nishina limit , where the cooling time scale becomes @xmath278 third , we consider cooling by the bethe - heitler ( bh ) interaction ( @xmath279 ) , which is important for protons . our treatment is the same as that described in ref . the bh cross section rises logarithmically with energy , @xmath280\}$ ] @xcite . the @xmath281 produced are at rest in the centre of mass frame of the proton and photon , and so the energy lost by the proton in each interaction is given by @xmath282 , where @xmath283 is the lorentz factor of the centre of inertia as seen in the comoving frame . the energy loss rate is @xmath284 , and thus the cooling time is @xmath285 . lastly we discuss cooling due to collisions with protons . assuming @xmath286 of the proton energy is lost in each collisions with a proton , the cooling time scale is @xmath287 for the proton , where @xmath163 @xcite . analogous equations hold for mesons , e.g. eq . ( [ mesonhadronic ] ) for the pion . s. ando , j. f. beacom and h. yuksel , phys . * 95 * , 171101 ( 2005 ) [ arxiv : astro - ph/0503321 ] . s. ando , jcap * 0406 * , 007 ( 2004 ) [ arxiv : astro - ph/0405411 ] . e. nakar , phys . rept . * 442 * , 166 ( 2007 ) [ arxiv : astro - ph/0701748 ] . m. a. aloy , e. mueller , j. m. ibanez , j. m. marti and a. macfadyen , astrophys . j. * 531 * , l119 ( 2000 ) [ arxiv : astro - ph/9911098 ] . a. burrows , l. dessart , e. livne , c. d. ott and j. murphy , arxiv : astro - ph/0702539 . t. takiwaki , k. kotake , s. nagataki and k. sato , astrophys . j. * 616 * , 1086 ( 2004 ) [ arxiv : astro - ph/0408388 ] . a. achterberg , y. a. gallant , j. g. kirk and a. w. guthmann , mon . not . * 328 * , 393 ( 2001 ) [ arxiv : astro - ph/0107530 ] .

superconducting qubits are solid - state artificial atoms , comprising lithographically defined josephson tunnel junctions and superconducting interconnects . when cooled to milli - kelvin temperatures , these qubits exhibit quantized states of charge , flux , or junction phase depending on the circuit design parameters @xcite . associated with these quantized states is a spectrum of energy levels , tunable via an external control parameter , e.g. , an applied electric or magnetic field . although generally only the lowest two energy eigenstates are utilized for quantum information science applications , the energy spectrum indeed extends to higher - energy levels corresponding to higher - excited states of the circuit . the separation between pairs of energy levels typically falls in the radio frequency and microwave regimes , and resonantly driving the artificial atom with a harmonic field can couple and induce quantum - state transitions . due to their relatively large size , superconducting artificial atoms can be strongly coupled to their external control fields . it is this feature , along with their quantum coherence , that we utilize in the present article . a large - amplitude harmonic control field can drive an artificial atom throughout its energy - level spectrum . when driven through an avoided level crossing , a landau - zener - stckelberg ( lzs ) transition occurs . this is a coherent process akin to a beamsplitter for photons , taking an input state of the atom and outputting a superposition of states . repeated passages through an avoided crossing act as an atom interferometer , causing the atomic superposition state to interfere quantum mechanically with itself . since the weighting of the superposition state depends sensitively on the size of the avoided crossing and the velocity ( change in relative energy between levels per unit time ) with which it is traversed , the quantum interference reflects the energy spectrum of the artificial atom . in turn , the quantum interference can be leveraged to facilitate non - adiabatic quantum control . we begin this article with an introduction to the superconducting persistent - current qubit @xcite and an overview of landau - zener - stckelberg ( lzs ) transitions . we then present three experimental works that utilize lzs transitions in a strongly - driven persistent - current qubit . the first is mach - zehnder - type interferometry between repeated lzs transitions @xcite , for which we observed quantum interference fringes in n - photon transition rates @xcite , with n=1 50 . the second is microwave - induced cooling @xcite , by which we achieved effective qubit temperatures less than 3 mk , a factor 10 - 100 times lower than the environmental temperature . the third is amplitude spectroscopy @xcite , a spectroscopy technique that monitors the system response to amplitude rather than frequency . amplitude spectroscopy allowed us to probe the energy spectra of our artificial atom from 0.01 - 120 ghz , while driving it at a fixed frequency near 0.2 ghz . finally , we conclude by considering the application of lzs transitions to quantum information science and technology . superconducting artificial atoms exhibit a high degree of quantum coherence , and there have been numerous proposals and demonstrations of quantum phenomena in these systems , many derived from the fields of atomic physics and quantum optics . a few examples include : coherent superpositions of macroscopic states @xcite , rabi oscillations @xcite , landau - zener transitions @xcite , stckelberg oscillations @xcite , microwave cooling @xcite , electromagnetically induced transparency @xcite , geometrical phase @xcite , and cavity quantum electrodynamics @xcite . significant progress has also been made toward their application to quantum information science @xcite , including state initialization @xcite , tunable @xcite and long - distance @xcite coupling , quantum control @xcite , quantum state @xcite and process @xcite tomography , and measurement @xcite . for a recent review of superconducting qubits , see ref . . here , we use a superconducting persistent - current qubit to implement an artificial atom @xcite . the persistent - current qubit is a superconducting loop interrupted by three josephson junctions ( fig . [ fig:1]a ) . when biased with a static magnetic flux @xmath0 , where @xmath1 is the superconducting flux quantum , the system assumes a double - well potential profile ( fig . [ fig:1]b ) . the diabatic ground state of the left ( right ) well corresponds to a persistent current @xmath2 with clockwise ( counterclockwise ) circulation . these two diabatic energy levels ( fig . [ fig:2]a ) have a separation @xmath3 linear in the flux detuning @xmath4 . higher - excited states of the double - well potential ( see fig . [ fig:5]c ) will be considered in sections [ sec : cooling ] and [ sec : as ] . the two - level hamiltonian for the lowest two diabatic states is shown in eq . [ eq : hamiltonian ] . at detuning @xmath5 , the double - well potential is symmetric and the diabatic - state energies are degenerate . at this `` degeneracy point , '' resonant tunneling between the diabatic states opens an avoided level crossing of energy @xmath6 . here , the qubit states are @xmath7 eigenstates , corresponding to symmetric and anti - symmetric combinations of diabatic circulating - current states . detuning the flux away from this point tilts the double well , allowing us to tune the eigenstates and eigenenergies of the artificial atom . far from the degeneracy point the qubit states are approximately @xmath8 eigenstates , the diabatic states with well - defined circulating current . the qubit is read out using a hysteretic dc squid ( superconducting quantum interference device ) , a sensitive magnetometer that can distinguish the flux generated by circulating current states . in addition to the static flux biases , the artificial atom is controlled and read out using the pulses illustrated in fig . [ fig:1]c . as we describe below , the qubit is first prepared in its ground state using a harmonic cooling pulse with amplitude @xmath9 and frequency @xmath10 . quantum - state transitions are then driven using a harmonic rf pulse with amplitude @xmath11 and frequency @xmath12 . these fields are mutually coupled to the qubit through a small antenna . this is followed by a squid readout current pulse using the `` sample and hold '' technique @xcite . if the sample current exceeds the squid switching current , a voltage pulse will appear at the output during the hold phase . threshold detection looks for the presence or absence of a squid voltage , and this constitutes a digital measurement of the qubit state . alternatively , although not used in these experiments , we have incorporated the squid into a resonant circuit and realized qubit readout via the shift in resonance frequency and phase for both the linear and non - linear resonance regimes @xcite . the `` qubit step '' is shown in fig . [ fig:1]d as a function of the squid sample current and the flux detuning . the diabatic states @xmath13 and @xmath14 correspond to different levels of sample current ( dashed lines ) located symmetrically about the degeneracy point @xmath15=0 . this plot constitutes a cumulative switching current distribution of the squid for each @xmath15 value . additionally , a 3-@xmath16s rf pulse at 1.2 ghz is applied to the qubit , and resonant transitions can be observed as fingers extending down ( up ) from state @xmath13 ( @xmath14 ) when @xmath171.2 ghz becomes resonant with the energy - level separation . a best - estimator ( dashed - dotted line ) can be determined to provide the best statistical discrimination between states @xmath13 and @xmath14 . the resulting qubit step with its saturated n - photon resonances along the best estimator line is shown in fig . [ fig:1]e . the device used in this work was fabricated at lincoln laboratory using a fully - planarized niobium trilayer process and optical lithography . it has a critical current density @xmath18 , and the characteristic josephson and charging energies are @xmath19 @xmath20 respectively . the ratio of the qubit josephson junction ( jj ) areas is @xmath21 , and @xmath22 . although dependent on the flux detuning , the approximate values for coherence times are : interwell relaxation time @xmath23 , intrawell relaxation time @xmath24 , homogeneous transverse decay time @xmath25 , inhomogeneous transverse decay time @xmath26 . the experiments were performed in a dilution refrigerator at a base temperature of 20 mk . the device was magnetically shielded , and all electrical lines were carefully filtered and attenuated to reduce noise ( see ref . for details ) . landau - zener - stckelberg ( lzs ) transitions occur when a quantum system is driven through an energy - level avoided crossing @xcite . the resulting quantum dynamics of the lzs mechanism in driven systems @xcite have been developed @xcite within a two - level coherent scattering formalism @xcite with potential application to quantum information science and technology @xcite . we model a lzs transition at a single crossing using the two - level hamiltonian @xmath27 in which @xmath28 is the energy difference between diabatic - state energy levels ( dashed lines in fig . [ fig:2]a ) , and @xmath6 is the size of the avoided crossing and corresponds to the coupling strength between diabatic states @xmath29 and @xmath30 . the system is prepared in state @xmath31 at a static value @xmath32 far from the avoided crossing ( blue dot , fig . [ fig:2]a ) and driven longitudinally from that point by making @xmath33 time dependent . we distinguish here between `` weak driving , '' in which the resulting excursion is not large enough to reach the avoided crossing , and `` strong driving '' for which the avoided crossing is traversed . under strong - driving conditions , the asymptotic probability @xmath34 of a transition between diabatic states , @xmath35 is governed by the relative - energy sweep rate @xmath36 @xmath37 evaluated at the time @xmath38 at which the system is swept through the avoided crossing . in the original lzs formulation , the system was driven with a fixed sweep rate , whereas we will later consider a harmonically driven system . we also note that we have elected to monitor the probability @xmath34 of a transition between the diabatic states @xmath39 and @xmath31 , because our readout detector is sensitive to changes in diabatic state . this can be written with no loss of generality in terms of the probability @xmath40 of a transition between the system eigenstates . there are two strong - driving limits characterized by the relative sizes of the sweep velocity @xmath41 and the avoided crossing @xmath6 . in the adiabatic limit ( fig . [ fig:2]a ) , the sweep velocity is small ( @xmath42 ) and the probability of a transition from state @xmath31 to state @xmath39 approaches unity , @xmath43 . in this case , the system dynamics are slow enough that the system adiabatically follows the ground eigenstate through the avoided crossing . in the nonadiabatic limit ( fig . [ fig:2]b ) , the sweep velocity is large ( @xmath44 ) and the probability of a transition approaches zero @xmath45 . in this case , the dynamics are too fast for the system to follow ; the system remains in diabatic state @xmath31 and , thereby , effectively jumps the energy gap at the avoided crossing . more generally , a superposition state @xmath46 results from an excursion through the avoided crossing , as illustrated in fig . [ fig:2]c . following an idea discussed by shytov _ et al . _ @xcite , the lzs transition acts as a beamsplitter for the atomic state . the amplitudes @xmath47 and @xmath48 are determined by a unitary transformation @xmath49 , effectively a @xmath50 `` beamsplitter matrix '' comprising complex reflection @xmath51 and transmission @xmath52 coefficients related to the adiabaticity parameter @xcite @xmath53 present in eq . [ eq : p_lzs ] such that @xmath54 , @xmath55 , and @xmath56 . note that we have defined @xmath57 and @xmath58 from the perspective of a beamsplitter , which `` transmits '' ( `` reflects '' ) an input state to the same ( opposite ) diabatic state , respectively . after the transition , a relative phase @xmath59 accrues due to the energy difference between the states @xmath39 and @xmath31 . if the drive @xmath33 then returns the system through the avoided crossing a second time , the atomic state collides and quantum mechanically interferes with itself during the second lzs transition ( fig . [ fig:2]d ) . the cumulative result is an atom - state interferometer whose output state depends on the lzs transition amplitudes and the interference phase . an analogy can be made to an optical mach - zehnder interferometer : the atomic states play the role of the photon modes , the lzs transitions play the role of the photon beamsplitters , and the energy - level splitting , which determines the interference phase , plays the role of optical path length difference . the quantum interference is robust provided the evolution time of the state through the interferometer is short compared with the atom s coherence times . in addition to superconducting artificial atoms @xcite , this concept is generally applicable to other solid - state artificial atoms @xcite and generalized spin systems @xcite ( e.g. , molecular magnets @xcite , natural atoms @xcite , and molecules @xcite ) that exhibit avoided level crossings , and it is extensible to multiple energy levels as we demonstrate in section [ fig:5 ] . the structure of the n - photon spectroscopy peaks seen in fig . [ fig:1]e consists of regularly - spaced resonances positioned according to the condition @xmath171.2 ghz being resonant with the energy level separation . notably , however , for this particular value of driving amplitude , the n=1 , 3 , and 6 photon peaks are missing . as we describe in this section , the presence and absence of these peaks arise from mach - zehnder - type quantum interference at a level crossing . the interference phase is tunable via the driving amplitude , leading to a `` bessel staircase '' interference pattern in the observed spectroscopy . the interference oscillations are known as stckelberg oscillations @xcite , and they have been observed in both artificial @xcite and natural @xcite atomic systems . in a conventional mach - zehnder interferometer , an optical signal is passed through two beamsplitters . the first beamsplitter coherently divides the signal into two output paths , which may have different effective optical path lengths . these paths are then recombined at the second beamsplitter , where the signal waves interfere and exit the interferometer through the two output ports . an intensity measurement at either output port exhibits interference fringes depending on the relative path length . here , we instead utilize lzs transitions at level crossings as beamsplitters for the atomic state @xcite . we drive the persistent - current qubit with a harmonic driving field of the form @xmath60 with @xmath61 the driving frequency and @xmath62 the field amplitude ( in units of energy ) , which is proportional to the rf voltage at the source . as illustrated in fig . [ fig:2]d , the qubit state undergoes two lzs transitions during one period of the driving field . the first lzs transition at time @xmath63 splits the qubit into a superposition of excited and ground states . a relative phase @xmath59 accumulates @xmath64 until the second lzs transition at time @xmath65 , at which point the qubit state collides with itself and interferes . interference fringes appear at half - integer and integer values of @xmath66 , which is tunable via the drive amplitude . it is clear , however , that a second phase must also play a role in this problem , since the qubit state continues to evolve for the remainder of the driving period . it is physically meaningful to consider the total phase @xmath67 accumulated over a single period : @xmath68 which is _ independent _ of the driving amplitude . over many periods of the driving field , the cascaded pairs of lzs transitions ( cascaded interferometers ) will constructively interfere provided @xmath69 . one can view this as a `` time - domain '' formulation for the familiar n - photon resonance condition @xmath70 where n is the number of photons involved in the transition . it is only when the product @xmath71 equals the energy separation @xmath72 that the cascaded lzs transitions lead to a net buildup of state population and , as a function of @xmath59 , the observed interference fringes . these oscillations are related to photoassisted transport @xcite and rabi oscillations @xcite in the multiphoton regime . mach - zehnder - type interference in the discrete - resonance limit driven towards saturation is shown in fig . [ fig:3]a for driving frequency @xmath73 mhz . this frequency is larger than the resonance linewidth ( fig . [ fig:3]b ) , and so individual n - photon resonances can be resolved . there are two main features observable in this plot . the first is the presence of equally - spaced n - photon transitions as a function of flux detuning , symmetrically located about the qubit step at @xmath5 . as one might expect , the onset of the higher - photon transitions requires larger driving amplitudes . remarkably , we observe up to 50-photon transitions in this scan . the second main feature is that for each n - photon resonance , the spectroscopy appears and disappears as a function of amplitude , which sweeps the interference phase @xmath59 . for example , 14 oscillation lobes are visible for the 1-photon transition . a vertical slice of the spectroscopy ( dashed white line in fig . [ fig:3]a ) is plotted in fig . [ fig:3]c ; at this particular amplitude , one can see the enhancement and suppression of the spectroscopy peaks as was observed in fig . [ fig:1]e . the mach - zehnder interference for an n - photon transition yields a modified amplitude - dependent matrix element @xcite @xmath74 where @xmath75 is the nth - order bessel - function , and its argument @xmath76 is the dimensionless driving amplitude . intuitively , the bessel - function dependence arises , because the qubit is driven harmonically through energy levels that are linear in flux detuning and , as a result , the interference phase accumulates with a harmonic time dependence . the transition rate and , therefore , the population transfer , at each n - photon resonance is related to the matrix element squared , @xmath77 . at specific amplitudes , despite driving the system resonantly , no net transition occurs due to a complete destructive mach - zehnder interference corresponding to the zeros of @xmath75 ; this is the coherent destruction of tunneling @xcite condition for driven n - photon transitions @xcite . more generally , as a function of the driving amplitude , we have observed the continuous evolution between regions of enhanced and suppressed tunneling as dictated by @xmath75 over several bessel - function lobes ( fig . [ fig:3]a ) . in the discrete resonance limit , n - photon resonances are distinguishable , because the coherence time of the qubit is sufficiently longer than the driving period . by reducing the frequency to @xmath78 mhz ( fig . [ fig:3]a , right panel ) , we effectively made the drive period comparable with the linewidth , @xmath79 ( fig . [ fig:3]b ) . in this spectroscopy plot , the individual n - photon resonances are no longer resolvable . however , the mach - zehnder interference fringes ( vertical slice , fig . [ fig:3]c ) are clearly visible , because the qubit remains coherent during the critical fraction of the drive period during which the phase @xmath59 accumulates and the mach - zehnder interference occurs . intuitively , provided @xmath80 , the output populations of each quantum interferometer are then preserved ( frozen ) until the subsequent interferometer is reached in the following period ; therefore , although the resonance condition is lost , the mach - zehnder quantum interference remains . this behavior can be contrasted with driven rabi oscillations , for which there would be no signature of coherence in the limit @xmath79 . the previous discussion involved driving transitions in the lowest two energy levels in the double - well potential of our artificial atom ( fig . [ fig:4]a ) , which constitute the two - level qubit subsystem of a more complex energy level diagram ( fig . [ fig:5]c ) . when higher - excited states are accessed , the driven system behavior can be markedly different from the population saturation observed when only two levels are involved . for example , at least three levels are required to achieve population inversion , and such a multi - level artificial atom coupled to a microwave cavity has been used to demonstrate masing ( microwave lasing ) @xcite . in that work , josephson quasi - particle states were driven to achieve inversion . alternatively , population inversion can be established by accessing an ancilliary excited state via direct or lzs transitions . this will be briefly discussed in the next section . here , by reversing the cycle that leads to population inversion , we show that one can pump population from the qubit excited state @xmath81 to the qubit ground state @xmath82 ( fig . [ fig:4]a ) via an ancillary energy level @xmath83 . in the case where the population in @xmath81 results from thermal excitation , the transfer of population to @xmath82 effectively cools the qubit by lowering its effective temperature . this kind of active cooling represents a means to initialize and reset qubits with high fidelity , key elements for quantum information science and technology . alternatively , the pumping mechanism can be used to refrigerate environmental degrees of freedom @xcite , or to cool neighboring quantum systems @xcite . more explicitly , for a qubit in equilibrium with its environment , the population in @xmath84 that is thermally excited from @xmath82 follows the boltzmann relation @xmath85 , \label{boltzmann}\ ] ] where @xmath86 are the qubit populations for energy levels @xmath87 , @xmath88 , @xmath89 is the boltzmann constant , and @xmath90 is the bath temperature . to cool the qubit subsystem below @xmath90 , a microwave magnetic flux of amplitude @xmath91 and frequency @xmath12 targets the @xmath92 transition , driving the state-@xmath84 thermal population to state @xmath83 , from which it quickly relaxes to the ground state @xmath82 . efficient cooling only occurs when the driving - induced population transfer to @xmath82 is faster than the thermal repopulation of @xmath84 . the hierarchy of relaxation and absorption rates required , @xmath93 , is achieved in our system owing to a relatively weak tunneling between wells , which inhibits the interwell relaxation and absorption processes @xmath94 and @xmath95 , compared with the relatively strong intrawell relaxation process @xmath96 . figure [ fig:4]b shows the qubit step at @xmath90 = 150 mk in equilibrium with the bath ( top ) and after a 3-@xmath16s cooling pulse at 5 mhz ( bottom ) . under equilibrium conditions , the average level populations exhibit a thermally - broadened qubit step about @xmath97 , with a width proportional to @xmath90 . the presence of microwave excitation targeting the @xmath92 transition , followed by relaxation , acts to increase the ground - state population and , thereby , sharpens the qubit step . cooling can thus be quantified in terms of an effective temperature @xmath98 , a signature that is evident from the narrowing of the qubit steps in fig . [ fig:4]b after cooling . more precisely , using the notation from fig . [ fig:2 ] , the effective qubit temperature is obtained by fitting an effective temperature that would have been required in equilibrium to achieve the observed qubit population @xmath99 , @xmath100.\ ] ] universal cooling ( cooling that is independent of flux detuning ) occurs near an optimal driving amplitude @xmath101 ( fig . [ fig:4]c ) . this is demonstrated in fig . [ fig:4]d where we present the @xmath82 state population @xmath102 measured as a function of the microwave amplitude @xmath91 and flux detuning @xmath103 for @xmath104 . cooling and the diamond feature can be understood in terms of the energy level diagram ( fig . [ fig:5]c ) . as the amplitude of the microwave pulse is increased from @xmath105 , population transfer first occurs when the @xmath106 avoided crossing is reached , _ i.e. _ @xmath107 ; this defines the front side of the observed diamond , symmetric about the qubit step ( see also fig . [ fig:3]a ) . for amplitudes @xmath108 , the @xmath109 ( @xmath110 ) side avoided crossing dominates the dynamics , resulting in a second pair of thresholds @xmath111 , which define the back side of the diamond . in the region outside of the diamond s backside , the qubit is cooled . as the diamond narrows to the point @xmath112 , the narrowest qubit step is observed . this is the universal cooling condition : only one of the two side avoided crossings ( @xmath109 or @xmath110 ) is reached and , thereby , strong transitions with relaxation to the ground state result for all @xmath103 . in contrast , for @xmath113 , both side avoided crossings ( @xmath109 and @xmath110 ) are reached simultaneously for @xmath114 , leading once again to a large population transfer between @xmath82 and @xmath84 , and opening the second diamond feature ( see fig . [ fig:5 ] ) . the cooling exhibits a rich structure as a function of driving frequency and detuning , resulting from the manner in which state @xmath83 is accessed ( fig . [ fig:4]c ) . transitions occur via a ( multiphoton ) resonant or adiabatic passage process when the driving frequency is high or low enough , respectively @xcite . at high frequencies ( 800 and 400 mhz in fig . [ fig:4]e ) well - resolved resonances of n - photon transitions are observed and cooling is thus maximized near resonances . at intermediate frequencies ( 400 and 200 mhz ) , mach - zehnder interference at the side avoided crossing @xmath115 becomes more prominent and modulates the intensity of the n - photon resonances @xcite . below @xmath116 , individual resonances are no longer discernible , but as in fig . [ fig:3]c , the modulation envelope persists @xcite . at the lowest frequencies ( @xmath117 ) , state @xmath83 is reached via adiabatic passage through the @xmath115 crossing ( fig . [ fig:4]c ) , and the population transfer and cooling become conveniently independent of detuning ( see @xmath118 in fig . [ fig:4]e ) . as shown in fig . [ fig:4]e , we achieve an effective qubit temperature @xmath119 = 3 mk , even for @xmath90 = 300 mk . in our qubit , our determination of @xmath119 was limited primarily by decoherence ( linewidth ) , which limited the resolution with which we could distinguish the states @xmath82 and @xmath84 near degeneracy . nonetheless , we can estimate the ideally resolvable cooling factor @xmath120 for this type of cooling process using eq . [ boltzmann ] , @xmath121 where @xmath122 ghz is the energy separation where the relaxation @xmath96 occurs and @xmath123 ghz for our qubit , yielding a cooling factor @xmath124 . for a bath temperature @xmath125 mk , this would correspond to an effective temperature @xmath126 @xmath16k in our qubit . cooling a qubit in equilibrium with the bath requires a characteristic cooling time . in turn , a cooled qubit will thermalize to the environmental bath temperature over a characteristic equilibration time . the relationship between these two times determines if it is possible to drive the qubit while it is still cold . we found in this qubit that equilibration times are at least one order of magnitude larger than cooling times at @xmath127 mk and up to three order of magnitude larger at @xmath128 mk @xcite . this allowed us ample time to drive the qubit after cooling it . the implementation of an active cooling pulse prior to a generic driving pulse is highly advantageous . on the one hand , it sensibly shortens measurement times , enabling us to acquire data at repetition rates that far exceed the intrinsic equilibration rate due to interwell relaxation after each measurement trial . by adopting active cooling , we gained a factor 50 in data acquisition speed , limited by the bandwidth of our readout circuit . on the other hand , by analogy to the cooling of trapped ions and atoms , active cooling greatly reduces thermal smearing , allowing us to analyze features in the data that would have been hidden otherwise . this type of active cooling protocol was required to obtain the detailed , clean data in fig . [ fig:3 ] , where we could clearly resolve resonances separated by only 270 mhz ( 13 mk ) . in fact , the necessity of active cooling becomes even more evident in the next section where , without the cooling pulse , the observed level of detail could not have been resolved over such a large parameter space in practical acquisition times . frequency - dependent absorption and emission spectroscopy has long played a fundamental role in the characterization of quantum systems . the development of coherent microwave ( maser ) and optical ( laser ) sources , high - intensity radiation with tunable , narrow spectral linewidth , has further enabled targeted absorption spectroscopy of atoms and molecules with high frequency resolution @xcite . however , the application of broadband frequency spectroscopy is not universally straightforward . this is particularly relevant for artificial atoms engineered from solid - state materials which , when cooled to cryogenic temperatures , assume quantized energy levels that extend into microwave and millimeter wave regimes ( 10 - 300 ghz ) . although certainly not an impossible task , a broadband frequency - based spectroscopic study of these atoms in excess of around 50 ghz becomes extremely challenging and expensive to implement due to numerous frequency - dependent effects ( _ e.g. _ , frequency dispersion and the requisite tolerances to control impedance ) . amplitude spectroscopy is a technique that allows broadband spectroscopic characterization of a quantum system . with amplitude spectroscopy , spectroscopic information is obtained from the system response to driving - field amplitude at a _ fixed _ frequency . the resulting spectroscopic interference patterns , `` spectroscopy diamonds , '' are mediated by multilevel lzs transitions and mach - zehnder - type interferometry , and they serve as a fingerprint of the artificial atom s multilevel energy spectrum ( figs . [ fig:5]a and [ fig:5]b ) . the energy spectrum is then determined by analyzing the atomic fingerprint . in this way , the amplitude spectroscopy technique complements frequency spectroscopy : although a less direct approach , it allows one to probe the energy level structure of a quantum system over extraordinarily large ( even practically prohibitive ) bandwidths by circumventing many of the challenges associated with a frequency - based approach . in general , the spectroscopy diamonds arise due to an interplay between the static flux detuning @xmath129 and driving amplitude @xmath11 . as described in sections [ sec : lzs ] and [ sec : mz ] , transitions occur when an avoided crossing is reached for a particular set of values ( @xmath129,@xmath11 ) . for example , at a flux detuning @xmath130 , the diamond boundaries occur at @xmath131 ( fig . [ fig:5 ] a and c ) . the mach - zehnder interference due to a phase accumulation @xmath132 at a given avoided crossing @xmath133 can be modulated by varying both @xmath129 and @xmath11 . there are two important contributions to the diamond spectroscopy patterns : lzs - mediated transitions and intrawell relaxation . in fig . [ fig:5]d , we show a subset of the diamond interference pattern in fig . [ fig:5]a . arrows indicate lines of constant phase accumulation @xmath134 in ( @xmath129,@xmath11)-space that leads to lzs transitions at each of the three listed avoided crossings , @xmath106 , @xmath109 , and @xmath110 . where these lines cross , competition ( coordination ) between avoided crossings act to suppress ( enhance ) the net transition rate between pairs of energy levels . the arrangement of these crossing lines leads to the checkerboard patterns observed inside and outside the diamonds . the second contribution , intrawell relaxation , provides another means to connect energy levels and results in both cooling and population inversion . in section [ sec : cooling ] , [ fig:4]c , @xmath109 mediated the transition @xmath92 , and intrawell relaxation then mediated the transition @xmath96 ; the net result was cooling , since the flux @xmath129 was positive , making @xmath82 the ground state . however , in fig . [ fig:5]d @xmath129 is negative . furthermore , in the upper - right corner of fig . [ fig:5]d , both crossings @xmath109 and @xmath110 are accessed . in the bright red regions , @xmath109 still tends to cause transitions @xmath92 , and relaxation puts that population in state @xmath82 . however , the interference condition at @xmath110 on the other side of the energy level diagram ( fig . [ fig:5]c ) tends to keep the population in @xmath82 . therefore , population builds in state @xmath82 , the first excited state for negative flux detuning , resulting in strong population inversion . varying the interference conditions at @xmath109 and @xmath110 by changing ( @xmath129,@xmath11 ) causes the observed modulation between population inversion and cooling . we have developed several techniques for extracting information about the energy levels from the spectroscopy interference patterns @xcite . the key metrics are the positions of the avoided crossings in flux , the values of the splittings @xmath133 , and the slopes of the energy levels . with this information , one can reconstruct a large portion of the energy level diagram . the positions of the avoided crossings can be determined precisely from the diamond boundaries , because the onset of each diamond is associated with lzs transitions at a particular level crossing . the splitting of each avoided crossing can be obtained essentially by fitting the lzs formula in eq . [ eq : p_lzs ] to the mach - zehnder interference patterns . alternatively , one can study the dynamical population transfer between states using the pulsed , short - time implementation of amplitude spectroscopy @xcite . the energy - level slopes can be determined by two means . the first is by relating the separation between mach - zehnder interference nodes to the expected phase accumulation @xmath59 , which depends sensitively on the energy - level slope . alternatively , we show with rudner _ _ in ref . that the two - dimensional fourier transform of the diamond patterns yields a family of _ one - dimensional _ sinusoids in fourier space ; the periods of these sinusoids are related to the energy - level slopes . the intuition is that the fourier transform inverts the energy domain of the spectroscopy to the time domain ( scaled by @xmath135 ) . this means that the sinusoids in fourier space image the time - dependence of the quantum phase of the qubit , which , in turn , depends sensitively on the energy - level slopes . using amplitude spectroscopy , we were able to scan the energy level diagram continuously beyond the fourth energy - level avoided crossing ( @xmath136 , @xmath137 ) with splitting values ranging from @xmath138 ghz to @xmath139 ghz @xcite . the equivalent information obtained using frequency spectroscopy would have required scanning frequencies from 0.01 ghz out to beyond 100 ghz ( in this device , each avoided crossing is separated by approximately 25 ghz , and so each subsequent crossing in fig . [ fig:5 ] is raised an additional 25 ghz above the ground state ) . remarkably , with amplitude spectroscopy , the entire scan performed in fig . [ fig:5]a was performed at a fixed frequency @xmath140 mhz . the scan in fig . [ fig:5]b shows the amplitude spectroscopy of the same system for a fixed frequency @xmath141 mhz , clearly in the discrete resonance limit . the resonance condition adds another constraint , making a more complex and interesting checkerboard pattern . strongly driving a superconducting artificial atom through an avoided level crossing results in a landau - zener - stckelberg transition , which , in general , creates a superposition of atomic states whose weighting depends on the size of the avoided crossing and the velocity with which it is traversed . in this sense , as we discussed in section [ sec : lzs ] , the lzs mechanism acts as a beamsplitter for artificial atoms . in section [ sec : mz ] , we described how harmonically driving the system cascades two lzs transitions per driving period , resulting in an atomic analog to a mach - zehnder interferometer . the relative phase acquired between lzs transitions is the interference phase of the interferometer , and it is tunable by the driving amplitude . the buildup of population over many driving periods exhibits stckelberg oscillations as a function of the driving amplitude ( interference phase ) due to the cascaded mach - zehnder - type interference effect . to observe these oscillations , the coherence time must only be longer than the small portion of the drive period during which the interference phase accrues , and the relaxation time must be long enough to maintain the population until readout . the total phase accumulated over one period , in contrast , is amplitude independent . for coherence times longer than the drive period , cascaded interferometers yield a net population change when this round - trip phase accumulation is @xmath142 per driving period , a condition which can be viewed as the `` time - domain '' counterpart to the @xmath143-photon resonance condition @xmath144 . by making the drive period commensurate with the coherence time , we showed that we could still observe the stckelberg oscillations , even though the discrete resonances were no longer discernable . we utilized strong driving and the lzs mechanism with higher - energy levels to achieve both cooling and population inversion in our artificial atom . in section [ sec : cooling ] , we described using a microwave pumping scheme to cool the atomic degrees of freedom a factor 10 - 100 times colder than the ambient dilution refrigerator temperature . the scheme involved pumping unwanted thermal population to an ancillary excited state , from which it relaxed to the ground state . in section [ sec : as ] , we showed that by reversing the order , we could pump population through an ancillary state to achieve inversion . the energy level structure can be probed over extraordinarily large bandwidth using the amplitude spectroscopy approach presented in section [ sec : as ] . since the lzs mechanism and mach - zehnder interference are sensitive to the defining features in the energy level diagram ( energy band slopes , level splittings , and their locations ) , the interference patterns that result from sweeping the amplitude are a `` fingerprint '' of the artificial atom s energy spectrum . using amplitude spectroscopy at a fixed driving frequency of only 185 mhz , we could access continuously multiple energy levels from about 10 mhz out to beyond 120 ghz . large - amplitude driving and the lzs mechanism have application to quantum information science and technology . active cooling has utility in state initialization and refreshing ancilla qubits in quantum error correction protocols . amplitude spectroscopy provides a means to ascertain over large bandwidth the energy level structure of a qubit system beyond the lowest two levels . the mach - zehnder - type interference can facilitate nonadiabatic control schemes , in which the quantum interference at an avoided crossing is used to achieve state transitions that approach the intrinsic coupling rate @xmath6 . in cold atoms and molecules , this kind of non - adiabatic control has been used to drive transitions that would otherwise be challenging to realize in a weak - driving limit . we thank t.p . orlando , d.m . berns , l.s . levitov , m.s . rudner , a.v . shytov , and k.k . berggren with whom we collaborated closely on this work ; j. bylander , s. gustavsson , b. turek , and j. sage for carefully reading the manuscript ; v. bolkhovsky , g. fitch , t.r jordan , e. macedo , p. murphy , k. parrillo , r. slattery , and t. weir at mit lincoln laboratory for technical assistance . this work was supported by afosr and lps ( f49620 - 01 - 1 - 0457 ) under the durint program , and by the u.s . the work at lincoln laboratory was sponsored by the us dod under air force contract no . fa8721 - 05-c-0002 . c. h. van der wal , a. c. j. ter haar , f. k. wilhelm , r. n. schouten , c. j. p. m. harmans , t. p. orlando , s. lloyd , j. e. mooij , quantum superposition of macroscopic persistent - current states , _ science _ * 290 * , 773 - 777 ( 2000 ) . a. j. berkley , h. xu , r. c. ramos , m. a. gubrud , f. w. strauch , p. r. johnson , j. r. anderson , a. j. dragt , c. j. lobb , f. c. wellstood , entangled macroscopic quantum states in two superconducting qubits , _ science _ * 300 * , 1548 - 1550 ( 2003 ) . a. izmalkov , m. grajcar , e. ilichev , n. oukhanski , th . wagner , h .- meyer , w. krech , m. h. s. amin , a. maassen van den brink and a. m. zagoskin , observation of macroscopic landau - zener transitions in a superconducting device , _ europhys . _ 65 , 844 - 849 ( 2004 ) . a. izmalkov , s.j.w . van der ploeg , s.n . shevchenko , m. grajcar , e. ilichev , u. hbner , a.n . omelyanchouk , and h .- g.meyer , consistency of ground state and spectroscopic measurements on flux qubits , _ phys . _ 101 , 017003 ( 2008 ) . m. grajcar , s.h.w . van der ploeg , a. izmalkov , e. ilichev , h .- meyer , a. fedorov , a. shnirman , and g. schn , sisyphus cooling and amplification by a superconducting qubit , _ nature phys . _ 4 , 612 - 616 ( 2008 ) . murali , z. dutton , w.d . oliver , d.s . crankshaw , and t.p . orlando , probing decoherence with electromagnetically induced transparency in superconductive quantum circuits , _ phys . * 93 * , 087003 ( 2004 ) . z. dutton , k.v.r.m . murali , w.d . oliver , and t.p . orlando , electromagnetically induced transparency in superconducting quantum circuits : effects of decoherence , tunneling , and multilevel crosstalk , _ phys . b _ * 73 * , 104516 ( 2006 ) . leek , j.m . fink , a. blais , r. bianchetti , m. goppl , j.m . gambetta , d.i . schuster , l. frunzio , r.j . schoelkopf , and a. wallraff , observation of berry s phase in a solid - state qubit , _ science _ * 318 * , 1889 - 1892 ( 2007 ) . a. wallraff , d. i. schuster , a. blais , l. frunzio , r .- s . huang , j. majer , s. kumar , s. m. girvin and r. j. schoelkopf , strong coupling of a single photon to a superconducting qubit using circuit quantum electrodynamics , _ nature _ * 431 * , 162 - 167 ( 2004 ) . j. johansson , s. saito , t. meno , h. nakano , m. ueda , h. tanaka , k. semba , and h. takayanagi , vacuum rabi oscillations in a macroscopic superconducting qubit lc oscillator system , _ phys . * 96 * , 127006 ( 2006 ) . f. deppe , m. mariantoni , e.p . menzel , a. marx , s. saito , k. kakuyanagi , h. tanaka , t. menon , k. semba , h. takayanagi , e. solano , and r. gross two - photon probe of the jaynes - cummings model and controlled symmetry breaking in circuit qed , _ nature phys . _ * 4 * , 686 - 691 ( 2008 ) . fink , m. goppl , m. baur , r. bianchetti , p.j . leek , a. blais , and a. wallraff , climbing the jaynes - cummings ladder and observing its root n nonlinearity in a cavity qed system , _ nature _ * 454 * , 315 - 318 ( 2008 ) . a. fragner , m. goppl , j.m . fink , m. baur , r. bianchetti , p.j . leek , a. blais , and a. wallraff , resolving vacuum fluctuations in an electrical circuit by measuring the lamb shift , _ science _ * 322 * , 1357 - 1360 ( 2008 ) . m. hofheinz , e.m . weig , m. ansmann , r.c . bialczak , e. lucero , m. neeley , a.d . oconnell , h. wang , j.m . martinis , and a.n . cleland , generation of fock states in a superconducting quantum circuit , _ nature _ * 454 * , 310 - 314 ( 2008 ) . van der ploeg , a. izmalkov , a. maassen van den brink , u. hbner , m. grajcar , e. ilichev , h .- meyer , and a.m. zagoskin , controllable coupling of superconducting flux qubits , _ phys . * 98 * , 057004 ( 2007 ) . j. majer , j.m . chow , j.m . gambetta , j. koch , b.r . johnson , j.a . schreier , l. frunzio , d.i . schuster , a.a . houck , a. wallraff , a. blais , m.h . devoret , s.m . girvin , and r.j . schoelkopf coupling superconducting qubits via a cavity bus , _ nature _ * 449 * , 443 - 447 ( 2007 ) . r. mcdermott , r. w. simmonds , m. steffen , k. b. cooper , k. cicak , k. d. osborn , s. oh , d. p. pappas , j. m. martinis , simultaneous state measurement of coupled josephson phase qubits , _ science _ * 307 * , 1299 - 1302 ( 2005 ) . m. steffen , a. ansmann , r.c . bialczak , n. katz , e. lucero , r. mcdermott , m. neeley , e.m . weig , a.n . cleland , and j. m. martinis , measurement of the entanglement of two superconducting qubits via state tomography , _ science _ * 313 * , 1423 - 1425 ( 2006 ) . m. steffen , a. ansmann , r. mcdermott , n. katz , r.c . bialczak , e. lucero , m. neeley , e.m . weig , a.n . cleland , and j. m. martinis , state tomography of capacitively shunted phase qubits with high fidelity , _ phys . lett . _ * 97 * , 050502 ( 2006 ) . m. neeley , m. ansmann , r.c . bialczak , m. hofheinz , n. katz , e. lucero , a. oconnell , h. wang , a.n . cleland , and j. m. martinis , process tomography of quantum memory in a josephson - phase qubit coupled to a two - level state , _ nature phys . _ * 4 * , 523 - 526 ( 2008 ) . n. katz , m. ansmann , r.c . bialczak , e. lucero , r. mcdermott , m. neeley , m. steffen , e.m . weig , a.n . cleland , j.m . martinis , a.n . korotkov coherent state evolution in a superconducting qubit from partial - collapse measurement , _ science _ * 312 * , 1498 - 1500 ( 2006 ) s. yoakum , l. sirko , p. m. koch , stckelberg oscillations in the multiphoton excitation of helium rydberg atoms : observation with a pulse of coherent , field and suppression by additive noise _ phys . * 69 * , 1919 - 1922 ( 1992 ) . f. lang , p.v.d . straten , b. brandsttter , g. thalhammer , k. winkler , p.s . julienne , r. grimm , and j. hecker denschlag , cruising through molecular bound - state manifolds with radiofrequency , _ nature physics _ * 4 * , 223 - 226 ( 2008 ) . p. k. tien and j. p. gordon , multiphoton process observed in the interaction of microwave fields with the tunneling between superconductor films _ phys . * 129 * , 647 ( 1963 ) .

the thermodynamic limit in which the system size is taken to be infinite is beloved by theorists as many calculations in many - body physics become dramatically simplified in this limit . this must be contrasted with experimental work , in which one tends to make measurements on systems of a finite size . for bulk thermodynamic measurements this discrepancy is typically not important , the reason being that avogadro s number is so large as to be practically indistinguishable from infinity . for example , phase transitions are strictly speaking only true transitions in the thermodynamic limit where the free energy exhibits singularities , however their signatures becomes arbitrarily sharp in large but finite systems as to make this mathematical technicality irrelevant . as nanotechnology improves and nano - devices become smaller , finite - size corrections to the thermodynamic limit become more relevant . the same is true as computers improve and computational approaches become more commonplace . particularly for strongly correlated systems , one often turns to methods based on exact diagonalization or stochastic sampling of the partition function ( monte - carlo methods ) , which tend to be limited to relatively small ( certainly compared to avogardo s number ) system sizes . to extrapolate from these results on small systems to systems of experimental relevance requires an understanding of the finite - size scaling ; even typical experimental nano - devices while tiny compared to bulk systems are large compared to the number of particles / sites that may be simulated in the fully quantum methods necessary for understanding strongly correlated systems . however one of the more fascinating results of modern theoretical physics is that , at least in a large class of models , the finite - size scaling often reveals more information about the system in question than the bulk result itself . the most beautiful example of this occurs in systems with conformal invariance @xcite , where the ground state energy of the system per unit length @xmath5 of a system of length @xmath6 follows the relationship @xcite @xmath7 the crucial thing in this relationship is that the coefficient @xmath8 appearing in the @xmath9 finite - size scaling term is the central charge of the system . this is a measure of the number of gapless degrees of freedom in the bulk system , and can not be simply determined from knowledge of @xmath10 alone . when one turns to transport through non - equilibrium nano - structures , one finds a similar dichotomy between theory and experiment in the time domain . this is most apparent in the study of full counting statistics ( fcs ) of charge transfer @xcite . the fcs is the study of the full probability distribution @xmath0 that @xmath1 electrons have moved from the source lead to the drain lead in the measuring time @xmath2 , typically when driven by a bias voltage @xmath11 . this is a probability distribution as the exact number measured in any given realization is subject to both quantum and thermal fluctuations . in a typical experimental measurement of the fcs @xcite , one counts ( having first found a way to do so which is not a trivial matter , but is not the subject of this manuscript ) the number of electrons that has flown through the nano - structure in a given time , then repeats the experiment many times to get the distribution . it is clear that in the long - time limit , the average number @xmath12 of electrons moved , and by extension all moments of the distribution @xmath0 , should be proportional to the measuring time @xmath2 , and indeed the theoretical work has mostly concentrated on the long - time limit @xmath13 . however in the experiments to date @xcite , the source - drain bias voltage @xmath14 is so low that the relevant dimensionless parameter @xmath15 is actually rather small . in addition , modern numerical techniques to determine the fcs in strongly correlated systems via real - time simulations @xcite are computationally very expensive and as such can only simulate the system for relatively short measurement times . one may therefore ask the question : can one understand the contributions to the fcs that are sub - leading in measurement time , and do such contributions give us useful information about the system in question ? early work in this direction @xcite has concentrated on non - interacting systems ; more recently the present authors have presented a conjecture that extends this work to interacting ( and in particular strongly correlated ) systems @xcite , although this result is limited to zero temperature . in these proceedings , we review this conjecture along with its background , and present the latest numerical evidence for the conjecture . in addition , we discuss more general questions about finite - time corrections to fcs , particularly with reference to the measurement of fractionally charged quasi - particles . to be concrete , we discuss all of this with specific reference to the interacting resonant level model . the structure of the paper is as follows : in section [ sec : fcs - ni ] , we review the background of full counting statistics in non - interacting systems in the scattering matrix ( landauer - bttiker ) approach . in section [ sec : fcs - i ] we conjecture how this may be extended to interacting systems , giving numerical evidence in support of our conjecture . in section [ sec : fcs - pt ] we then discuss one of the more fascinating properties of the interacting resonant level model a non - equilibrium phase transition characterised by a bifurcation in the cumulant generating function corresponding to fractionalisation of the charge - carrying quasi - particles . in particular , we show that the conjecture about finite - time corrections must break down at this critical bias voltage , and the relevance of this to the measurement of fractional charge . finally in section [ sec : end ] , we summarise our results and open questons . throughout the paper , we use units where planck s constant @xmath16 , and the unit of electrical charge @xmath17 unless otherwise specified . note the slightly unusual choice to use units in which planck s constant ( @xmath18 ) rather than the reduced constant ( @xmath19 ) is one . using this choice , the fundamental unit of conductance @xmath20 becomes unity in the dimensionless units , which is convenient for the quantum transport phenomena described in these proceedings . rather than study directly the probability distribution @xmath0 of @xmath1 electrons ( or in other words , a charge of @xmath21 in our dimensionless units ) being moved from the source lead to the drain lead in measuring time @xmath2 , it is often more convenient to take the fourier transform which gives the generating function of the distribution @xmath22 here , @xmath23 is known as the counting field , and it is usual to further take the logarithm of the above expression to obtain the _ cumulant generating function _ ( cgf ) , @xmath24 the derivatives of the cgf give the irreducible moments @xmath25 substituting the definition in eqs . and into this expression , we see that the first cumulant @xmath26 where @xmath27 is the current flowing through the system , while the second cumulant @xmath28 where @xmath29 is the zero - frequency shot noise . in fact to make the above expressions rigorous in quantum mechanics , one must supplement the definition with a prescription for the appropriate time - ordering of the operator @xmath30 as the current operator does nt commute with itself at different times ; we refer to the literature @xcite for a full discussion on this . the cgf encodes all information about the statistics of charge transfer through the nanostructure , however one of the easiest quantities to obtain from the cgf is the charge on the ( quasi)particles involved in transport . from the basic definition , one sees that if only single charges @xmath31 ( @xmath32 if units are restored ) may be transferred between the leads , then @xmath33 must be a @xmath34 periodic function . if however one has , for example , a superconductor where only cooper pairs may be transferred with charge @xmath35 , one will see that @xmath36 @xcite . more interestingly , if one sees a periodicity larger than @xmath34 , then this is an indication that the particles involved in transport are fractionally charged @xcite . such fractionally charged quasi - particles appear in strongly correlated systems , where the charge may even change with bias voltage @xcite . we will discuss this point in more detail in section [ sec : fcs - pt ] . we add here one final comment about the periodicity of the cgf . as @xmath33 is given by a complex logarithm , , the imaginary part of @xmath37 is only defined modulo @xmath34 . the imaginary part of the cgf can therefore contain a linear component @xmath38 and still be classed as periodic . as we will see later , such a linear component is associated with perfect transmission through the system , and thus is often seen in model systems as it has a particular physical significance . as mentioned earlier , the average charge transferred between the leads in time @xmath2 is expected to grow linearly in measuring time @xmath2 , and by extension one expects all cumulants to exhibit similar behavior . hence the cumulant generating function is expected in the long time limit @xmath39 . it has been shown that the subleading corrections to this at zero temperature @xmath40 are logarithmic @xcite in nature , and hence one may write @xmath3 as a formal expansion in the small parameter @xmath41 : @xmath42 for plotting and comparison to numerical work it is more convenient to take the time derivative of this , @xmath43 from hereon when we refer to the cgf , we will most typically mean @xmath44 , however this should always be clear from context . for non - interacting system , it is most convenient to use the scattering ( landauer - bttiker ) formalism ( see e.g. @xcite ) which describes transport properties in terms of the transmission @xmath45 through the system . within this approach , the leading term in the expansion of the cgf was derived by levitov and lesovik @xcite and is given by @xmath46 . \label{eq : ll}\ ] ] here , @xmath47 and @xmath48 are the source and drain voltages respectively , the applied bias @xmath49 . in order to understand this expression we note that if the transmission is independent of energy @xmath50 , then this gives @xmath51 $ ] , which is just the cgf for the bernoulli distribution where each event occurs with probability @xmath52 . a full introduction to this may be found in ref . . we now turn our attention to the subleading correction , which has been found @xcite to be given by @xmath53 . \label{eq : ni - f1}\ ] ] while we refer to the original papers for the technical details of the derivation of this expression , it is worth making some comments on the origin of these contributions to the cgf that are logarithmic in measuring time . the cgf may be represented as the determinant of a matrix @xcite which has a toeplitz form . the leading contribution to this determinant is then given by szego s theorem @xcite ; this is the levitov - lesovik result . however it is known that if the matrix elements have certain singularities then the determinant acquires logarithmic corrections according to the fisher - hartwig conjecture @xcite . in this case , the singularities come from the fermi edges which are infinitely sharp at zero temperature , the correction being eq . which was first derived in ref . . the crucial point above is that the logarithmic corrections are due the sharp fermi edges . the first implication of this is that at non - zero temperature when the fermi edges are smooth , the long time expansion _ does not _ contain the ( non - analytic ) logarithmic term . at low temperatures @xmath52 however , there is a crossover time @xmath54 ; at measurement times @xmath55 the cgf will include the logarithmic terms as if the system were at @xmath40 , crossing over at long times @xmath56 to a different temperature - dependent result @xcite . the second implication is that the coefficient of the logarithmic term , , depends on the system _ only _ in the vicinity of the fermi edges . this is clearly true in eq . where the expression depends on the transmission only at the fermi energies of the left and right leads ( which are different due to the bias voltage ) . we emphasize this point as it will turn out to be very important when we generalize these results to interacting systems in section [ sec : fcs - i ] . before we do this however , we look at an example of the evolution of the cgf as a function of measuring time in a non - interacting system . , then the system is the non - interacting resonant level model . ] the resonant level model ( rlm ) is a fermionic model of a single ( spinless ) state hybridized with two leads , which for convenience are modelled as tight - binding chains . the hamiltonian of the system , which is represented schematically in fig . [ fig : irlm_lattice ] is given by @xmath57 in this expression , @xmath58 is the fermionic creation operator on lead @xmath59 ( the left and right leads correspond to the source and drain in the transport set up ) and site @xmath1 , while @xmath60 is the creation operator on the resonant level . the hopping parameter of the leads is taken to be @xmath61 ( this is conventionally notated as @xmath62 in condensed matter , however we find @xmath61 more convenient in non - equilibrium situations when @xmath62 is the time ) , and the length of each lead is @xmath63 . while this is typically taken to be infinite in analytic calculations , numerical simulations are done with finite @xmath63 , see appendix [ sec : size ] . finally , the hybridization between the resonant level and the leads is given by @xmath64 . as this is a non - interacting problem , the transport through the resonant level when a bias voltage is applied between the two leads may be calculated in the scattering matrix formalism . the transmission @xmath45 through the dot as a function of energy @xmath65 in the rlm , is given by ( see e.g. @xcite ) @xmath66 where @xmath67 and @xmath68 . this equation is for the specific model which has tight - binding leads with a cosine band . one often takes the limit of a wide band , corresponding to @xmath69 which gives the more usual lorentzian shape of the transmission @xmath70 where @xmath71 . for comparison with high - precision numerics however , it is necessary to keep the full structure of . substituting the expression into the general levitov - lesovik formula , and performing the integral gives us the leading ( in inverse measuring time ) term of the cgf : @xmath72 \nonumber \\ & & + \frac{4e^{{\ensuremath{\mathrm{i}}}\chi/2}\tan^{-1}\left[\frac{{\ensuremath{v_{\mathrm{sd}}}}}{2}e^{-{\ensuremath{\mathrm{i}}}\chi/2}\sqrt{b^2-a^2(e^{{\ensuremath{\mathrm{i}}}\chi}-1)}\right]}{\sqrt{b^2-a^2(e^{{\ensuremath{\mathrm{i}}}\chi}-1 ) } } \nonumber \\ & & - \frac{4\tan^{-1}(b{\ensuremath{v_{\mathrm{sd}}}}/2)}{b}. \label{eq : f0-rlm}\end{aligned}\ ] ] the subleading term is then determined from to be @xmath73 having obtained analytic expressions and for the coefficients of the series governing the time evolution of the cgf for the rlm , we can now compare this series with the full time evolution as obtained through a real time numerical simulation of the model ( see appendix [ sec : size ] for a brief outline of the numerical method ) . this will allow us to see how well the series , which only contains two terms , compares to real numerical data , and furthermore allows us to see what features the time dependence may exhibit beyond this series . the advantage of doing this for a non - interacting system is that one may obtain numerical data to relatively long times , which is not possible for strongly correlated interacting systems . this will be a useful benchmark from which to interpret interacting results presented later in these proceedings . in fig . [ fig : rlm - time - evolution ] , we plot the real and imaginary parts of the cgf @xmath74 as a function of inverse measuring time @xmath75 for various values of the counting field @xmath23 and fixed values of the bias voltage @xmath76 and model parameters @xmath77 . the hopping parameter on the leads @xmath61 gives an overall energy scale to the problem and is fixed to be @xmath78 . the data points are obtained from numerical simulation , while the solid lines are from the series . we note that in the series , rather than using bare measuring time @xmath2 , we have replaced this by the conformal time @xmath79 , eq . , where @xmath80 is the system size . the conformal time , introduced in ref . @xcite takes into account the leading finite _ size _ effects , and is briefly discussed in appendix [ sec : size ] . we also note that the imaginary part of @xmath81 tends to be positive , while the real part tends to be negative ; the reasons for this can be seen by looking at the leading terms of the cumulant expansion eq . . in order to make the plots as clear as possible , we therefore plot both the real and imaginary parts as positive in most of the graphs , meaning that we plot @xmath82 rather than @xmath83 . the plotting convention is always clearly stated on the axes of each graph . looking at fig . [ fig : rlm - time - evolution ] , one can make several observations . the first observation is that the agreement between the numerical data and the series is excellent over a number of orders of magnitude . it is clear from the plots that the long time asymptotic of @xmath74 is indeed captured by the analytic expressions . the second observation is that the finite measuring time corrections @xmath84 beyond the infinite measuring time limit @xmath85 are much more important for the real part of this particular model . in particular , if one is interested in the real part of @xmath81 and has numerical data only up to @xmath86 which is typical time scales reached in numerics on interacting systems , then one must extrapolate over several orders of magnitude to obtain the infinite time value @xmath85 . we will come back to this point in more detail when we discuss interacting models . it is also worth mentioning that although the contribution from @xmath84 is not visible on the scale of the plot of the imaginary part in fig . [ fig : rlm - time - evolution ] , this contribution is there and agrees well with the analytic formula ; this is demonstrated in fig . [ fig : dfxi ] . the third observation from fig . [ fig : rlm - time - evolution ] is that at smaller times , and in particular for larger values of @xmath23 , there appears to be oscillations on top of the basic series . we believe that these ( decaying ) oscillations are due to the sudden quench of the system when the counting field is switched on , in other words they depend on details of how the stopwatch for the fcs is started and stopped and are therefore do not depend only on the physics of the system in question ( i.e. they are not universal ) . we stress however that this is currently a hypothesis ; future work is needed to understand these oscillations . from the numerical data involved in each of these plots , one may fit the series to the time evolution in order to numerically determine the parameters @xmath85 and @xmath84 . this is plotted in fig . [ fig : dfxi ] and compared to the analytic expressions and to good agreement . in ref . we presented a conjecture for the form of the subleading ( in inverse measuring time ) term in the time evolution of the cgf . here , we review this conjecture , and present the latest numerical evidence supporting it . the scattering formulation which is used in eqs . and can not be used in interacting systems , as such systems are not fully characterised by a single - particle transmission amplitude @xmath45 . one can however ask the question : suppose we have managed to calculate the leading term in the cgf , @xmath88 . from this knowledge , can we derive the size of the subleading term @xmath89 ? by simple comparison of eqs . and , we see that @xmath90 , \label{eq : conjecture1}\ ] ] where @xmath91 . this formulation makes no reference to the single - particle scattering problem and we _ conjecture _ that this expression remains true when interactions are added to the problem . we allow in the conjecture that the constant of proportionality @xmath92 may depend on the interaction strength , although the numerical work indicates that at least for the interacting resonant level model in section [ sec : sdirlm ] , the value @xmath91 is unchanged from the non - interacting value . we stress however that this result can only hold for single channel systems as soon as there is a sum over different transmission channels , no such simple formula can exist . for convenience we limit ourselves to cases which have particle - hole symmetry ; meaning that @xmath93 . in the example of the non - interacting rlm , which satisfies @xmath94 , if one then applies the source and drain biases symmetrically @xmath95 , the conjecture simplifies to become @xmath96 as previously mentioned , in the non - interacting case the logarithmic term in the series arises due to the discontinuity in the fermi function at zero temperature , the above derivative formula indeed relates @xmath84 to the physics at the fermi edges . perturbatively , one may then certainly imagine an effective landau picture of quasi - particles valid near the fermi - surface . these would make their contribution to the overall quantities such as current and noise ( in essence @xmath85 ) , although the calculation of these quantities require knowledge of all of the states in the energy window between the two fermi edges so could not be reliably obtained simply from knowledge of the quasi - particles at the fermi surface . on the other hand the physics around the fermi edge is all that is needed in the determination of @xmath84 , which gives some hope that may remain valid in interacting systems . we emphasize however that this line of reasoning , though intuitively appealing , has no rigour , and remains a conjecture . we can however give two pieces of evidence that it remains true . the first comes from looking at the lowest cumulants , while the second piece of evidence comes from verifying numerically that it holds , at least for one strongly interacting model . let us expand @xmath88 as @xmath97 where @xmath27 is the current and @xmath29 is the zero - frequency shot noise . according to conjecture , one then should have the corrections ( expanded to quadratic order in @xmath23 ) as @xmath98 where @xmath99 is the conductance of the system . the consequences of this are twofold . firstly , we see that there are no ( leading order ) corrections from finite measuring time to the current . secondly , we see that the corrections to noise due to finite measuring time are @xmath100 the significance of this relates to the frequency dependence of noise , which in non - interacting systems is known to be @xmath101 in @xcite it was shown that this relationship remains true to all orders in perturbation theory when interactions are added , at least in a wide class of models , although the constant of proportionality @xmath92 may depend on interaction strength . other work @xcite has suggested that the relationship may break down in the nonperturbative regime ; however recent numerical work @xcite has verified that this relation does indeed hold . a full analytic understanding of this is still lacking however . if one understands that a finite measuring time @xmath2 means that noise can not be probed at zero frequency , and instead is measured at a characteristic frequency @xmath75 , one sees that relations and the conclusion from our conjecture become equivalent . this means that our conjecture is intimately related to the validity of the expression in the presence of interactions . while an analytic proof of this is still an open question , the existing literature is in support of this . we also note that a similar relation between the measuring time dependence and the frequency dependence of the third cumulant has been previously discussed in non - interacting systems @xcite . however there is very little work on the frequency dependence of the third cumulant in the presence of interactions ; this also remains an interesting direction for future study . having discussed the conjecture in the context of the low cumulants , we now turn to the full cgf and show that holds for one interacting model . the interacting resonant level model is the same as the resonant level model , with the addition of an interaction @xmath102 between the resonant level and the leads . the model is schematically represented in fig . [ fig : irlm_lattice ] and the hamiltonian is given by : @xmath103 the quartic fermionic term here is the interaction , while the subtractions of @xmath104 is convenient to retain particle hole symmetry . it turns out that for a certain value @xmath105 , this model has a certain self - duality @xcite that allows the model to be solved exactly via bethe ansatz , even out of equilibrium @xcite the analytic expression for the cgf of the self - dual interacting resonant level model , as derived from thermodynamic bethe ansatz , is given by @xcite : @xmath106 here @xmath107 @xmath108 , and @xmath109 is the radius of convergence of each series . we note that unlike the non - interacting model , the interacting one can not be solved exactly on the lattice , and the given result is only for the wide - band ( field - theoretic ) limit . the non - universal pre - factor @xmath110 relates the regularization of the field theory to the lattice model , and is taken from earlier work @xcite which studied the non - linear @xmath111 characteristics . this means that when we come to compare analytic and numerical results , there are no free fitting parameters and the plots simply show a direct comparison . in order to do this however , parameters must be chosen carefully so that there is not a strong contribution from the finite band width in the numerical result . we will return to the physical interpretation of the cgf in section [ sec : fcs - pt ] ; in this section we concentrate on the finite time corrections and the conjecture . applying the conjecture , and furthermore taking the constant of proportionality @xmath91 to be the same as the non - interacting value , one should then see the subleading term @xmath112 ^ 2 & v_\mathrm{sd}<v_c , \\ \frac{1}{\pi } \left [ \sum_{m > 0 } a_{1/4}(m ) \frac{1 - 3m/2}{m } \left ( \frac{v_{\mathrm{sd}}}{t_b ' } \right)^{-3m/2 } \left ( e^{{\ensuremath{\mathrm{i}}}m \chi/2 } -1 \right ) \right]^2 & v_\mathrm{sd } > v_c . \end{cases } \label{eq : f1-irlm}\ ] ] of the cgf for the interacting resonant level model at different large bias voltages @xmath113 . the symbols give the numerical results , while the solid line follows the analytic result . at these high biases , the subleading correction @xmath84 is too small to be reliably extracted from numerics . ] we are now in a position to compare these analytic predictions for @xmath85 and @xmath84 with results from real time numerical simulations see appendix [ sec : size ] for an overview of the numerical technique , as well as the two original papers @xcite where these results were first presented . numerically , one chooses a value of the counting field @xmath23 , and time evolves the system to obtain @xmath74 . [ fig : sdirlm - time ] shows some sample results of @xmath81 as a function of measuring time @xmath2 for different values of @xmath23 . we mention here that a number of other plots of a similar nature are shown later in fig . [ fig : timeslices ] . in the plots of fig . [ fig : sdirlm - time ] , we also include the analytic result given by the series and the conjecture . these appear to be in good agreement in the figure however one can go a stage further and extract the parameters @xmath85 and @xmath84 by fitting the series to the numerically obtained time evolution . the results of this are shown in figs . [ fig : smallvoltage ] and [ fig : largevoltage ] . we begin by discussing the case @xmath114 , which is shown in fig . [ fig : smallvoltage ] . we begin by discussing the real part . as seen in the figure , there is a very good agreement between the numerically determined @xmath85 and the analytic result . the incredible thing here is that to obtain the numerical result , one must extrapolate @xmath81 about an order of magnitude from the finite - time results available . in other words , for the time scales that can be simulated on the computer , the result is dominated by the @xmath84 contribution one should carefully read the scale on the axis of the graphs of @xmath85 and @xmath84 . looking at the result for the real part of @xmath84 , we also see that the numerical data agrees well with the analytic result . this is strong numerical evidence supporting our conjecture that the formula does indeed hold in the presence of interactions . in fig . [ fig : smallvoltage ] , we have also plotted the quadratic approximation @xmath115 , which is the finite time ( frequency ) correction to the shot noise that has been previously discussed in @xcite . from the plot , it is clear that the conjecture correctly captures the correction to cumulants beyond this . we next look at the imaginary part of @xmath116 , also plotted in fig . [ fig : smallvoltage ] . here it is clear that the expression for @xmath85 fits rather well , however it would be difficult to say the same for @xmath84 . there are several issues going on here though . first , one must again look at the vertical scales of these graphs . for the imaginary part of @xmath81 , the finite measuring time corrections @xmath84 are a very small correction on top of the long - time limit ; as an order of magnitude this is @xmath117 of the signal at @xmath118 , and less as @xmath2 increases . combined with this , there are other contributions to @xmath81 beyond the terms given in the series . in particular , looking at one of the first panels of fig . [ fig : timeslices ] that appears later , one see s that the imaginary part of @xmath81 exhibits certain oscillations presumably of the same ( as yet unknown ) origin as those previously seen in the non - interacting case . this oscillations make it difficult to fit the small overall slope of the numerical data reliably , as the answer depends on the exact range of data used for fitting . combining this with the small overall size of @xmath84 makes the numerical task daunting . we emphasize that this is work in progress , and will be reported elsewhere when enough data is available to make reliable predictions ; however with this in mind , the discrepancy between the analytic formal and the numerical data ( remembering that there are no fitting parameters whatsoever in the analytic formula ) is not as severe as it may look . for completeness , we finally turn our attention to the high voltage regime , @xmath113 , where the analytic result is given by the other series in eqs . and . for these large bias voltages , the overall scale of @xmath84 as compared to @xmath85 is even worse than the previously discussed case and we therefore unfortunately can not fit @xmath84 reliably to the numerical data . instead , we give a comparison between the numerical and analytic @xmath85 in fig . [ fig : largevoltage ] . it is clear there is a good agreement . the advantage of the self - dual interacting resonant level model in testing the conjecture is that analytic results exist for the cgf , which may be easily differentiated and compared to numerics to either support or disprove the conjecture . while we are not aware of exact results for any other models , certain things may be checked , see appendix [ sec : otheru ] where the latest results in this direction are presented . we now return to the analytic formula for the ( long time limit of the ) cgf for the self - dual interacting rlm , and discuss the physical interpretation . we being with the small bias voltage case @xmath119 . the first term of eq . , @xmath120 , corresponds to the cgf of a system with perfect transmission ; in such a case there is no source of noise and all higher cumulants are zero . the fact the ` impurity ' level is tuned to resonance makes it unsurprising that the transport ( at low bias voltages ) may be considered as a weak correction to perfect transmission . the remaining terms in the series give a power series in @xmath121 , and may be interpreted as the weak backscattering current , and it s associated statistics . each term is proportional to @xmath122 which is the cgf for a poissonian process moving charges of @xmath123 from the right to the left ( due to the minus sign this validates the interpretation as the backscattering correction to perfect transmission ) @xcite . the constants are nt all positive , due to quantum interference between the different processes . the important point however is that the backscattered charges are all multiples of @xmath124 , or in other words the quasi - particles that may be backscattered have charge @xmath125 if units are resurrected . in the opposite limit @xmath126 , one notices that the periodicity of @xmath85 is different this time the sum may be interpreted as a sum of poissonian processes carrying charges @xmath127 from left to right ; or in other words the charge on the quasi - particles that may be transmitted in this limit is @xmath128 ( restoring units ) . this fractionalisation , by which the fundamental charge changes from @xmath124 at low bias to @xmath104 at high bias was first observed by studying the fano factor of the shot noise @xcite . the shot noise however gave no indication of how this fractionalisation occurs . from studying the fcs however , we will be able to interpret this as a bifurcation of the cgf at @xmath129 ; in some sense we will show that @xmath129 is the point of a non - equilibrium phase transition @xcite . of the cgf of the self - dual interacting rlm just above and just below the critical voltage @xmath130 . this plot demonstrates the bifurcation at @xmath131 which may be interpreted as a non - equilibrium phase - transition . ] while the series representation of the cgf , is useful away from @xmath129 , at @xmath132 all terms of the series become important and the the behavior of @xmath88 is mathematically opaque . fortunately , one may recognise that the series for high and low voltages are actually hypergeometric and both give different series representations of the same ( generalised ) hypergeometric function . while one may do this for @xmath85 itself , it is far more convenient to take a derivative and study the counting current @xmath133 . carrying out the summation gives @xcite @xmath134 where @xmath135 . while this representation as a rather exotic hypergeometric function , @xmath136 , is not useful in terms of numerical evaluation , the analytic properties of hypergeometric functions are well documented ( see , e.g. @xcite ) . the function above @xmath137 has two branch points at @xmath138 , with branch cuts stretching from these points to @xmath139 . this indicates that for @xmath140 , i.e. for @xmath141 , the function has a single branch , and from the definition of @xmath142 we see that @xmath143 is @xmath144 periodic in @xmath23 , giving the fundamental charge of @xmath124 previously mentioned . on the other hand , if @xmath145 , i.e. @xmath113 , the function crosses the branch cuts at @xmath146 . by choosing @xmath143 to be continuous as a function of @xmath23 , this implies that @xmath143 must now be @xmath147 periodic , which in turn implies a fundamental charge of @xmath104 . we now see that the fractionalisation of charge happens discontinuously at @xmath148 ; the branch point leading to a bifurcation in the cgf . this can be seen in fig . [ fig : bifurcation ] , where the cgf @xmath88 is plotted for @xmath11 slightly above and slightly below @xmath130 . it is worth pointing out however that while this branch cut leads to a discontinuity in the cgf as a function of @xmath11 , this discontinuity occurs at @xmath149 . all of the cumulants , which are given as derivates around @xmath150 , are smooth analytic functions of @xmath11 at @xmath148 . in other words , one needs all of the cumulants , i.e. the _ full _ counting statistics in order to see and understand this transition . the previous subsection analysed only the ( infinitely ) long time limit , @xmath85 , and found a sharp transition due to singularities ( branch points ) in the cgf . we would ask : what is the effect of finite - time on this transition ? there are many important aspects of this question , including : 1 . in analogy to finite size on equilibrium phase transitions , is the singularity present at finite time , or is the dependence on @xmath11 smooth at any finite time becoming singular only in the @xmath4 limit ? 2 . does finite time play a role in our ability to see fractional charges ? in other words , does the @xmath34 periodicity of the cgf remain at all finite times , becoming larger only in the infinite time limit ? while we are not yet able to answer these questions , we now present results about the measuring time dependence of the cgf near the critical voltage @xmath130 that will give a start in understanding these points . it is clear that the conjecture relies on @xmath85 being a smooth function of @xmath11 ; near the critical bias voltage @xmath130 , this is no longer the case . this means that in the vicinity of @xmath130 , the conjecture can not hold , or at the very minimum there must be a lot of physics involved in the measuring time dependence of @xmath81 that is not captured by the simple series and the conjecture . in order to study this numerically , we take a fixed value of @xmath149 ( we choose @xmath151 ) and study @xmath81 as a function of measuring time @xmath2 and bias voltage @xmath11 chosen such that @xmath152 . the results are shown as a density plot in fig . [ fig : coolplot ] . for the model parameters chosen ( @xmath153 ) , one has @xmath154 . superimposed on the plots are various lines @xmath155 , which show that the time evolution of the cgf has distinctive features due to the proximity to the critical point . we would like to point out that there is an analogy between our data in fig . [ fig : coolplot ] and the standard view of ( equilibrium ) quantum critical points @xcite , where inverse measuring time in our case takes the role usually played by temperature . whether or not this analogy is a useful way of thinking of the non - equilibrium problem however is an open question . to investigate further the features in the time evolution that are indicated in fig . [ fig : coolplot ] , we take some vertical cross sections of this plot , in other words we plot @xmath81 as a function of @xmath2 for a selection of values of @xmath11 close to the critical value . this is shown in fig . [ fig : timeslices ] , where the analytic result according to the series and conjecture are also plotted for comparison . we start with the first two graphs , which are for @xmath114 . the imaginary part of @xmath81 fits very well to the conjecture , although with certain oscillations on top of the straight line . we believe these oscillations to be of the same origin as already seen in the non - interacting case , i.e. related to quenching on the counting field , however this still requires further investigation . the real part of @xmath81 also appears to converge to the line given by the conjecture after some initial transients . there seems to be some evidence that the closer @xmath11 is to @xmath130 , the longer one has to wait before convergence however this also requires further investigation . we turn now to the next two plots for @xmath156 and @xmath157 which is just below and just above the critical bias @xmath130 . just below @xmath130 , one sees a continuation of the previous trend @xmath158 fits well to the conjecture though with additional oscillations ; while @xmath159 looks like it will converge to the line of the conjecture , although in this case there is not enough data to say this for certain . the plot for @xmath160 , just greater than @xmath130 is however very revealing . by comparison with the previous plot , we see that the analytic results have jumped , due to the discontinuity at @xmath130 . however , the numerical results are very close to the previous panel for @xmath114 ; in other words the evolution of the numerical ( finite - time ) results appears to be smooth as a function of @xmath130 . this gives strong evidence to support the idea that there is no discontinuity for any finite time as suggested in question one above . assuming that the infinite - time limit results are correct , we expect that the time evolution will eventually cross over to the analytic values according to the conjecture ; the numerics suggest that this crossover will happen at a time scale @xmath161 this statement is further supported by the final two panels of fig . [ fig : timeslices ] which shows the time evolution at slightly larger times . here , we see that the short(er ) time behaviour ( which is similar to an analytic continuation of the cgf below at @xmath114 ) eventually evolves to reach strong peaks in @xmath81 in both the real and imaginary parts . the location of these peaks has already from fig . [ fig : coolplot ] been seen to be at a time given by above . we assume that at times ( significantly ) longer than the location of these peaks , the evolution eventually settles to the analytic long - time values certainly for larger values of @xmath11 the numerics and analytics are in good agreement ( c.f . fig . [ fig : largevoltage ] ) , although at these larger values the strong peaks are not seen in the time evolution . clearly more data is necessary to make any definitive conclusions ; we defer this to a future work . as an alternative visualisation , we can take horizontal cross sections of fig . [ fig : coolplot ] , in other words , plot @xmath81 as a function of @xmath11 for given values of measuring time @xmath2 . this is shown in fig . [ fig : fvv ] . the long time limit @xmath85 shows discontinuities at @xmath148 , however the numerical results from different finite times appear to all be smooth functions of @xmath11 . for @xmath114 , the time sequence converges towards the analytic long - time limit in fact , as shown in fig . [ fig : timeslices ] , it does so according to the conjecture presented in this work . for @xmath113 however , the finite time results develop additional structure in the form of peaks that become narrower as @xmath2 is increased . one can certainly imagine how these peaks tend to the discontinuous analytic result as @xmath4 ; however more data is necessary to investigate this further . finally , let us comment briefly on the role of finite - time on the measurement of fractionally charged quasiparticles . the background of this problem is that if one does a projective measurement of the number of electrons in the lead , one must obtain one of the eigenvalues of the number operator , i.e. an integer . the measurement of fractional charge is therefore intimately related to the quantum mechanical measurement problem . we will not review this in any detail here , but would like to add an extra ingredient to the discussion : the role of finite measuring time . in section [ sec : bif ] , we showed that , at least for the interacting rlm , the fractionalisation of charge occurs mathematically via a bifurcation at a branch point of the cgf . we also demonstrated that at finite times , the cgf appears to be a smooth function . at present , our data does not allow us to hypothesize on the periodicity of @xmath81 as a function of measuring time however we believe this is an interesting question that could shed some light on the experimental detection of fractional charge . to summarise , we have discussed the cumulant generating function @xmath3 of full counting statistics , and in particular the evolution of this function with measuring time . we demonstrated that the leading corrections to the long time limit @xmath162 are logarithmic in nature at zero temperature , and have conjectured that the coefficient of this subleading term is universally related to the long time limit via eq . . while we have presented strong numerical evidence for this conjecture for one specific model , its more general validity is an open question . one promising direction in this regard is based on a keldysh expansion of the cgf @xcite . another open question which is of great importance for extrapolating numerical results to longer times is about the nature of other finite - time corrections to the cgf , such as the oscillations seen in fig . [ fig : rlm - time - evolution ] . while we do nt expect these to be universal , it is important to understand their form and origin in order to fit numerical data to obtain reliable results about the long - time limit of strongly correlated models . our numerical test of the conjecture were on the self - dual interacting resonant level model , , where we further demonstrated that the long - time limit of the cgf , @xmath85 , has branch cuts associated with fractionalize of charge at some critical voltage @xmath130 . we demonstrated that the cgf at finite times however shows no discontinuities ; but does show distinctive features in the time evolution associated with this critical point ; occurring at times satisfying eq . . we furthermore suggested that this may have strong implications in the attempted experimental measurement of fractional charge . understanding the physics at this non - equilibrium critical point is an open question , and we believe one of the most fascinating future directions in the theoretical study of non - equilibrium systems . we thank a. komnik , d. bagrets and d. gutman for insightful discussions . hs work was supported by the french agence nationale pour la recherche ( anr projet 2010 blanc simi 4 : dime ) and the us department of energy ( grant number de - fg03 - 01er45908 ) . to study transport properties in the real time numerical method @xcite , one initially finds the ground state of the system subject to a non - uniform potential of @xmath163 on the left ( right ) lead . this charge imbalance potential is then quenched off at time @xmath164 , and the system is evolved numerically according to the schrdinger equation @xmath165 by calculating the time evolution of the state , one can then evaluate the desired operator . in particular , the protocol by which one can obtain the cgf of fcs by this method was presented in ref . . for interacting systems , we numerically evolve the system using the density matrix renormalization group ( dmrg ) @xcite , while for the noninteracting systems , we use a much more efficient method based on single - particle evolution and slater determinants @xcite . however these particular numerical methods of time evolution have one thing in common : they all work on systems of a finite size , @xmath80 . this means that the leads are not considered infinite and instead consist of @xmath166 lattice sites . there are a number of consequences of evolution on a system of finite size @xcite , the most crucial being that there is a transit time @xmath167 ( @xmath168 being the fermi velocity in the leads ) after which the charge imbalance has flowed across the system , bounces of the boundary and starts coming back thus destroying the steady - state current flow . this therefore provides a hard limit on how long the system may be evolved for ; the numerical equivalent of the battery going flat . the typical sizes ( and therefore maximum times ) that may be simulated on a computer are limited by computer resources , although clearly the single - particle basis for the non - interacting case means that much larger systems may be used there . while we refer to previous work @xcite for a more detailed discussion of the manifestations of finite - size in non - equilibrium transport , we mention here one extremely useful result that is used in these proceedings . in a ( conformally invariant ) system of finite size , one can capture the leading effects of times close to the transit time when the current is beginning to reverse direction by replacing the measuring time with a _ _ @xcite . in the simulations , the counting begins at a time @xmath169 , which is chosen to be sufficiently long after the initial quench at @xmath164 so that the largest transients have died away and the system is in a ( quasi-)steady state . the conformal substitution therefore amounts to replacing the measuring time @xmath170 by the conformal time @xmath171 for short times @xmath172 , we find @xmath173 . the difference only becomes important at times near the transit time . equation is based on the usual conformal transformation between the plane and a cylinder of finite extent in imaginary time @xcite , and analytically continuing to real time . we note that while this procedure is not entirely justified in this setup @xcite , a comparison to numerics shows that it captures the leading effect of the back - reflection off the edges of the system remarkably well @xcite . while there is a lot of work on the fcs of other interacting models out of equilibrium ( see e.g. @xcite ) , there are no exact results available for the cgf . in particular , there is a lot of work on the interacting rlm , for more general values of @xmath102 @xcite , however without exact results for comparison to numerics , it is not easy to check the validity of the conjecture [ eq : conjecture2 ] . this notwithstanding , the conjecture [ eq : conjecture2 ] relates the finite - time corrections @xmath84 to the long - time limit itself @xmath85 for any individual model . hence with enough data over a range of voltages , one can fit the numerical data to the series , numerically differentiate the obtained @xmath174 and hence check the relation completely numerically . this requires a lot of high - quality numerical data and is work in progress . here we present only the first step of this , which is verifying that the corrections to the long time behavior of @xmath81 are indeed proportional to @xmath75 ( see eq . [ eq : series ] ) for the interacting rlm but _ away _ from the self - dual point . in fig . [ fig : nonselfdual ] , we show data for the measuring time evolution of the cgf for the interacting rlm at @xmath175 . this data shows clear @xmath176 behavior , thus demonstrating the validity of the series . from the graphs of @xmath81 , it is also clear what is the numerical difficulty in testing the conjecture . in particular , in the real part of @xmath81 , we see that we must extrapolate the numerical data from typical maximum times available over at least one order of magnitude to get the value of @xmath85 . as @xmath85 must be obtained with high precision in order to carry out a numerical derivative , this is a daunting task . the results of this when complete will be presented elsewhere . m. e. fisher and r. e. hartwig , _ toeplitz determinants : some applications , theorems , and conjectures _ , in _ advances in chemical physics : stochastic processes in chemical physics _ , volume 15 , edited by k. e. schuler ( john wiley and sons , 1969 ) .

a classical nova eruption plays an important role in the evolution of cataclysmic variable stars ( cvs ; see the books by * ? ? ? * ; * ? ? ? it represents the primary mechanism for mass - loss in these systems , preventing ( most of ) them to end up as supernovae ia . irradiation of the donor star by the eruption - heated white dwarf is furthermore suspected to cause an increase in the mass - transfer rate , which would have a strong effect on the evolutionary time - scale @xcite . still , the connection between the nova eruption and the physical parameters of cvs is not very well known , mainly due to the fact that a large fraction of the post - nova systems remain unidentified @xcite . we are carrying out a systematic observing campaign in order to recover and study classical novae that erupted at least 30 yr ago ( earlier than 1986 ) . this implies that sufficient time has passed since the nova eruption for the optical signal to become dominated by the underlying cv instead of representing mainly the ejected material . the targets presented here were selected from @xcite and refer to fields with hitherto ambiguous identification of the post - nova ( el aql , v606 aql , v908 oph , v1149 sgr and v1583 sgr ) . they were investigated by means of @xmath0 photometry to identify potential candidates for the post - nova , based on their colours . the selected candidates were subsequently observed spectroscopically to confirm the post - nova via their spectroscopic characteristics ( mainly the presence of characteristic emission lines ) . in addition , we also obtained spectrograms of candidates for the post - novae v1301 aql , v1151 sgr and v3964 sgr that were included in @xcite . other collections of post - nova spectra can be found in @xcite , @xcite and parts of our own series tappertetal12 - 1,tappertetal14 - 1,tappertetal15 - 2 . @llllllll object & ra ( 2000.0 ) & dec . ( 2000.0 ) & rms ( arcsec ) & date & filter / grating & @xmath1 ( s ) & mag + el aql & 18:56:01.87 & @xmath203:19:18.8 & 0.21 & 2012 - 07 - 18 & @xmath3/@xmath4/@xmath5/@xmath6 & 1576/420/160/120 & 20.9v + & & & & 2015 - 04 - 27 & b600 ( 1.5 arcsec ) & 4760 & 19.9r + v606 aql & 19:20:24.29 & @xmath200:08:07.8 & 0.30 & 2012 - 06 - 24 & @xmath3/@xmath4/@xmath5/@xmath6 & 1576/420/160/120 & 20.4v + & & & & 2015 - 06 - 15 & b600 ( 1.5 arcsec ) & 3600 & 20.1r + v908 oph & 17:28:04.58 & @xmath227:43:04.4 & 0.33 & 2014 - 04 - 03 & @xmath3/@xmath4/@xmath5/@xmath6 & 1655/495/200/170 & 20.5v + & & & & 2015 - 04 - 28 & b600 ( 1.5 arcsec ) & 3600 & 18.5r + v1149 sgr & 18:18:29.85 & @xmath228:17:24.9 & 0.21 & 2012 - 08 - 15 & @xmath3/@xmath4/@xmath5/@xmath6 & 1576/420/160/120 & 18.4v + & & & & 2015 - 06 - 06 & b600 ( 1.5 arcsec ) & 1350 & 18.1r + v1583 sgr & 18:15:26.53 & @xmath223:23:17.5 & 0.27 & 2012 - 07 - 24 & @xmath3/@xmath4/@xmath5/@xmath6 & 1576/420/160/120 & 20.5v + & & & & 2015 - 04 - 21 & b600 ( 1.5 arcsec ) & 3600 & 19.5r + v3964 sgr & 17:49:42.42 & @xmath217:23:34.8 & 0.17 & 2015 - 04 - 26 & b600 ( 1.5 arcsec ) & 3600 & 19.0r + the @xmath0 photometric data for the objects with ambiguous identification were taken in two service mode runs in 2012 ( el aql , v606 aql , v1149 sgr and v1583 sgr ) and in 2014 ( v908 oph ) at the antu 8 m telescope that is part of the very large telescope of the european southern observatory ( eso ) , at cerro paranal , chile , using the focal reducer / low dispersion spectrograph ( fors2 ; * ? ? ? * ) system with the high - throughput broad - band filters u_high , b_high , v_high and r_special . series of four and three frames per filter were obtained in the 2012 and the 2014 observations , respectively . in service mode , fors2 usually operates with a mosaic of two 2k@xmath74k mit ccds . for the data obtained on 2012 august 18 , exceptionally , the blue - sensitive e2v ccds were used . the field of view of fors2 is 6.8@xmath76.8 arcmin@xmath8 . after combining the two mosaic ccd frames into one with the fsmosaic routine from the fors instrument mask simulator package , the further reduction was performed using the ccdred package of iraf . the bias - subtracted and flat - fielded data were then corrected for the telescope offsets between them and averaged to a single frame per filter . a 3@xmath9 clipping algorithm was applied to exclude bad pixels and cosmic rays . subsequent photometry was performed using iraf s daophot package and the stand - alone daomatch and daomaster routines @xcite . the aperture radius for the photometry was chosen to be slightly smaller than the full width at half - maximum ( fwhm ) of the point spread function ( psf ) . the conversion to calibrated magnitudes was obtained using observations of standard fields @xcite . the @xmath3 passband is not a part of eso s standard calibration plan , and in most nights , the standard field observations in this filter covered only a very limited range of airmasses that did not allow us to derive a value for the extinction from them . thus , a standard value for the extinction was used that was taken from the la silla observatory website . while this observatory is located roughly 480 km south of cerro paranal , comparison of the @xmath10 passbands and also the @xmath3 data for the few nights , where a sufficiently large airmass range was covered , showed identical values within the uncertainties . similarly , for some nights , also the @xmath10 standard data proved insufficient . for those filters , median values for a 60 d range are available at the observatory s health check monitor . the spectroscopic data were obtained using the gemini multi - object spectrograph ( gmos ; * ? ? ? * ) mounted at the gemini - south 8 m telescope situated on cerro pachn , chile . three spectra were taken per object , in most cases subsequently . since 2014 , gmos is equipped with a row of three 2048@xmath74176 hamamatsu ccds . grating b600 was employed with a 1.5 arcsec slit and a two and four pixels binning in spatial and dispersion direction , respectively . this combination yielded a typical wavelength range of 4100 7100 and a resolving power of @xmath11800 . the grating was operated at a central wavelength of 5600 to put the two 0.915 mm wide gaps between the chips at regions of comparatively low scientific interest . the typically affected wavelength ranges are 5030 5080 and , in principle , 6100 6050 . however , all observations suffered from a bright large - scale distortion in spatial direction that severely affected the signal at a much broader wavelength range of roughly 5900 6200 and could not be properly corrected for . the data were reduced using the gemini gmos extension for iraf as contained in the ureka software package . after subtraction of an averaged bias frame and subsequent division by a flat - field that was normalized by fitting a high - order cubic spline , a two - dimensional wavelength correction was applied that was obtained from spectra taken with a cuar comparison lamp . after this transformation , the individual spectra were combined employing a 2@xmath9 clipping algorithm to correct for cosmic rays . the sky was subtracted by fitting a chebychev function of typically third order in spatial direction . finally , the spectra were extracted and corrected for the instrumental response function by comparison with a spectrum of the standard star ltt 3864 taken on 2015 february 25 . since all calibrations are based on this single observation of a standard , we emphasize that , while the spectral energy distribution ( sed ) of our objects should be sufficiently well reproduced , the zero - point is not well determined , and thus the data can not be regarded as flux calibrated . above steps were performed both for the combined and the individual spectra to examine them for variability . the details of the observations of the confirmed post - novae are summarized in table [ obslog_tab ] . the coordinates in columns two and three were determined by calculating an astrometric correction on the photometric @xmath6 fors2 images or the @xmath12 gmos acquisition frames using starlink s gaia tool ( version 4.4.6 ) with the us naval observatory ccd astrograph catalogue version 4 @xcite . the associated root mean square ( rms ) is given in column four . the fifth column states the date of the observations referring to the start of the night , while the sixth column gives details on the instrumental setup , i.e. filter or grating and slit , and column seven contains the total exposure time per object . finally , the brightness of the object at the time of the observations is presented in column eight , with the corresponding passband indicated by a letter . for the spectroscopic data , the magnitudes were determined via differential photometry on the acquisition frames with respect to five comparison stars in the vicinity of the target . these values were calibrated with respect to previous photometric data , where available , or with the guide star catalogue , version 2.2 @xcite . @lllll object & @xmath5 & @xmath13 & @xmath14 & @xmath15 + el aql & 20.88(03 ) & @xmath20.10(15 ) & 0.62(08 ) & 0.70(08 ) + v606 aql & 20.41(03 ) & @xmath20.28(15 ) & 0.40(07 ) & 0.44(07 ) + v908 oph & 20.54(11 ) & @xmath20.10(15 ) & 1.02(15 ) & 1.23(17 ) + v1149 sgr & 18.36(05 ) & @xmath21.44(15 ) & 0.30(12 ) & 0.30(07 ) + v1583 sgr & 20.54(04 ) & @xmath20.16(18 ) & 0.96(09 ) & 0.83(06 ) + data . in each field , the post - nova is indicated by a circle . ] @llllllllllll object & & & bowen & & civ + & 4340 & 4861 & 6563 & 4922 & 5016 & 6678 & 7065 & 4645 & 4686 & 5412 & 5812 + el aql & 26(1 ) & 35(1 ) & 47(1 ) & 4(1 ) & 3(1 ) & 8(1 ) & 5(1 ) & & 3(1 ) & 4(1 ) + v606 aql & 18(1 ) & 21(1 ) & 26(1 ) & 4(1 ) & 2(1 ) & 4(1 ) & 2(1 ) & & & + v908 oph & 1(1)@xmath8 & 1(1)@xmath8 & 3(1 ) & & & & & & & & + v1149 sgr & 3(1 ) & 2(1 ) & 6(1 ) & & & 1(1 ) & & 5(1 ) & 3(1 ) & 1(1 ) & 3(1 ) + v1583 sgr & 6(1 ) & 9(1 ) & 26(5 ) & 3(1 ) & & 4(1 ) & 2(1 ) & 5(1 ) & 2(1 ) & & + v3964 sgr & 3(1 ) & 5(1 ) & 12(1 ) & 1(1 ) & & 2(1 ) & & 3(1 ) & 2(1 ) & & + + @xmath16 bowen/ blend . + @xmath8 emission cores in absorption troughs . colour colour diagrams for the fields of reported novae with ambiguous identification for the post - nova are presented in fig . [ ubvr_fig ] . we have excluded objects with photometric errors @xmath170.1 mag or @xmath180.005 mag to minimize contamination by very faint sources and artefacts , respectively . candidates for subsequent spectroscopy were selected based on their position in the colour colour diagram with respect to the other stars in the field . an additional selection criterion was the distance to the reported coordinates of the nova , taking into account their detection history . for example , the search area for novae that were identified on photographic plates was limited to the central 1 arcmin@xmath8 . subsequent long - slit spectroscopic data were used to confirm the post - nova . the spectra of the confirmed novae are presented in fig . [ allspec_fig ] , their photometric data are summarized in table [ ubvr_tab ] . we briefly describe the methods used to analyse the spectroscopic properties that apply to all targets . the emission line strength is characterized via the equivalent width @xmath19 , which is calculated with respect to three manually selected wavelength ranges . two of them are used to define the continuum and the noise on either side of the line , while the third one contains the line itself . we then integrate over the range of the latter divided by a linear fit to the continuum . the uncertainty of the resulting @xmath19 is estimated via a monte carlo simulation by repeating the measurement a thousand times to a variation of the data with respect to the addition of a random value within the noise interval . the resulting equivalent widths are presented in table [ eqw_tab ] . we make an attempt at recovering the intrinsic sed by correcting for the interstellar extinction in the line of sight . we use the interface on nasa s infrared science archive ( irsa ) web pages to obtain values @xmath20 from @xcite . these serve as input for the dereddening routine implemented in iraf and based on the relations from @xcite . a standard value for the ratio of the total to the selective extinction @xmath21 is employed . in order to characterize the corrected continuum slope , we fit the wavelength range between 5000 and 7000 with a power law @xmath22 . for the intrinsic sed , the exponent @xmath23 then depends on the mass - transfer rate and the system inclination . larger values for @xmath23 correspond to higher mass - transfer rates ( hotter and thus bluer discs ) and lower inclinations ( a larger fraction of the hotter inner disc is visible ) . however , there are large uncertainties involved in the reddening correction , because these are large - scale extinction maps with a resolution of 2@xmath72 deg and local fluctuations may be present . furthermore , the exact depth position of the nova with respect to the absorbing material is also unknown . these uncertainties propagate to the derived @xmath23 , which should thus be taken only as an approximate indicator of the intrinsic sed . line profiles for the four individual spectra of el aql in sequence from bottom to top . the vertical line indicates the rest wavelength of h@xmath23 . ] the nova was discovered by @xcite as a star of ninth magnitude on photographic plates taken on 1927 july 30 and 31 . examination of pre - discovery harvard plates by @xcite revealed the object having reached its peak brightness before that date . the brightest recorded value is 6.4 mag on 1927 june 15 , which is thus taken as maximum brightness , although observational gaps before and after that date leave room for an even brighter peak value . the decline of the nova is comparatively well covered , especially in the first 200 d after eruption . description of the spectral evolution during the decline is found in a number of publications . the most detailed data can be found in @xcite , while @xcite present the only drawing of the spectrum . in spite of the good photometric coverage , there is no complete representation of the long - term light curve in the literature . the plot in @xcite is limited to the first 100 d , and @xcite does not include the harvard data . in fig.[elaqllc_fig ] , we collect the observations from @xcite , @xcite , @xcite , @xcite , @xcite , @xcite and @xcite . the dates and brightness values from @xcite were extracted from his plot with a precision of 0.25 d and 0.1 mag , respectively . all other data were available in tabulated form . for data where only the start of night was given , ut = 0 h was assumed for the observation . we omitted from the plot the first four data points from @xcite , where he describes the target as ` blurred ' , because they differ drastically from those of other authors . the upper limits in the plot are from @xcite and @xcite . for reasons of visualization , we omit another upper limit ( @xmath1716 mag ) from 1927 may 28 @xcite . after the initial decline , the nova experienced a series of minor rebrightenings and then settled on a prolonged plateau phase . there also appear to be further fluctuations around 400 and 500 d after eruption . these are based on the @xcite data only , but since in the overlapping regions his magnitudes usually agree within 0.2 mag to those of other authors , there is no reason to doubt the reality of these rebrightenings . to our knowledge , there are no further data on the nova after the seasonal gap between 540 and 720 d after eruption . assuming that this is not due to the lack of observations , the final decline must have occurred sometime during that period . @xcite observed the field again on 1940 june 5 and reported a brightness of 19 mag for the nova , but , as we argue below , this is a likely misidentification . the decline light curve is classified in @xcite as ` ba ' that includes novae with ` standstills or other minor irregular fluctuations during decline ' . although el aql is not included in the @xcite catalogue , its light curve likely corresponds to the ` p ' ( plateau ) class , and in particular , shows some similarity to hs sge , especially with respect to the rebrightenings . in el aql , the major peak around day 63 makes it difficult to define the end of the initial decline and the start of the plateau phase . if we select the end of this event as the corresponding point , we find the following approximate values for the light - curve parameters as defined by @xcite : the time of the start of the plateau phase , and thus the duration of the initial decline , @xmath2487 d , the duration of the plateau phase @xmath25450630 d , the brightness difference between the start of the plateau phase and the peak magnitude @xmath263.9 mag , and the slope of the decline during the plateau phase to @xmath27 . these are typical values for class p novae and justify the inclusion of el aql into this category . @xcite find that this class contains a large fraction ( 6090% ) of the recurrent novae and suggest that some of the other systems in this class are yet unrecognized recurrent novae . the remaining members of the p class novae appear to be magnetic systems . a candidate for the post - nova was selected from our @xmath0 photometric data on the basis of its colours ( fig . [ ubvr_fig ] , table [ ubvr_tab ] ) and its proximity ( 3 arcsec ) to the coordinates reported in @xcite . the spectrum of this object presents emission lines typical for a cv and we conclude that it is indeed the post - nova . its brightness is @xmath28 mag and thus considerably fainter than the 19th magnitude star observed by @xcite in 1940 . @xcite gives a value of @xmath29 mag for el aql for observations made in 1989 . it is , in principle , possible that both observations above caught the post - nova in a bright state ( or our data were taken in a low state ) . however , we consider this unlikely , because cvs in bright states are also expected to have bluer colours than in low states , and the object observed by @xcite is significantly redder in the visual range ( @xmath30 mag compared to our 0.62 mag , see table [ ubvr_tab ] ) . we thus conclude that those observations did not concern the post - nova , but a different object . similarly , @xcite gives infrared data @xmath31 mag , @xmath32 mag , @xmath33 mag for el aql . this led @xcite to suspect the nova to contain a giant secondary star . however , the spectral appearance of the post - nova ( fig . [ allspec_fig ] ) and its position in the colour colour diagram does not support this interpretation , and it thus again appears more likely that @xcite observed a different object . we furthermore note that the resulting eruption amplitude ( defined as the difference between the brightness of the post - nova and the maximum brightness ) of 14.5 mag basically excludes el aql as a recurrent nova candidate . el aql presents a rich emission line spectrum ( fig . [ allspec_fig ] , table [ eqw_tab ] ) . the balmer lines are exceptionally strong for an old nova , which might be indicative of a low mass - transfer rate @xcite . the strength of heii @xmath344686 and the presence of heii @xmath345412 points to a magnetic system , but since the former does not quite match the strength of h@xmath35 , el aql is more likely to be an intermediate polar rather than a discless system . the continuum has a slightly red slope , but as can be seen from the comparison of the colour colour diagrams ( fig . [ ubvr_fig ] ) , the field of el aql is significantly affected by interstellar reddening . from the irsa website , we find @xmath20 = 1.11(08 ) mag , and when corrected for that value , an inverse power law with an exponent @xmath36 provides an excellent fit to the continuum slope . the spectroscopic data on el aql consist of four individual spectra . in fig . [ elaqllprof_fig ] , we examine the variation of the h@xmath23 line profile over the course ( @xmath111 h ) of our observations . we find that the line as a whole shows a significant doppler shift . estimating the shift in radial velocity units by fitting a single gaussian function to the line profiles yields a variation of @xmath37@xmath11100 km s@xmath38 over the observed time range . furthermore , also the shape of the line profile appears to be variable , pointing to the presence of an additional emission source in the system , such as from an irradiated secondary star or a bright spot ( e.g. * ? ? ? * ) . from the gathered evidence , we conclude that el aql is a likely intermediate polar seen at a moderately high inclination . in spite of its faintness , its orbital period might therefore be accessible with time series photometry . line profiles for the three individual spectra of v606 aql in sequence from bottom to top . the vertical line indicates the rest wavelength of h@xmath23 . ] this is a little observed nova that was discovered by fleming as an object of seventh magnitude on a photographic plate from 1899 april 21 , as reported by @xcite . the brightness of the nova on this plate was later more precisely determined to 6.75 mag by @xcite . there is a @xmath11200 d gap of observations prior to the discovery , and it is thus possible that the real maximum has been missed . @xcite estimates the peak brightness to 5.5 mag . the observed decline light curve as shown in @xcite covers @xmath11160 d , and suggests a classification as type ` p ' @xcite . in 1991 , @xcite obtained spectroscopy of the suspected post - nova , but found a g7k4 type spectrum for a @xmath5 = 16.3 mag object . the south - western component of an unresolved visual binary that was marked in @xcite as the possible post - nova showed promising colours ( fig . [ ubvr_fig ] , table [ ubvr_tab ] ) . subsequent spectroscopy of the @xmath5 = 20.4 mag object confirmed the nova ( fig . [ allspec_fig ] ) . the spectrum presents comparatively strong balmer emission lines , although not as strong as in el aql . the high excitation lines are even considerably weaker , and we do not detect heii @xmath345412 . as discussed in section [ elaql_sec ] , most members of the ` p'-class appear to be recurrent or magnetic novae . even if we use the actually observed brightness maximum to calculate the eruption amplitude , the result of @xmath39 mag appears much too large for a recurrent nova @xcite . on the other hand , the spectroscopic properties do not present particularly strong evidence for a magnetic nature either . apparently , v606 aql is an unusual member of the p class . the spectrum presents a slightly blue continuum slope , and the colour colour diagram shows that the field is only moderately affected by interstellar extinction . after dereddening with a value @xmath20 = 0.35(03 ) mag , it became apparent that the continuum was distorted by a number of large ` bumps ' . after examining the three individual spectra , we concluded that the bumpy structure is not intrinsic but an artefact . however , the presence of the bumps rendered large parts of the continuum unusable for fitting . we thus chose a number of small regions of continuum between the blue emission lines as well as a couple of regions around 6000 to determine an inverse power - law exponent of @xmath40 2.02(02 ) . when examining the h@xmath23 line profile of the individual spectra , the changes are not as pronounced as in el aql , and the dopper shift of the whole profile is not as large ( fig . [ v606aqllprof_fig ] ) . still , the shape of the line profile is clearly variable , and the width of the emission lines is basically identical to those in el aql . we thus tentatively conclude that also v606 aql is likely seen at a moderately high inclination , which makes it a potential target for follow - up time series observations . very few data are available for v908 oph . its discovery was reported by @xcite who detected oi emission in a near - infrared spectrum taken on 1954 , july 2 . he estimated the brightness in this spectral range to 9 mag , but suspected that the nova was observed about a week after its maximum brightness . the only other observations consist of two objective prism spectra taken on 1954 july 27 by @xcite . apart from strong h@xmath23 emission , they could not distinguish more detailed spectral features due to the object having declined to a blue magnitude ` probably less than 14 ' . if the object was indeed observed several days after having reached its maximum brightness , and accounting for the colour differences between the two observations and for the strong interstellar reddening , the time @xmath41 to have declined by 3 mag below maximum brightness is likely to be in the order of 3040 d , which would classify the object as a moderately fast nova . the coordinates of the nova had only been recorded with a precision in the order of arcmin , but fortunately , a candidate for the post - nova can be easily identified in the colour colour diagram ( fig . [ ubvr_fig ] ) . its spectrum presents a few weak hydrogen emission lines , with those bluewards of h@xmath23 being embedded in absorption troughs , and a red continuum slope ( fig . [ allspec_fig ] ) . from the @xmath0 photometry , it is clear that the field is strongly affected by interstellar reddening , and the respective catalogue yields an accordingly high @xmath20 = 1.26(02 ) mag . once corrected for this value , the sed corresponds to a steep blue power law with a negative exponent @xmath42 . thus , the spectral characteristics are consistent with a high mass - transfer system . in fig . [ v908ophlprof_fig ] , we show the soectral regions around h@xmath35 and h@xmath23 . in more detail . we note that high excitation emission lines like the bowen blend and heii are absent or , at least , very weak . this is in stark contrast to other likely high mass - transfer post - novae like gr sgr . towards the red side of h@xmath23 emission there appears to be another peak that could be identified with the cii @xmath436578,6583 doublet . this line has been previously detected in a few old novae with sufficiently narrow h@xmath23 emission lines @xcite , and the latter also applies to v908 oph . however , the signal - to - noise ratio ( s / n ) in that part of the spectrum is low , and there are a number of similar large features close - by , which we were unable to identify and thus suspect to be artefacts . in consequence , the detection of cii in this object remains ambiguous . we finally remark that we did not detect any significant positional variation of the h@xmath23 line in the three individual spectra . together with the narrowness of that line , this suggests that the system is seen at low inclination , and that it will be difficult to determine its orbital period . the nova was discovered by @xcite on an objective prism plate at a photographic brightness of about 9 mag . later research by @xcite revealed the maximum brightness to 7.4 mag . the light curve is described as a ` very steep increase in brightness [ ... ] followed by a rapid initial decrease ' , reaching a plateau by 1946 may 28 . the nova was visible from 1945 may 16 to 1948 may 30 , and afterwards fainted below 14.7 mag . large gaps in the observations prevent a more detailed analysis . from our @xmath0 photometry ( fig . [ ubvr_fig ] ) , we selected two potential candidates for the post - nova based on their colours and coordinates . spectroscopy showed one of those to present emission lines and a blue continuum , and we thus identify this object as the post - nova . it is located @xmath1111 arcsec south - west of the previously reported position . we note that these coordinates do not coincide with the objects labelled as v1149 sgr in @xcite and in @xcite . the spectrum shows a blue continuum , that becomes somewhat steeper ( @xmath44 ) when corrected for the catalogued interstellar reddening of @xmath20 = 0.39(03 ) mag , and weak emission lines ( fig . [ allspec_fig ] ) . in contrast to v908 oph ( section [ v908oph_sec ] ) , v1149 sgr counts with a strong bowen blend and heii @xmath344686 emission ( fig . [ v1149sgrlprof_fig ] ) . since no other heii lines can be identified , the system is not a strong candidate for a magnetic cv . the balmer lines are comparatively narrow , and a bump on the red wing of h@xmath23 could tentatively identified with cii ( see also section [ v908oph_sec ] ) . the three individual spectra do not show any significant radial velocity variations , and thus , like v908 oph , this object is likely to be seen at a rather low system inclination . line profiles for the three individual spectra of v1583 sgr in sequence from bottom to top . the vertical line indicates the rest wavelength of h@xmath23 . ] an examination of archival photographic plates by @xcite led to the discovery of this nova almost 30 yr after its eruption . still , in contrast to other novae , the maximum brightness of 8.9 mag on 1928 june 24 appears to be well established . the authors report that two days before that night , the nova was fainter than 15.6 mag and that it fell again below the plate detection limit 75 d after maximum brightness . the system is classified as a moderately fast nova with a decline time @xmath45 d @xcite . from its colours ( fig . [ ubvr_fig ] , table [ ubvr_tab ] ) and its position only 3.5 arcsec off the reported coordinates , we identified a candidate for the nova . subsequent spectroscopy show a slightly reddish continuum ( fig . [ allspec_fig ] ) , which is likely the consequence of a comparatively strong interstellar reddening ( @xmath20 = 1.09(04 ) mag ) . like for most of the other systems , the corrected sed presents a blue slope , with @xmath46 . superposed are moderately strong emission lines ( table [ eqw_tab ] ) of the balmer and hei series . similar to v908 oph ( section [ v908oph_sec ] ) , the bowen blend and heii @xmath344686 lines are weak and hardly detectable . the emission lines are also comparatively narrow , but still the line profile shows some clear asymmetric structure and marked differences between the individual exposures ( fig . [ v1583sgrlprof_fig ] ) . we especially notice that the line strength in the third spectrum of that series is significantly diminished , which indicates that the emission source is being obscured . a possible explanation for this apparent discrepancy between narrow lines and an obscuration effect is that the system is seen at high inclination , but that the doppler - broadened component of the line profile produced by the disc is weak and that the line profile is dominated by a narrow component that originates on the secondary star . in any case , the observed obscuration makes v1583 sgr an attractive target for time series photometry . line profiles for the three individual spectra of v3964 sgr in sequence from bottom to top . the vertical line indicates the rest wavelength of h@xmath23 . ] in , we had observed a candidate for this nova that turned out to be a k1v star . however , examination of the 2d spectra showed the possibility of h@xmath23 emission from a different star in the slit , and a likely source of this emission was identified . here , we present further spectroscopic observations of that target . the spectrum in fig . [ allspec_fig ] clearly reveals it as the post - nova system , presenting the typical hydrogen and helium emission lines , as well as a weak bowen blend . the continuum has a red slope , but like in the other objects , this is due to interstellar reddening . still , even after the correction for @xmath20 = 0.56(01 ) mag , the slope appears significantly flatter than in the other systems , being represented by an exponent @xmath23 = 1.43(03 ) . the emission lines are not particularly strong ( table [ eqw_tab ] ) , but broad , and even show , at times , a hint of a double - peaked disc profile ( fig . [ v3964sgrlprof_fig ] ) . furthermore , like in v1583 sgr ( section [ v1583sgr_sec ] ) , the emission line in the third spectrum is significantly diminished compared to the other two spectra . on the other hand , a possible doppler shift of the line is less clear . still , the width of the line , the potential obscuration effect and the flat slope of the corrected continuum suggest that v3964 sgr is seen at a comparatively high inclination . thus , it should be feasible to determine the orbital period with time series photometry . together with v3964 sgr ( section [ v3964sgr_sec ] ) , we had also reported on candidates for the novae v1301 aql and v1151 sgr that turned out to be misidentifications . for both targets , we had indicated alternative candidates in the immediate vicinity of the original targets . in the current observing run , we incorporated some of these new candidates , namely the objects marked as ` w1 ' and ` w2 ' for v1301 aql , and the one marked as ` 1 ' in the case of v1151 sgr . all objects show rather unremarkable spectra with reddish continua and balmer lines in absorption . ` 1 ' turned out to be a close visual binary , but neither component presented any emission lines . these two post - novae thus remain unidentified . likely , additional photometric data will be needed to select further candidates for spectroscopy . @lllllllllllllllll object & @xmath47@xmath16 & @xmath48@xmath16 & @xmath49 & @xmath50 & @xmath41 & @xmath20 & @xmath23 & @xmath51 & & @xmath52 + & ( mag ) & ( mag ) & ( mag ) & ( yr ) & ( d ) & ( mag ) & & ( ) & ( ) & ( km s@xmath38 ) & ( arcsec ) + el aql & 6.4@xmath53 & 20.9@xmath5 & 14.5 & 88 & 25 & 1.11 & 3.12 & 47 & 27 & 1230 & 3 + v606 aql & 6.8@xmath53 & 20.4@xmath5 & 13.6 & 116 & 65 & 0.35 & 2.02 & 26 & 28 & 1280 & 2 + v908 oph & @xmath189@xmath54 & 20.5@xmath5 & @xmath1711.5 & 61 & @xmath5530 & 1.26 & 3.38 & 3 & 12 & 550 & 234 + v1149 sgr & 7.4@xmath53 & 18.4@xmath5 & 11.0 & 70 & @xmath18270 & 0.39 & 2.41 & 6 & 21 & 960 & 11 + v1583 sgr & 8.9@xmath53 & 20.5@xmath5 & 11.6 & 87 & 37 & 1.09 & 3.27 & 26 & 17 & 780 & 4 + v3964 sgr & @xmath188.5@xmath53 & 19.0@xmath6 & @xmath1711.5 & 40 & 32 & 0.56 & 1.43 & 12 & 29 & 1320 & 3 + + @xmath16 @xmath53 : photographic , @xmath5 : @xmath5 band , @xmath6 : @xmath6 band , @xmath54 : near - infrared . + in table [ prop_tab ] we have collected several distinguished properties of the here presented post - novae . these are the maximum brightness @xmath47 , the brightness of the post - nova @xmath48 at the time of our observations and the resulting eruption amplitude defined as @xmath56 . in the next two columns , we give the time @xmath50 that has passed between the nova eruption and our spectroscopic data and the time @xmath41 by which the nova had declined to 3 mag below maximum brightness . the parameters in the subsequent two columns are associated with the determination of the continuum slope , representing the catalogued value for the interstellar reddening @xmath20 and the negative exponent @xmath23 of the fitted power law ( section [ results_sec ] ) . the next three columns give selected properties of the h@xmath23 emission line : the equivalent width @xmath51 and the fwhm of a gaussian fit to the line , the latter both in wavelength and in velocity units . the final column states the deviation of the position of an object from the originally recorded coordinates of the nova eruption as given by @xcite . since the absolute maximum brightness of novae shows a scatter of typically only 23 mag @xcite , but the absolute brightness of cvs in quiescence is distributed over a range of @xmath119 mag ( e.g. , * ? ? ? * ) , differences in @xmath49 will mostly reflect the difference in the brightness of the post - nova . the latter principally depends on the brightness of the accretion disc ( or , more generally , on the emission from the accretion process ) , and thus on the mass - transfer rate @xmath57 and on the system inclination @xmath58 @xcite . because @xmath47 can be assumed to be largely independent of these two quantities , @xmath49 can be taken as an indicator of the absolute brightness of the post - nova and of the involved parameters . apart from the spread in the absolute maximum brightness , additional uncertainties are introduced by the possibility that the recorded @xmath47 does not correspond to the real maximum brightness , but that the nova was observed at some later point , and by the different passbands used for @xmath47 and for @xmath48 . while in general , the colour differences in the optical range for cvs will amount to less than 1 mag ( e.g. , * ? ? ? * ) , these differences will be enhanced in regions with strong interstellar reddening . a thorough analysis of @xmath49 will thus have to correct for these effects . however , here we only aim at a rough classification of our post - novae within the cv subtypes , and we will do so by also considering other parameters , thus reducing the weight of above uncertainties . further quantities of table [ prop_tab ] that are indicative of @xmath57 are @xmath23 and @xmath51 . the significance of the former and the involved uncertainties have already been presented in section [ results_sec ] . the indicative power of the latter stems from low-@xmath57 , optically thin discs producing stronger emission lines than high-@xmath57 , optically thick ones @xcite . however , in cvs , typically the accretion disc is not the sole contributor to the emission line profile @xcite . because post - novae contain an eruption heated white dwarf , irradiation of the secondary star is likely to produce an additional emission component . furthermore , there is also the possibility that the ejected material still contributes to the line profile especially in ` younger ' , i.e. more recent post - novae . finally , because the continuum emission depends on @xmath58 , so does @xmath51 @xcite , and thus in our interpretation , we will also use the width of the line fwhm , which is indicative of @xmath58 due to the larger projected velocity values @xmath59 sampled in the line profile . el aql and v606 aql are the novae with the largest eruption amplitudes in the present sample . indeed , a comparison with fig.7 in shows that there are only @xmath1110 novae in our sample of pre-1986 eruptions with similar large @xmath49 the idea of this parameter being an indicator for the disc brightness in these two systems fits well with the comparatively strong emission lines . on the other hand , the lines are also broad , and thus additionally , the inclination will have contributed to the high value of @xmath49 . while , for v606 aql , also @xmath23 is in agreement with a comparatively low-@xmath57 disc , this is not the case for el aql . however , we note that the steepest continuum slopes are found for the systems with the largest interstellar reddening . this appears suspicious and suggests a tendency to overcorrect for this effect . v908 oph and v1149 sgr have the smallest @xmath49 and are also spectroscopically consistent with an optically thick disc and thus high @xmath57 in that they present only weak emission lines . in v908 oph , the bluer hydrogen lines are even embedded in absorption troughs . still , also here the line width suggests that the inclination is such that it amplifies that effect . in conclusion , because of the strong dependence of @xmath49 and @xmath19 on @xmath58 , those four post - novae could have very similar intrinsic properties with respect to @xmath57 in spite of their different spectroscopic appearance . the picture is somewhat different for the remaining two systems . v1583 sgr presents narrow , but comparatively strong emission lines with a variable line profile and medium @xmath49 . we interpret this as the combination of a broad but weak disc line profile and a narrow component that likely originates in the secondary star . in this way , all parameters would agree with v1583 sgr having a medium to high @xmath49 and a medium to high inclination . assuming that the narrow component is produced by irradiation , we would also expect that either the white dwarf in this system is still very hot or that the orbital period is comparatively short . because of the ` age ' of the post - nova ( @xmath50 = 87 yr ) , the latter appears more likely . finally , v3964 sgr has broad and weak lines and shows a possible obscuration effect in one spectrum . this points to a high-@xmath57 system seen at high inclination . the recorded medium @xmath49 represents only a lower limit that , if the speculations by @xcite are correct , could be up to 3 mag larger . thus , it would also be in agreement with a high @xmath58 . last , not least , the continuum slope appears rather flat for the suspected high @xmath57 , but again might reflect a high system inclination . in , we have argued that there is a high probability that the cvs detected by us do , in fact , represent the post - novae corresponding to the respective nova eruption and not different cvs that coincidentally have similar coordinates . our first point was that the low space density of cvs ( * ? ? ? * and references therein ) makes it unlikely that more than one cv could be present in a field of a few square arcmin and a limiting magnitude of @xmath6024 mag , although a proper quantitative treatment would have to take into account a detailed model of the galactic structure . as a second point of support to our argumentation , we estimated our detection probability of a specific post - nova system . this is based on the limiting magnitude @xmath61 of our photometry , calculated as the faintest object that is detected in @xmath0 and has s / n = 5 in @xmath5 , corresponding to an uncertainty in the @xmath5 brightness of 0.198 mag . this gives us the maximum amplitude @xmath62 that a nova with a brightness at maximum @xmath47 can have to still be detected in that specific photometry . finally , the fraction of novae with @xmath63 translates into the detection probability . this probability is not fixed , but depends on our current knowledge of the @xmath49 distribution . as of 2016 may 15 , we count 92 pre-1986 post - novae with known @xmath49 , with the shape of the distribution suggesting that it is biased to bright , small-@xmath49 post - novae . our estimate furthermore ignores the possibility that a nova could have very unusual colours and thus is not recognized as such . table [ prop_tab ] shows that four of our objects were detected within 5 arcsec of the reported eruption and thus in its immediate vicinity . we estimate the possibility of a different cv being present within this radius to have a very low probability , making it unnecessary to additionally calculate the detection probability for them . v1149 sgr is slightly farther off , but only v908 oph shows a very large discrepancy , due to the low precision of the coordinates of the nova eruption . for the photometries of v1149 sgr and of v908 oph we find @xmath61 = 23.0 and 20.5 mag , and thus @xmath64 = 15.6 and 11.5 mag , respectively . the much lower brightness and @xmath64 limit for v908 oph is not due a lower quality of the photometry but a result of the strong interstellar reddening causing less than 1 per cent of the objects detected in @xmath10 being also detected in @xmath3 . the corresponding detection probabilities then calculate to 99 and 55 per cent . additionally taking into account the small positional deviation to the reported coordinates , we can thus be confident that v1149 sgr is indeed the post - nova . for v908 oph , the case is less clear , and both parameters leave room for speculation . here , basically the only argument supporting the identification of this system with the post - nova of nova oph 1954 is the absence of other candidates in the field . 1 . we have used @xmath0 photometry and optical spectroscopy to recover the post - novae el aql , v606 aql , v908 oph , v1149 sgr , v1583 sgr and v3964 sgr . we furthermore report on spectroscopy of candidates for the post - novae v1301 aql and v1151 sgr , 2 . historic observations of el aql were collected to build a complete eruption light curve that allows it to be classified as a class p nova @xcite . its eruption amplitude and colours indicate that the previously suspected possibility that el aql is a recurrent nova , is based on source confusion . furthermore , its spectroscopic features suggest that the object belongs to the group of intermediate polars . 3 . comparison of several properties of the novae indicates that the differences in the spectroscopic appearances of el aql , v606 aql , v908 oph and v1149 sgr regarding the aspect of the mass - transfer rate are likely to be an effect of the system inclination and thus do not allow for a conclusive classification . the shape of the h@xmath23 emission line profile in v1583 sgr indicates that it is dominated by a non - disc emission component , likely originating in the irradiated donor star . inspection of the emission line profiles of the individual spectra proposes el aql , v606 aql , v1583 sgr and v3964 sgr as good candidates for time series observations in order to determine the orbital period . we have calculated the probability that the detected objects are , in fact , the post - novae corresponding to the reported eruptions . we have found that the only potentially ambiguous case is that of v908 oph , where the low precision of the reported nova coordinates and the strong interstellar reddening strongly reducing the number of objects detected in @xmath3 leave some room for doubt . still , in the absence of other candidates , the here presented object should be regarded as the post - nova . this paper is based on observations with gemini - south , program ids gs-2015a - q-54 and gs-2015a - q-75 , and eso telescopes , proposal numbers 089.d-0505(a ) and 092.d-0225(a ) . we are indebted to the gemini and eso astronomers who performed the service observations . eso furthermore provided office space and infrastructure for the two week stay of ae in chile , during which large parts of the work were discussed and prepared . many thanks for the kind hospitality ! this research was supported by fondecyt regular grant 1120338 ( ct and nv ) . ae acknowledges support by the spanish plan nacional de astrononoma y astrofsica under grant aya2015 - 66211-c2 - 1 . ct , nv and ifm acknowledge support by the centro de astrofsica de valparaso . we gratefully acknowledge ample use of the simbad data base , operated at cds , strasbourg , france , and of nasa s astrophysics data system bibliographic services . iraf is distributed by the national optical astronomy observatories . the guide star catalogue - ii is a joint project of the space telescope science institute and the osservatorio astronomico di torino . space telescope science institute is operated by the association of universities for research in astronomy , for the national aeronautics and space administration under contract nas5 - 26555 . the participation of the osservatorio astronomico di torino is supported by the italian council for research in astronomy . additional support is provided by european southern observatory , space telescope european coordinating facility , the international gemini project and the european space agency astrophysics division . gemini observatory is operated by the association of universities for research in astronomy , inc . , under a cooperative agreement with the nsf on behalf of the gemini partnership : the national science foundation ( united states ) , the national research council ( canada ) , conicyt ( chile ) , the australian research council ( australia ) , ministrio da cincia , tecnologia e inovao ( brazil ) and ministerio de ciencia , tecnologa e innovacin productiva ( argentina ) . 58 natexlab#1#1 i. et al . , 1998 , the messenger , 94 , 1 n. , strashny g. , 1928 , astron . , 234 , 213 m. , 1929 , astronomische nachrichten , 235 , 427 v. m. , 1954 , , 1461 , 1 f. , 1931 , astronomische nachrichten , 244 , 207 a. j. , 1927 , harvard college observatory bulletin , 851 , 9 j. a. , clayton g. c. , mathis j. s. , 1989 , , 345 , 245 m. , livio m. , 1995 , , 452 , 704 j. , hoffleit d. , 1955 , , 60 , 259 r. a. , webbink r. f. , shara m. m. , ritter h. , kolb u. , duerbeck h. w. , 2005 , journal of astronomical data , 11 , 2 h. w. , 1981 , , 93 , 165 h. w. , 1987 , , 45 , 1 t. e. , 1992 , , 259 , 17p m. , 1927 , harvard college observatory bulletin , 851 , 10 c. , 2001 , cataclysmic variable stars . springer i. m. , jrgensen i. , allington - smith j. r. , davies r. l. , metcalfe n. , murowinski r. g. , crampton d. , 2004 , , 116 , 425 m. m. , cenko s. b. , kulkarni s. r. , ofek e. o. , quimby r. , rau a. , 2011 , , 735 , 94 a. , prialnik d. , shara m. m. , 1988 , , 325 , 828 a. u. , 1983 , , 88 , 439 a. u. , 1992 , , 104 , 340 b. m. et al . , 2008 , , 136 , 735 h. s. , 1933 , annals of harvard college observatory , 84 , 121 i. , stenholm b. , 1976 , , 2997 , 2 i. , stenholm b. , 1977 , information bulletin on variable stars , 1351 , 1 m. w. , 1949 , , 54 , 191 p. et al . , 2015 , , 219 , 26 a. , schaefer b. e. , 2014 , , 788 , 164 j. , 1984 , , 54 , 443 j. et al . , 2013 , , 434 , 1902 e. c. , 1900 , , 12 m. , 2014 , in the x - ray universe 2014 , p. 164 d. , shara m. m. , 1986 , , 311 , 172 f. a. , naylor t. , mukai k. , 1996 , , 281 , 192 r. k. et al . , 2013 , , 554 , a123 g. , nikonoff w. , 1928 , astronomische nachrichten , 233 , 221 e. f. , finkbeiner d. p. , 2011 , , 737 , 103 c. k. , yoss k. m. , 1954 , harv . coll . card , 1260 m. m. , livio m. , moffat a. f. j. , orio m. , 1986 , , 311 , 163 p. b. , 1992 , in astronomical society of the pacific conference series , vol . 25 , astronomical data analysis software and systems i , worrall d. m. , biemesderfer c. , barnes j. , eds . , p. 297 p. b. , 2000 , , 112 , 925 r. j. , schaefer b. e. , henden a. a. , 2010 , , 140 , 34 p. , 1994 , , 108 , 639 c. , ederoclite a. , mennickent r. e. , schmidtobreick l. , vogt n. , 2012 , , 423 , 2476 c. , hanuschik r. , 2001 , lnp vol . 573 : astrotomography , indirect imaging methods in observational astronomy , 573 , 119 c. , schmidtobreick l. , vogt n. , ederoclite a. , 2013 , , 436 , 2412 c. , vogt n. , della valle m. , schmidtobreick l. , ederoclite a. , 2014 , , 442 , 565 c. , vogt n. , schmidtobreick l. , ederoclite a. , 2015 , , 450 , 943 ( paper v ) t. , swierczynski e. , mikolajewski m. , ilkiewicz k. , 2015 , , 576 , a119 r. , 1981 , , 31 , 127 j. , 1927 , , 4 , 106 b. , 1986 , , 222 , 11 b. , 1987 , , 227 , 23 b. , 1995 , cataclysmic variable stars . cambridge astrophysics series , cambridge , new york : cambridge university press j. , 1965 , aavso abstracts , 8 m. , 1927 , astronomische nachrichten , 230 , 421 a. b. , 1940 , publications of lick observatory , 14 , 217 n. , finch c. t. , girard t. m. , henden a. , bartlett j. l. , monet d. g. , zacharias m. i. , 2013 , , 145 , 44 w. , 1928 , astronomische nachrichten , 231 , 383 [ lastpage ]

classical spin liquids @xcite , such as the coulomb phase @xcite in spin ice @xcite and related systems , are examples of phases whose behavior is not captured by the standard landau picture of broken symmetries @xcite . their two defining characteristics are fractionalization , the emergence of excitations not constructed from finite combinations of the elementary degrees of freedom , and topological order , the presence of structure that can only be discerned by observing the system globally @xcite . in the case of the classical spin ices @xcite , a family of magnetic pyrochlore oxides , the consequences of these properties have been of sustained interest , from both theoretical and experimental perspectives . the spins in these materials carry large magnetic moments , and their fractional excitations take the form of magnetic monopoles @xcite , acting as sources for the physical magnetic field . furthermore , transitions between phases where these excitations are confined and deconfined have particularly interesting properties that can not be captured by the landau ginzburg theory of phase transitions @xcite . spin ice also has the unusual feature that the magnetization is a topologically constrained quantity , and so topological order can be probed directly @xcite . in particular , within the low - energy configuration space relevant at low temperatures , any local dynamics conserves the uniform magnetization @xcite . this fact is known to have interesting consequences for critical properties at certain confinement transitions , such as the kasteleyn transition in an applied magnetic field @xcite . in this work , we investigate the interplay between these two aspects of the coulomb spin liquid by studying confinement phase transitions in spin ice across the full range of magnetization . as noted by henley @xcite , one can distinguish three categories of phase transition from the coulomb phase : the first , where the magnetization is zero across the transiton , was studied systematically in ref . @xcite . a second , where the ordered state has saturated polarization , was considered in refs . here we consider the phase structure and transitions more generally , including the third case , where the magnetization is nonzero but unsaturated , and may or may not change across the transition . to do so , we introduce a mean - field theory ( mft ) designed to capture the distinction between confined and deconfined phases . standard mean - field approaches , based on the landau picture of ordered phases , are not well suited to describing confinement transitions . instead , we take inspiration from the well - known mft for the bose hubbard model @xcite , making use of a mapping from spin ice to a quantum model of bosons @xcite . we apply the method to a simplified model of classical spin ice , in order to illustrate the general approach and its physical interpretation . we also briefly discuss extensions to the method that could be used to treat more physically realistic perturbations . in addition , we present arguments for the generality of our results , beyond mean - field theory and our particular choice of hamiltonian . we argue that the phase diagram generically consists of a set of confined phases at low temperature in which the magnetization is , to a very good approximation , fixed . this plateau structure is strongly reminiscent of the lobes present in the phase diagram of the bose hubbard model @xcite , a connection that helps clarify the general phase structure . while the focus here is on classical spin ice in a @xmath0 $ ] field , the distinction between transitions in different flux sectors is more general . for example , spin ice in a field along the @xmath1 $ ] direction also exhibits magnetization plateaux @xcite , while it has been suggested that the zero - field ground state of quantum spin ice may have nonzero magnetization @xcite . in section [ secmodel ] , we introduce the model to be used and briefly outline the general properties of the coulomb phase and confinement transitions . the mean - field method is then described and applied to the model in section [ secmeanfieldtheory ] . we present general arguments for the phase structure and the nature of the phase transitions in section [ secphasestructure ] , before concluding in section [ secconclusions ] . our presentation will be based on the case of classical spin ice in a magnetic field applied along the @xmath0 $ ] crystallographic direction . we first introduce the nearest - neighbor model of spin ice ( nnsi ) that describes the physics of the coulomb spin liquid , before discussing perturbations , such as an applied field , and the resulting ordering transitions . the spin ices @xcite are a family of frustrated magnetic materials with moments on the sites @xmath2 of a pyrochlore lattice , a network of corner - sharing tetrahedra . prominent examples are the `` classical '' spin ices , such as @xmath3 and @xmath4 , which are well described in terms of classical spins @xmath5 . each spin is subject to strong easy - axis anisotropy along the local @xmath6 direction @xmath7 which joins the centres of the two tetrahedra sharing site @xmath2 , as shown in fig . [ figtetrpair ] . ( shown as green arrows ) lie at the sites @xmath2 of the lattice and are constrained to point along the local @xmath6 axes ( dashed lines ) that join the centers of adjacent tetrahedra.,scaledwidth=35.0% ] the anisotropy results from a large ( of order ) crystal - field splitting between states with maximum projection along @xmath7 and all other states in the single - moment hilbert space . this large gap ensures that tunneling processes between the low - energy states are strongly suppressed , rendering the moments effectively classical @xcite . for the low temperatures of interest , it is therefore sufficient to treat the spins as effectively constrained to @xmath8 , where @xmath9 is a classical ising variable . each tetrahedron @xmath10 in the pyrochlore lattice can be labeled according to its orientation @xmath11 , and the lattice structure is such that all neighbors of @xmath10 have orientation @xmath12 . the fixed unit vector @xmath7 for each pyrochlore site @xmath2 , directed along the line joining the centers of its two tetrahedra , can be chosen to point towards the tetrahedron with @xmath13 , and so @xmath14 indicates whether @xmath5 points out of ( @xmath15 ) or into ( @xmath16 ) tetrahedron @xmath10 . with this convention , @xmath17 and hence @xmath18 , for all nearest neighbors @xmath2 and @xmath19 . the minimal hamiltonian that captures the spin - liquid physics of the classical spin ices @xcite contains only interactions between nearest - neighbor pairs @xmath20 , @xmath21 the coupling between moments @xmath5 is ferromagnetic ( @xmath22 ) , incorporating the net effect of the dipolar interactions between nearest neighbors and ( antiferromagnetic ) exchange @xcite . the effective interactions between the ising variables @xmath23 are therefore antiferromagnetic , and hence frustrated . further - neighbor interactions , both dipolar and exchange , are significant in spin ice , but can be treated as a relatively small perturbation to @xmath24 @xcite ; we will return to these in section [ secperturbations ] . because two sites are nearest neighbors if and only if they share a tetrahedron , the interactions can be rewritten as @xmath25 where @xmath26 includes all tetrahedra ( of both orientations ) and @xmath27 is referred to as the ( fictional ) `` charge '' on tetrahedron @xmath10 in terms of a sum over its sites @xmath28 . the energy is therefore minimized by configurations with @xmath29 , which are those where two spins point into and two out of each tetrahedron . configurations that satisfy this condition are said to obey the _ ice rule _ @xcite ; all configurations that do so at every tetrahedron are degenerate minimal - energy states of @xmath24 . tetrahedron configurations with three spins in and one out , or vice versa , have energy @xmath30 above the ice - rule configurations ; they have charge @xmath31 and are hence referred to as `` monopoles '' . ( note that @xmath32 , and so all configurations are globally charge - neutral . ) for @xmath3 , the effective nearest - neighbor coupling is @xmath33 @xcite , corresponding to a monopole cost of @xmath34 . the corresponding boltzmann weight @xmath35 is called the monopole fugacity . while the significance of the term `` monopole '' is here only that they carry effective `` charge '' @xmath36 , it has been argued that monopole excitations in the classical spin ice materials in fact carry true magnetic charge @xcite . for temperature @xmath37 significantly below @xmath38 , the system is effectively constrained by the ice rule , and the ensemble of ice - rule configurations therefore determines its behavior @xcite . the number of such configurations grows exponentially with the total number of spins @xmath39 , and so the entropy density remains nonzero even well below @xmath38 , the temperature scale at which an unfrustrated system would order . in spite of the absence of order , the ensemble exhibits correlations that decay only algebraically with distance . the resulting correlated paramagnet is referred to as a coulomb phase @xcite , and is an example of a classical spin liquid @xcite . the long - wavelength properties of the coulomb phase can be captured by coarse - graining the vector spins @xmath40 to give a continuum field @xmath41 referred to as the effective `` magnetic field '' @xcite . the ice - rule constraint applied to @xmath40 implies that @xmath42 is divergenceless , while monopoles act as sources or sinks , according to an effective gauss law . the resulting predictions for spin spin correlation functions are in good agreement with neutron - scattering experiments in classical spin ice @xcite . the ice - rule configurations also display interesting topological properties @xcite . for our purposes , the most important is that any pair of ice - rule configurations is connected by flipping all spins along one or more loops . for a system with periodic boundary conditions ( which we will assume ) , such a loop can only change the total magnetization @xmath43 if it has nontrivial winding number . for other systems that host coulomb phases @xcite , such as classical dimer models , an analogous quantity , called the `` flux '' , can be defined , but does not necessarily correspond to a quantity that is so directly accessible experimentally . a second important property of the coulomb phase is that monopoles , defects in the ice rule , are deconfined @xcite . to make this statement precise , imagine imposing a pair of monopoles with opposite charges @xmath44 at tetrahedra @xmath10 and @xmath45 in a background that otherwise obeys the ice rule . let @xmath46 denote the set of spin configurations @xmath47 compatible with this charge configuration . their number @xmath48 is a function of the separation @xmath49 that decreases with @xmath50 but reaches a finite limit as @xmath51 . the result is an effective entropic interaction @xmath52 between the charges that is bounded , and hence allows them to be separated to infinity at finite free - energy cost . in fact , since monopoles are charges in the continuum field @xmath42 , coarse - graining predicts that the interaction obeys the coulomb law , @xmath53 , for large @xmath54 @xcite . more generally , consider the partition function @xmath55 where @xmath56 is a perturbation that splits the degeneracy of the ice - rule configurations . this will tend to suppress fluctuations , and can eventually , as the temperature is reduced compared to the scale of the perturbation , drive the system out of the coulomb phase . a transition into a phase where @xmath57 increases without limit as @xmath51 , and so @xmath58 , is referred to as a confinement transition @xcite . ( in terms of the effective continuum field @xmath42 , gauss law implies that an imposed pair of charges must be joined by a fixed quantity of flux . linear confinement , where @xmath59 for large @xmath60 , occurs when the flux forms a narrow tube connecting the monopoles , with finite tension . ) while confinement may occur simultaneously with appearance of conventional symmetry - breaking order @xcite , there is no local order parameter that provides a signature of confinement . the confinement criterion , that @xmath61 approaches zero as @xmath51 in a confined phase and conversely has a finite limit in a deconfined phase , instead resembles the characterization of a conventional symmetry - broken phase in terms of long - range order . this connection can be pursued further by considering the partition function with an unconstrained charge distribution and a position - dependent fugacity @xmath62 , @xmath63 = \sum_{\{\sigma_i\ } } { \mathrm{e}}^{-v / t } \prod_t \zeta_t^{\lvert q_t \rvert}{\,{\text{,}}}\ ] ] in terms of which @xcite @xmath64 \right\rvert_{\zeta_\cdot = 0}{\,{\text { , } } } \label{eqmonopoledistributionfunction}\ ] ] where @xmath65\right\rvert_{\zeta_\cdot = 0}$ ] is the partition function restricted to ice - rule configurations . the relationship between the monopole distribution function @xmath66 and the fugacity @xmath67 expressed by eq . ( [ eqmonopoledistributionfunction ] ) is exactly analogous to the fluctuation dissipation theorem relating , for example , a spin spin correlation function and an applied field . this connection motivates a mean - field theory , analogous to standard weiss theory @xcite , that describes the confinement transition in terms of an effective self - consistent fugacity . it should be noted that the confinement distinction is only sharp in the limit @xmath68 , because any nonzero density of monopoles screens the effective interaction @xmath69 at large separation . there are nonetheless implications for the critical properties away from this limit @xcite . if the phase transition also involves a different type of order ( such as spontaneous symmetry breaking ) , then it may survive for @xmath70 , but with a different universality class @xcite . within the nearest - neighbor model @xmath24 , all configurations that obey the ice rules are degenerate and the system exhibits a deconfining coulomb phase . to drive a confinement transition , one must add perturbations that split the degeneracy of the ice - rule configurations . here we consider a hamiltonian of the form @xmath71 where @xmath72 is the coupling to an applied magnetic field and @xmath73 contains additional interactions described below . coupling to a magnetic field @xmath74 is described by a zeeman term @xmath75 where a factor dependent on the size of the magnetic moments has been included in the definition of @xmath74 . the case that we consider here has @xmath76 where @xmath77 is a unit vector along @xmath0 $ ] , and so @xmath78 . the coupling can therefore be written as @xmath79 where @xmath80 . using the choice described in section [ secnnsi ] , where @xmath7 points towards the tetrahedron with @xmath13 , @xmath81 alternates in sign on successive @xmath82 planes of spins . the component of the magnetization along the field can likewise be expressed as @xmath83 the perturbation @xmath72 alone can drive a confinement transition , referred to as a kasteleyn transition @xcite . for @xmath84 , all spins align with the field ( to the extent allowed by the easy - axis anisotropy ) , an arrangement that satisfies the ice rules . any excitation within the ice - rule states involves flipping a loop of spins , but starting from the fully polarized state , the only available loops span the system in the @xmath0 $ ] direction . such excitations therefore have energy cost proportional to the linear system size @xmath85 and are suppressed in the thermodynamic limit . the result is a strictly saturated magnetization @xmath86 even for nonzero @xmath87 ( but in the limit @xmath88 ) . above a temperature @xmath89 , however , the energy cost is outweighed by the gain in entropy , also @xmath90 , associated with the multitude of available paths for a spanning loop , and a kasteleyn transition occurs from the saturated paramagnet to the coulomb phase . the transition is possible only because the ice rule restricts to loop - like excitations ; away from the limit @xmath91 , the transition is replaced by a crossover @xcite . to study the interplay between the topological constraints on the magnetization , inherent in the ice - rule configurations , and conventional ordering transitions , we also consider an additional interaction @xmath73 that induces spontaneous symmetry breaking . a natural choice would be the further - neighbor coupling present in the spin - ice compounds , due to a combination of further - neighbor exchange @xcite and long - range dipolar interactions . the latter have been shown using monte carlo @xcite to lead to a phase transition into the ordered configuration shown in fig . [ figmdg ] , which we will call the melko den hertog gingras ( mdg ) state @xcite . and ordering wavevector along @xmath0 $ ] ( vertical ) . the spins in each @xmath82 plane are aligned ( to the extent allowed by the local easy axes ) , and between successive planes the magnetization rotates in the right - handed sense by @xmath92 . the perturbation @xmath73 reduces the ferromagnetic coupling of the red bonds by @xmath93 , and hence favors the configuration shown , or the one with all spins inverted.,scaledwidth=85.0% ] the actual perturbation @xmath73 that we will use here is one that is simpler to treat within our mean - field theory but that is expected to lead to an ordered phase of the same type . it is given by @xmath94 where @xmath95 is the set of nearest - neighbor bonds highlighted in fig . [ figmdg ] , which form a set of left - handed screw chains . the effect of @xmath96 is to reduce the strength of the interactions on these bonds and to favor the mdg configuration shown , or the one with all spins inverted . [ figtetrgrid ] shows the energy for each of the six configurations of a single tetrahedron that obey the ice rule , including both the applied field and the perturbation @xmath73 . , defined in eq . ( [ eqhu ] ) , as well as an applied field @xmath97 along @xmath0 $ ] ( upwards ) . the red bonds , as in fig . [ figmdg ] , are those where the ferromagnetic coupling is reduced by @xmath93 . the configurations favored by this perturbation are therefore those in the right column where the spins joined by the red bond are antialigned . ] unlike dipolar interactions , the perturbation @xmath73 reduces the symmetry of the lattice by picking an axis and a chirality . the 12-fold degeneracy of the mdg state is therefore reduced by a factor of @xmath98 , but the symmetry under inversion of all spins remains , and so an ordering transition into this state spontaneously breaks an ising symmetry . this transition ( at @xmath99 ) is in fact a type considered in ref . @xcite , where it was argued to be most likely of first order . as we will show , this simplified model has a phase diagram that illustrates general features emerging from the interplay between deconfinement , ordering , and topological constraints . in section [ secconclusions ] , we will briefly discuss the prospects for including more realistic perturbations in the mean - field theory . some aspects of the phase structure can be elucidated using a mapping between classical spin ice and an effective quantum model of hard - core bosons on a lattice @xcite . this proceeds by identifying the @xmath0 $ ] direction , along which the external field is applied , with the imaginary - time axis in a path - integral representation of the quantum problem . the thermodynamic limit of the original classical system therefore corresponds to the zero - temperature limit in the quantum problem . this mapping has previously been studied in detail for the hamiltonian @xmath100 @xcite . the fully polarized state ( along the @xmath101 direction , say ) is identified with the vacuum ; flipped spins relative to this configuration form system - spanning loops , which are viewed as the world lines of bosons . the result is an effective model of hard - core bosons on a lattice with density @xmath102 and hence chemical potential @xmath103 . in general , one expects all other local terms compatible with symmetry , such as hopping between nearby sites and further - neighbor interactions , with parameters related in nontrivial ways to those of the spin model @xcite . as @xmath104 is increased , the system passes from the vacuum through a quantum phase transition @xcite to a superfluid , which corresponds to the coulomb phase of spin ice . ( off - diagonal long - range order in the superfluid can be associated with deconfinement in the coulomb phase @xcite . ) for sufficiently large positive @xmath104 , the system is driven into a fully occupied mott insulator , with @xmath105 , corresponding to a saturated paramagnet with @xmath106 in an inverted applied field . critical properties at the phase transitions in spin ice are , _ mutatis mutandis _ , equivalent to those in the quantum model @xcite . the presence of interactions between bosons on different lattice sites also makes possible mott insulators with fractional filling . these can be identified with confining phases of spin ice ( because they lack superfluid order ) in which the magnetization is fixed to a value less than @xmath107 . the mdg phase , for example , is an ordered phase with @xmath108 , and corresponds to a mott insulator at half filling . in general , more exotic phases such as supersolids are also possible in the effective quantum model . a supersolid would correspond to a coulomb phase with broken spatial symmetry ; such phases have been proposed in spin ice @xcite , but are not captured by the mean - field theory used here . the phase structure of @xmath109 in eq . ( [ eqhamiltonian ] ) is expected to contain : the coulomb phase , characterized by deconfinement of monopoles ; the saturated paramagnet , in which fluctuations are suppressed by the ice - rule constraints ; and the mdg phase , with conventional order that breaks spin - inversion and spatial symmetries . a mean - field theory that describes all three must therefore capture the confinement deconfinement distinction , unlike mean - field approaches based on finite clusters @xcite , and also respect the spatial structure , unlike bethe - lattice calculations @xcite . here we apply an approach that is analogous to the standard mean - field theory for the bose hubbard model @xcite , but is applied directly to the classical partition function , written in terms of a transfer matrix . ( for other applications of the variational method to the transfer matrix , see refs . we start by partitioning the lattice into chains that span the system , which are the minimal units that can obey the ice rule . an effective model is then defined for each chain , with the coupling between chains treated self - consistently . in standard mean - field theory for a symmetry - breaking transition @xcite , coupling between lattice sites is replaced by an effective field related to the order parameter . here , in the absence of a local order parameter , we capture the confinement deconfinement transition by introducing an effective self - consistent fugacity @xmath110 for monopoles . since , according to eq . ( [ eqmonopoledistributionfunction ] ) , the monopole distribution function can be interpreted as the correlation function corresponding to a monopole fugacity , a nonzero @xmath110 implies deconfinement . this decoupling is closely analogous to that used for the bose hubbard model , where superfluid order is captured by introducing a self - consistent source term for bosons . following the mapping between this quantum problem and the classical statistical mechanics of spin ice , we apply the method at the level of the transfer matrix . it should be emphasized that this differs from the standard mean - field approach for classical statistical systems , where one chooses a trial distribution to minimize the free energy . ( the latter method can not be used , because a trial configuration with nonzero fugacity does not enforce the ice rule and therefore has infinite energy . ) while we treat deconfinement using an effective mean - field fugacity @xmath110 , we set the true monopole fugacity @xmath35 to zero . nonzero @xmath67 could straightforwardly be included in the approximation and is analogous to standard weiss mean - field theory in the presence of an applied magnetic field @xcite . we choose to decompose the pyrochlore lattice into right - handed screw chains with axes parallel to the @xmath0 $ ] direction , as illustrated in fig . [ figmdgchains ] . $ ] crystallographic direction . in the mdg state , the chains are ordered in a checkerboard pattern , with spins along each chain aligned and neighboring chains antialigned.,scaledwidth=85.0% ] for our purposes , this choice has two advantages : first , the chains span the system in the imaginary - time direction under the mapping to a bosonic model , and so the connection to the bose hubbard mean - field theory is transparent . second , the ordered states that we aim to describe have natural interpretations in terms of chain order : in the mdg state , shown in fig . [ figmdgchains ] , the chains order in a checkerboard pattern , while in the fully polarized paramagnet , all chains are aligned . other orderings , such as the partially polarized state observed by lin and kao @xcite , can be described in terms of ordered configurations of the same chains , while the general approach can likely be adapted for ordered states that are not captured by this decomposition of the lattice . ( for example , mcclarty et al . @xcite consider ordered states described in terms of linear chains of spins . ) the division of the lattice into screw chains has the property that each chain @xmath111 has @xmath112 neighbors , which it meets in turn at tetrahedra in successive @xmath82 layers . the set of tetrahedra at a given layer , at vertical position @xmath113 , amounts to a partition of the chains into pairs , which we denote @xmath114 . we also define @xmath115 as the chain that meets chain @xmath111 in a tetrahedron at layer @xmath113 . the pyrochlore lattice is symmetric under translations by @xmath116 layers in the @xmath0 $ ] direction @xcite , and so @xmath117 and @xmath118 . in the quantum mapping , the vertical direction is treated as imaginary time . applying the mapping at a microscopic level would therefore lead to a time - evolution operator that couples different pairs of lattice sites at successive time steps . this `` stroboscopic '' @xcite implementation of the hopping in the effective quantum hamiltonian has been shown to be irrelevant for the critical properties @xcite , but must be handled with some care when implementing the mean - field theory . the partition function @xmath119 can be expressed exactly in terms of a set of operators acting on degrees of freedom associated with the chains . to each chain @xmath111 is associated a pair of basis vectors , @xmath120 and @xmath121 , representing the orientation , up or down ( i.e. , @xmath122 ) , of the spin at a given vertical position @xmath113 along the chain . denoting by @xmath123 the space spanned by these vectors , the full hilbert space for a layer of the system is @xmath124 . ( note that there is a single two - dimensional hilbert space for each chain , and so the basis vectors are labeled only by their chain @xmath111 , and not by their position @xmath113 along the chain . ) the partition function can then be written as @xmath125 where @xmath85 is the system size ( number of layers ) in the @xmath113 direction and @xmath126 is the transfer operator for the layer of tetrahedra at vertical position @xmath113 . ( we assume periodic boundary conditions in the @xmath113 direction . ) the tetrahedron transfer operator @xmath127 , acting on @xmath128 , accounts for the possible configurations of the chains @xmath111 and @xmath129 on entering and leaving this tetrahedron . using the symmetry of the pyrochlore lattice under translations by @xmath116 layers in the @xmath113 direction , the product in eq . ( [ eqpartitionfunction ] ) can be rewritten as @xmath130^{l_z/\lambda_z}{\,{\text{,}}}\quad\text{where}\quad{\mathcal{t}}^{(\lambda_z)}=\prod_{z=\lambda_z}^{1}{\mathcal{t}}_z { \,{\text{.}}}\ ] ] in the thermodynamic limit , the partition function is therefore given by @xmath131^{l_z/\lambda_z}$ ] , where @xmath132 is the spectral radius ( largest absolute eigenvalue ) . it should be noted that , even if @xmath133 is hermitian , in general @xmath134\neq 0 $ ] for @xmath135 , and so @xmath136 is not hermitian . the approximation we propose is to replace the system with an effective model of decoupled chains , whose parameters are determined self - consistently . the partition function for chain @xmath111 is taken as @xmath137 where the effective transfer operator for chain @xmath111 is @xmath138 an operator on @xmath123 . the normalized vector @xmath139 , which is interpreted as an ansatz for the configuration of chain @xmath111 immediately above layer @xmath113 , is determined self - consistently as the thermal state of the one - dimensional system defined by eq . ( [ eqpartitionfunctionmf ] ) . assuming a self - consistent solution exists with at least periodicity under translation in @xmath113 by @xmath140 , it is given by the normalized ( right ) eigenvector with largest absolute eigenvalue of @xmath141 for the types of phases that we aim to describe , it will in fact be sufficient to consider @xmath142 uniform along each chain . this leads to the simplification that @xmath143 which is therefore hermitian . for such a solution to exist , @xmath144 must be a simultaneous eigenvector of @xmath145 for all @xmath113 . for the case with an applied field @xmath74 and a chain - favoring interaction @xmath93 , the tetrahedron transfer operator can be written explicitly as @xmath146 & { \mathrm{e}}^{\frac{2}{\sqrt{3}}h / t } \ket{\uparrow}_c \ket{\uparrow}_{c'}\bra{\uparrow}_{c}\bra{\uparrow}_{c ' } + { \mathrm{e}}^{-\frac{2}{\sqrt{3}}h / t}\ket{\downarrow}_c \ket{\downarrow}_{c'}\bra{\downarrow}_c \bra{\downarrow}_{c'}\\ & { } + { \mathrm{e}}^{2u / t}\ket{\uparrow}_c\ket{\downarrow}_{c'}\bra{\uparrow}_c \bra{\downarrow}_{c ' } + { \mathrm{e}}^{2u / t}\ket{\downarrow}_c\ket{\uparrow}_{c'}\bra{\downarrow}_c \bra{\uparrow}_{c'}\\ & { } + \ket{\uparrow}_c\ket{\downarrow}_{c'}\bra{\downarrow}_c \bra{\uparrow}_{c ' } + \ket{\downarrow}_c\ket{\uparrow}_{c'}\bra{\uparrow}_c \bra{\downarrow}_{c'}{\,{\text{. } } } \end{aligned}\ ] ] each term corresponds to one of the six ice - rules configurations of the tetrahedron , illustrated in fig . [ figtetrgrid ] , and is associated with a boltzmann weight implementing the terms @xmath72 and @xmath73 in the hamiltonian . ( terms breaking the ice rule could also be included ; we are here treating such configurations as having zero boltzmann weight . ) the first four terms correspond to cases where the two chains maintain their configurations after passing through the tetrahedron i.e . , the two spins on each chain are aligned while in the last two terms the chains exchange orientations . taking the inner product in @xmath147 gives @xmath148 & \left[\frac{1}{2}(1+m{^{\text{mf}}}_{c'}){\mathrm{e}}^{\frac{2}{\sqrt{3}}h / t } + \frac{1}{2}(1-m{^{\text{mf}}}_{c'}){\mathrm{e}}^{2u / t}\right]\ket{\uparrow}_c\bra{\uparrow}_c \\ & { } + \left [ \frac{1}{2}(1-m{^{\text{mf}}}_{c'}){\mathrm{e}}^{-\frac{2}{\sqrt{3}}h / t } + \frac{1}{2}(1+m{^{\text{mf}}}_{c'}){\mathrm{e}}^{2u / t}\right]\ket{\downarrow}_c\bra{\downarrow}_c\\ & { } + \zeta{^{\text{mf}}}_{c'}\ket{\uparrow}_c\bra{\downarrow}_c + \left({\zeta{^{\text{mf}}}_{c'}}\right)^*\ket{\downarrow}_c\bra{\uparrow}_c{\,{\text { , } } } \end{aligned } \label{eqteff}\ ] ] where the expectation values @xmath149 are mean fields describing the magnetization and density of monopoles on chain @xmath150 . the effective model for chain @xmath111 , described by the transfer operator @xmath151 , involves an applied field that is a function of @xmath152 and an effective monopole fugacity @xmath153 . this is the crux of the approximation : the exchange of orientation between chains , described by the last two terms of eq . ( [ eqtetrahedrontransferoperator ] ) , has been replaced by terms where a single chain flips orientation , thus creating or destroying a monopole . the criterion for deconfinement , that a distant pair of monopoles can be inserted with finite free - energy cost , is trivially satisfied whenever @xmath154 . ( in the analogous mean - field theory for the bose hubbard model @xcite , the expectation value of the bosonic annihilation operator is used as a mean field . ) in general , the solution of the self - consistent approximation requires finding simultaneous eigenvectors of @xmath151 , given in eq . ( [ eqteff ] ) , for all the values of @xmath113 . for the types of ordered phases expected in our model of spin ice , however , some further simplifications are possible . in particular , to describe the mdg and saturated phases , it is sufficient to divide the chains into two `` chain sublattices '' , which we label @xmath155 and @xmath156 , with one chain of each type meeting in every tetrahedron , as illustrated in fig . [ figmdgchains ] . the operator @xmath151 is therefore independent of @xmath113 , and it is sufficient to find @xmath157 and @xmath158 such that @xmath144 is the eigenvector of @xmath159 with largest eigenvalue @xcite , for both @xmath160 , @xmath161 and vice versa . because @xmath162 is hermitian , this is equivalent to maximizing @xmath163 with respect to normalized @xmath157 and @xmath158 . as an aside , we note that this amounts to maximizing the matrix element of the layer transfer operator @xmath133 in the subspace of factorizable vectors @xmath164 . in these terms , it is clear how this is related to the mean - field theory for the bose hubbard model , where one minimizes the matrix element of the hamiltonian @xmath109 . ( recall the standard connection between classical and quantum statistical mechanics , where @xmath165 is the evolution operator in imaginary time @xmath166 . ) in both cases , the operator is hermitian and so the matrix element is bounded by its extremal eigenvalue , but only for the bh model is the hamiltonian constant in time and the matrix element bounded by the true ground - state energy . in the present case , the product of extremal matrix elements @xmath167 provides an approximation to the exact partition function , but this approximation is not variational . to maximize @xmath168 , we express both vectors as @xmath169 where @xmath170 and @xmath171 , in terms of which @xmath172 the distinct maxima of this expression correspond to phases of the mean - field theory , shown in fig . [ figphasediagram ] . , the melko den hertog gingras ( mdg ) phase @xcite with @xmath173 , and a saturated paramagnet with @xmath174 . at higher - temperatures is a coulomb spin liquid whose magnetization is a smooth function of the applied field @xmath97 . ( dotted lines are contours of constant magnetization . ) ] the phase boundaries can be found by expanding in small deviations from the ordered phases and identifying the points where they cease to be stable maxima . ( we have verified numerically that the ordered states maximize the matrix element globally if and only if they are local maxima . ) at low temperature , there are three confined phases with different , fixed , values of the magnetization @xmath175 for @xmath176 , the maximum of eq . ( [ eqtv1v2 ] ) has @xmath177 ( @xmath178 ) or @xmath179 ( @xmath180 ) for all @xmath111 , representing a fully polarized paramagnet with @xmath181 . for @xmath182 , it has @xmath177 for @xmath160 and @xmath179 for @xmath183 , or vice versa . the two chain sublattices therefore have opposite orientations , @xmath184 , and the total magnetization vanishes . comparison with fig . [ figmdg ] confirms that this corresponds to the mdg phase . at higher temperature , the maximum has @xmath185 for all @xmath111 and @xmath186 , corresponding to a paramagnet with continuously varying magnetization . this last phase is deconfined , having @xmath187 , and is therefore identified as the coulomb phase . [ figphasediagram ] also shows the magnetization @xmath188 evaluated at the maximum of @xmath168 . the resulting phase diagram consists of a set of low - temperature confined phases with fixed magnetization , surrounded by the coulomb phase in which the magnetization varies smoothly with the parameters . the mdg phase consists of a lobe , whose width , i.e. , the range of fields for which it is stable , decreases with increasing temperature . as a benchmark for the method , we note that our model reduces for @xmath189 to that considered by jaubert et al . @xcite ( in the limit of zero monopole fugacity ) . in this case , one can restrict to confined phases with all chains equivalent ( i.e. , @xmath190 ) , and it is straightforward to show that the magnetization is given by @xmath191 where @xmath192 . our method therefore agrees , in this case , with the results of the bethe - lattice calculation @xcite . we have shown that applying mean - field theory to the model defined by @xmath109 in eq . ( [ eqhamiltonian ] ) produces a phase diagram containing the coulomb phase @xcite as well as a fully - polarized paramagnet @xcite and the mdg phase @xcite . in this section , we argue that the general features in the phase diagram of fig . [ figphasediagram ] , and in particular the lobe structure , can be understood by considering the interplay of confinement and flux in the coulomb phase . we first consider the limit of vanishing monopole density , approximately valid at temperatures well below @xmath38 , before discussing the qualitative effect of including monopoles . the crucial observation underlying the phase structure is that confinement implies suppression of fluctuations in the magnetization @xmath193 , and hence vanishing uniform magnetic susceptibility @xmath194 . by contrast , phases where monopoles are deconfined , such as the coulomb phase , are magnetizable ( i.e. , have nonzero susceptibility @xmath194 ) . this relationship between confinement and vanishing susceptibility ( or , more generally , flux variance ) occurs in any local model with a confining phase @xcite . the lobe structure of the phase diagram follows immediately from this observation . starting within a confining phase and changing the applied field will eventually tip the balance in favor of the coulomb phase , whose free energy is quadratic in the field . the higher entropy in the coulomb phase implies that the range of applied fields for which a given ordered state is favored narrows as the temperature increases . the confined phases therefore comprise lobes with fixed magnetization , each centered around favorable values of the applied field . ( the fully polarized phases are exceptions , extending to arbitrarily large @xmath195 as a result of their saturated magnetization . ) this qualitative structure is confirmed in the mean - field phase diagram , fig . [ figphasediagram ] . with more realistic interactions , we expect more lobes with other commensurate values of magnetization to appear in the gaps between the lobes shown , as indeed seen in the mc results of ref . @xcite . the general phase structure can also be understood through the mapping to a quantum model of bosons , described in section [ secbosonmapping ] and ref . @xcite . for an extended hard - core bose hubbard model in the grand canonical ensemble , the phase diagram contains a set of mott - insulator lobes at small hopping , consisting of different possible ordered states @xcite . returning to spin ice , each of these corresponds to a confined phase with order in the spin degrees of freedom and fixed magnetization . the susceptibility strictly vanishes in the ordered phases only in the thermodynamic limit at zero monopole fugacity @xmath35 . with finite monopole cost , confinement of monopoles is no longer precisely defined and the susceptibility becomes nonzero , though remains exponentially small at low temperature . the plateaux are therefore rounded for nonzero @xmath67 . those phases that have conventional order , such as the mdg phase , remain separated from the paramagnet by a symmetry - breaking transition . certain properties of the phase transitions can also be determined on general grounds , following similar arguments to those for the bose hubbard model @xcite , but with density replaced by magnetization . we first note that phase transitions can be classified by whether the magnetization changes across the transition . the magnetization necessarily changes in the case where the ordered phase has saturated magnetization , but it can also change across a transition into a magnetically ordered phase . at the tip of the lobe , the magnetization is identical in the two phases ( ordered within and coulomb outside ) , and so this is the only point at which the magnetization stays the same across the transition . the argument is that , inside the ordered phase but sufficiently close to the tip , an arbitrarily small change in the applied field pushes the system into the coulomb phase , regardless of its sign . if the two phases had different magnetizations at this point , their linear coefficients in the free energy would be different and this could not be the case . ( this argument is essentially identical to the corresponding one for the bose hubbard phase diagram @xcite . ) for any continuous transitions with fixed magnetization , critical properties can be calculated using the approach of ref . @xcite , but with a modified projective symmetry group due to the nonzero magnetization . while mdg found a strongly first - order transition @xcite and the transition into the partially polarized phase was seen to be of first order in ref . @xcite , it remains possible that transitions in the presence of different perturbations could be continuous . ( the possibility of unconventional transitions between a superfluid and fractional mott insulators has been discussed in the extended bose hubbard model @xcite . ) when the magnetization changes across the transition , i.e. , everywhere except at the tip , the transition is in the same universality class as the kasteleyn transition @xcite . it should be noted , however , that only the transitions into the saturated phases , with @xmath181 , are true kasteleyn transitions , which are distinguished by the complete absence of fluctuations on the ordered side . we have presented a mean - field theory designed to study confinement transitions , based on the analogy between the confinement criterion and long - range order in conventional ordering transitions . the method has been applied to a simplified model of classical spin ice and shown to produce a phase diagram containing , besides the coulomb spin liquid @xcite , a set of magnetization plateaux . these include the saturated paramagnet @xcite as well as the mdg phase expected to occur in dipolar spin ice at low temperature @xcite . we have argued that the general properties of the phase diagram follow from the interplay between confinement and magnetization that characterizes the ice - rule states and pointed out the analogy with the phase diagram of the extended bose hubbard model @xcite . the simplified model that was studied here reproduces the mdg phase and is particularly amenable to the mean - field approach . it would be desirable to extend the approach to allow for more realistic further - neighbor interactions between spins , particularly the dipolar interactions that are known to be significant in the classical spin - ice materials . further - neighbor interactions can also stabilize the partially magnetized phase observed using mc in ref . ( this phase can be described in terms of the chains shown in fig . [ figmdgchains ] , but distinguishes four chain sublattices , rather than two . ) a transfer operator written in the form of eq . ( [ eqtetrahedrontransferoperator ] ) has the limitation , however , that it can represent interactions only within a single tetrahedron . ( one can not even define the layer transfer operator for general interactions . ) a potential route to include further - neighbor interactions is through an additional mean - field decoupling , replacing a pairwise interaction @xmath196 with terms such as @xmath197 , representing a mean field acting on site @xmath2 . this extension , and others such as including nonzero monopole fugacity @xmath67 , are left to future work . k. matsuhira , z. hiroi , t. tayama , s. takagi , and t. sakakibara , _ a new macroscopically degenerate ground state in the spin ice compound @xmath3 under a magnetic field _ , j. phys . : condens . matter * 14 * , l559 ( 2002 ) . p. henelius , t. lin , m. enjalran , z. hao , j. g. rau , j. altosaar , f. flicker , t. yavorskii , and m. j. p. gingras , _ refrustration and competing orders in the prototypical @xmath3 spin ice material _ , , 024402 ( 2016 ) . this nomenclature follows lin and kao @xcite . mcclarty et al . @xcite call the same state a `` cubic antiferromagnet '' , while henelius et al . @xcite call it the `` single - chain state '' . in ref . @xcite , it was called a `` spin spiral '' . mdg themselves @xcite refer to it by its ordering wavevector , @xmath199 , where @xmath200 is the side length of the cubic unit cell . one can imagine creating a monopole antimonopole pair , winding one of the pair around the periodic boundaries , and then recombining them @xcite . the result of this charge transport is to increase the flux through the system ; since the separation required diverges in the thermodynamic limit , it is possible with finite energy cost only in a deconfined phase .

recent developments in the physics of ultra - cold gases have led to the discovery of bose - einstein condensation ( bec ) in trapped clouds of alkali atoms @xcite and stimulated a tremendous boost in theoretical studies of weakly interacting bose gases . as in previous years , these studies rely on the binary approximation for the interparticle interaction . the latter is characterized by the 2-body scattering length @xmath0 , which assumes the presence of a small gaseous parameter @xmath1 ( n is the gas density ) . especially intensive are the attempts to reach beyond the ordinary mean field approach and to develop a theory which can properly describe the behavior of finite temperature elementary excitations of a trapped bose - condensed gas and in particular , explain the jila @xcite and mit @xcite experiments on energy shifts and damping rates of the excitations . the commonly used mean field theory ( for @xmath2 ) is based on the bogolyubov quasiparticle approach developed originally for a spatially homogeneous bose - condensed gas at @xmath3 @xcite and employed by lee and yang @xcite ( see also @xcite ) at finite temperatures . the generalization of the bogolyubov method for spatially inhomogeneous systems has been described by de gennes @xcite . in the case of a bose - condensed gas it should be completed by the equation for the wavefunction of the spatially inhomogeneous condensate , derived by pitaevskii @xcite and gross @xcite . for spatially homogeneous gases the theory beyond the mean field approach was also developed . beliaev @xcite constructed the zero - temperature diagram technique which allows one to find corrections to the energies of bogolyubov excitations , proportional to @xmath4 , where @xmath5 is the condensate density . the corrections are provided by the interaction between the excitations ( in particular , through the condensate ) and contain both real ( energy shift ) and imaginary ( damping rate ) parts . at @xmath6 the latter originates from spontaneous decay of a given excitation ( @xmath7 ) to two other excitations ( @xmath8 and @xmath9 ) , with smaller energies and momenta : @xmath10 . a universal expression for the chemical potential in terms of the self - energy functions has been found by pines and hugenholtz @xcite . it should be emphasized that the corrections proportional to @xmath11 already depend on the contribution of 3-body interactions and , hence , can not be obtained within the binary approximation . the beliaev approach was employed by popov @xcite at finite temperatures . in this case the corrections to the beliaev self - energies contain infra - red singularities , i.e. they tend to infinity for momenta @xmath12 . this prompted popov to make a renormalization of the theory , which links the microscopic approach with phenomenological landau hydrodynamics @xcite . the popov theory eliminates the infra - red singularities and allows one to describe the behavior of low - energy excitations ( phonons ) at temperatures much smaller than the mean field interparticle interaction @xmath13 ( @xmath14 , with @xmath15 being the atom mass ) . the damping of phonons in this temperature range is determined by the beliaev damping processes and has also been calculated by hohenberg and martin @xcite . a simplified approach within the dielectric formalism was used by szepfalusy and kondor @xcite for calculating the damping rates of excitations in the phonon branch of the spectrum . they found that at temperatures @xmath16 the damping rate of a given excitation ( @xmath7 ) originates from the scattering of thermal excitations ( @xmath8 and @xmath9 ) on the excitation @xmath7 through the processes @xmath17 . since the characteristic energies of the thermal excitations @xmath8 , @xmath9 turn out to be much larger than the energy of the excitation @xmath7 , this damping channel can be treated as landau damping . it should be noted that the damping rates can be simply found by considering the interaction between the excitations as a small perturbation and using fermi s golden rule . this allows one to properly take into account the bogolyubov nature of the thermal excitations . the damping rates of phonons in a spatially homogeneous bose - condensed gas , in particular for the szepfalusy - kondor mechanism , have been calculated in the recent contributions @xcite . in order to reach beyond the mean field theory at @xmath18 one should further develop the popov approach . one can also proceed along the lines of the beliaev theory , since any physical quantity should be determined by combinations of the beliaev self - energies , which do not contain the infrared singularities . in this paper we choose the latter way and construct the perturbation theory for a bose - condensed gas , which allows us to find the next to leading order terms ( the terms proportional to @xmath4 ) in the energy spectrum of the elementary excitations . as in @xcite , we consider the excitations in the so - called collisionless regime , where their de broglie wavelength is much smaller than the mean free path of the thermal excitations . we start with the case of a spatially homogeneous bose - condensed gas and find temperature - dependent energy shifts and damping rates for bogolyubov excitations of any energy . at temperatures @xmath16 the small parameter of the theory proves to be @xmath19 in contrast to @xmath20 for @xmath6 . the appearance of the extra factor ( @xmath21 ) originates from the bose occupation numbers of thermal excitations with energies of order @xmath13 , which are the most important in the perturbation theory . as shown below , the damping of excitations with energies @xmath22 is determined by both the szepfalusy - kondor ( @xmath23 ) and beliaev ( @xmath24 ) processes , and can no longer be treated as landau damping . the theory is generalized for the case of excitations in a spatially inhomogeneous ( trapped ) bose - condensed gas . a new ingredient here is related to the inhomogeneous density profile of the condensate and the discrete structure of the excitation spectrum . we develop the method of calculating the self - energy functions and derive the equations for finding the wavefunctions and energies of the excitations ( generalized bogolyubov - de gennes equations ) . the use of the theory is demonstrated by two examples . the first one concerns quasiclassical low - energy excitations of a trapped bose - condensed gas in the thomas - fermi regime . the term `` low - energy '' assumes that the excitation energy @xmath25 is much smaller than the mean field interparticle interaction @xmath26 ( @xmath27 is the maximum condensate density ) , and the quasiclassical character of the excitations requires the condition @xmath28 , where @xmath29 is the characteristic trap frequency . we consider anisotropic harmonic traps , where the discrete structure of the excitation spectrum is not important ( see below and in @xcite ) . on the contrary , the inhomogeneity of the condensate density profile has a crucial consequence for the damping rates and energy shifts of quasiclassical low - energy excitations . the most important turns out to be the boundary region of the condensate , where @xmath30 @xcite . therefore , the result for the damping rates and energy shifts is completely different from that in spatially homogeneous gases . finally , we analyze the frequency shifts and damping of axially propagating sound waves in cylindrical bose condensates . as found , the nature of damping is similar to that in the case of phonons in spatially homogeneous bose condensates . we show that the attenuation of axially propagating sound wave packets in the recent mit experiment @xcite can be well explained as a consequence of this damping . we consider a weakly interacting bose - condensed gas confined in an external potential @xmath31 . the grand canonical hamiltonian of the gas can be written as @xmath32 , where ( hereinafter @xmath33 ) @xmath34 and the term @xmath35 assumes a point interaction between atoms . the field operator of atoms @xmath36 can be represented as the sum of the above - condensate part @xmath37 and the condensate wavefunction @xmath38 which is a _ c_-number . as the interparticle interaction @xmath39 contains both terms conserving the number of above - condensate particles and terms transferring two above - condensate particles to the condensate ( or two condensate particles to the above - condensate part ) , the diagram technique should include both the normal green function @xmath40 and the anomalous green function @xmath41 ( see , e.g. @xcite ) . the sums of the contributions of all irreducible diagrams will be represented by the normal ( @xmath42 ) and anomalous ( @xmath43 ) self - energies ( see fig.1 and fig.2 ) . the former corresponds to the processes conserving the number of above - condensate particles , and the latter describes absorption ( or emission ) of two particles to ( out of ) the condensate . the green function and self - energy operators satisfy beliaev - dyson equations @xcite @xmath44 where the green functions @xmath45 and @xmath46 describe forward and backward propagation of a particle characterized by the hamiltonian @xmath47 . we confine ourselves to the case of repulsive interaction between the atoms ( @xmath48 ) . to develop the finite temperature perturbation theory for calculating dynamic properties and finding the excitation spectrum of a weakly interacting bose - condensed gas we will use the non - equilibrium generalization @xcite of the matsubara diagram technique . in eqs . ( [ boperg]),([boperf ] ) we perform an analytical continuation of the matsubara frequencies @xmath49 ( @xmath50 is an integer number ) to the upper half - plane , which corresponds to the replacement @xmath51 . then , multiplying both sides of eqs . ( [ boperg ] ) and ( [ boperf ] ) by @xmath52 and @xmath53 , respectively , we arrive at the system of equations in the frequency - coordinate representation : @xmath54 { \cal g}% ( \varepsilon ; { \bf r},{\bf r}^{\prime } ) + \sigma _ { a}(\varepsilon ) { \cal f}% ( \varepsilon ; { \bf r},{\bf r}^{\prime } ) + \delta ( { \bf r}-{\bf r}^{\prime } ) , \label{bcoordg } \\ -\varepsilon { \cal f}(\varepsilon ; { \bf r},{\bf r}^{\prime } ) & = & \left [ -% \frac{\delta } { 2}+v({\bf r})-\mu + \sigma ^{+}(\varepsilon ) \right ] { \cal f}% ( \varepsilon ; { \bf r},{\bf r}^{\prime } ) + \sigma _ { a}(\varepsilon ) { \cal g}% ( \varepsilon ; { \bf r},{\bf r}^{\prime } ) . \label{bcoordf}\end{aligned}\ ] ] here the action of the integral self - energy operators on the green functions is written in a compact form @xmath55 ( and similar relations for the other combinations ) . ( [ bcoordg]),([bcoordf ] ) should be completed by a generalized gross - pitaevskii equation for the condensate wavefunction : @xmath56 \psi _ { 0}({\bf r})=0 \label{gpeq}\ ] ] and by the normalization condition @xmath57 where @xmath58 is the condensate density , @xmath59 is the density of above - condensate particles , and @xmath60 the total number of particles in the gas . ( [ bcoordg]),([bcoordf ] ) can be solved by using the bogolyubov transformation for the green functions : @xmath61 where the index @xmath7 stands for the set of quantum numbers , and the functions @xmath62 , @xmath63 satisfy generalized bogolyubov - de gennes equations @xmath64 \!\!u_{\nu}({\bf r } ) \!-\!\!\sigma _ { a}(\!\varepsilon_{\nu}\!)v_{\nu}({\bf r } ) , \!\!\ ! \label{bdggenu } \\ \!-\varepsilon_{\nu}v_{\nu}({\bf r})\!&\!=\!&\!\left[\!-\frac{\delta } { 2}\ ! + \!v({\bf r})\!-\!\mu\!+\!\!\sigma ^{+}\!(\!\varepsilon_{\nu}\!)\!\right]\ ! \!v_{\nu}({\bf r})\!-\ ! \!\sigma_{a}(\!\varepsilon_{\nu}\!)u_{\nu}({\bf r}).\!\!\ ! \label{bdggenv}\end{aligned}\ ] ] in the bogolyubov - de gennes approach only the terms bilinear in @xmath65 operators are retained in the interaction hamiltonian @xmath39 , which assumes that the condensate density is much larger than the density of above - condensate particles . then , the self - energy operators take the form @xmath66 the result of their action on the condensate wavefunction @xmath67 and the functions @xmath68 , @xmath69 is reduced to @xmath70 and similar relations for the other combinations . then , eq.([gpeq ] ) becomes the ordinary gross - pitaevskii equation @xmath71 \psi _ { 0}({\bf r})=0 , \label{gpeq0}\ ] ] and eqs . ( [ bcoordg]),([bcoordf ] ) are transformed to the ordinary bogolyubov - de gennes equations @xmath72 u_{\nu}({\bf r})-n_{0}({\bf r } ) \widetilde{u}v_{\nu}({\bf r } ) , \label{bdgu } \\ -\varepsilon_{\nu}v_{\nu}({\bf r } ) & = & \left [ -\frac{\delta } { 2}+v({\bf r})- \mu+2n_{0}({\bf r})\widetilde{u}\right ] v_{\nu}({\bf r})-n_{0}({\bf r } ) \widetilde{u}u_{\nu}({\bf r } ) . \label{bdgv}\end{aligned}\ ] ] taking into account eq.([gpeq0 ] ) , in terms of the functions @xmath73 these equations can be rewritten as @xmath74 for a trapped bose - condensed gas in the thomas - fermi regime , where @xmath75 is much larger then the spacing between the trap levels , the kinetic energy term in eq . ( [ gpeq0 ] ) can be omitted and one has @xcite @xmath76 if the argument of the square root is positive and zero otherwise . for the low - energy excitations ( @xmath77 ) of thomas - fermi condensates eqs . ( [ fplus0]),([fmin0 ] ) can be reduced to hydrodynamic equations obtained by stringari @xcite and solved in the case of spherically symmetric harmonic potential @xmath78 and for some excitations in a cylindrically symmetric potential . an analytical method of solving eqs . ( [ fplus0]),([fmin0 ] ) ( or the corresponding hydrodynamic equations ) for the low - energy excitations of thomas - fermi condensates in an anisotropic harmonic potential @xmath31 has been developed in @xcite . for a spatially homogeneous gas the generalized gross - pitaevskii equation ( [ gpeq ] ) is equivalent to the pines - hugenholtz identity @xcite . in the bogolyubov approach it simply gives @xmath79 , and eqs . ( [ bdgu]),([bdgv ] ) lead to the bogolyubov spectrum @xmath80 where @xmath81 is the momentum of the excitation . under the condition @xmath82 , for which the bogolyubov approach was originally developed , one can simply put @xmath83 equal to the total density @xmath84 in eq.([bogspectrum ] ) . for @xmath85 , which can be the case at @xmath86 , the dispersion law becomes essentially temperature dependent @xcite . in a spatially homogeneous gas the temperature dependence predominantly originates just from the presence of above condensate particles , with @xmath87 where @xmath88 is the bec transition temperature . this leads to the replacement @xmath89 in eq.([sigmas0 ] ) and gives @xmath90 . the dispersion law will be still given by eq.([bogspectrum ] ) in which the condensate density is now temperature dependent : @xmath91 $ ] . in this section we present the results for the damping rates and energy shifts of elementary excitations in an infinitely large spatially homogeneous bose - condensed gas . as one can see from eqs.([bdgu]),([bdgv ] ) , for finding the energy spectrum and wavefunctions of the excitations it is sufficient to calculate the self - energies @xmath42 , @xmath92 and @xmath43 . we will perform the calculations in the frequency - momentum representation and for physical transparency consider temperatures @xmath93 ( the opposite limiting case has been discussed by popov @xcite with regard to the phonon branch of the excitation spectrum ) . in the zero order approximation in the parameter @xmath94 we have the well - known mean field result : @xmath95 , @xmath96 , with @xmath97 ( see above ) . in this approach we obtain the bogolyubov quasiparticle excitations with the spectrum ( [ bogspectrum ] ) , which we use in order to calculate the next order in @xmath98 . the latter is determined by the contribution of diagrams containing one quasiparticle loop @xcite ( see figures 1 and 2 ) . actually in this approach we represent the hamiltonian as the sum of the ( diagonalized ) bogolyubov hamiltonian and the perturbation @xmath99 originating from @xmath100 ( [ hpert ] ) and containing the terms proportional to @xmath101 and @xmath102 : @xmath103.\ ] ] retaining only the temperature - dependent contributions , after laborious calculations for the normal self - energy we obtain @xmath104 , where @xmath105 here @xmath106 , @xmath107 , @xmath108 , @xmath109 is given by eq.([bogspectrum ] ) , @xmath110 is the equilibrium occupation number , @xmath111 and @xmath112 . similarly , the correction to the anomalous self - energy is given by @xmath113 the resonant parts @xmath114 originate from the terms where one of the intermediate quasiparticles is created and another one annihilated , and the non - resonant parts @xmath115 from the terms where both intermediate quasiparticles are created ( annihilated ) . temperature independent terms in the non - resonant parts , found by beliaev @xcite , are omitted in eqs.([sigman])-([sigmaar ] ) . each of the self - energies ( [ sigman])-([sigmaar ] ) is singular at @xmath116 and at least for small momenta the corrections become larger than the mean field values ( [ sigmas0 ] ) . nevertheless , keeping in mind that any physical quantity is determined by the combinations of the self - energies , which do not contain the infra - red singularities , we will still treat @xmath117 and @xmath118 as perturbations . for a spatially homogeneous gas the pines - hugenholtz identity @xmath119 gives the first order correction to the chemical potential @xmath120 and the relation between @xmath5 and the chemical potential , @xmath121 , coincides with that found by popov @xcite . the @xmath122 functions in generalized bogolyubov - de gennes equations ( [ bdggenu ] ) , ( [ bdggenv ] ) can be written as @xmath123 and @xmath124 , and in terms of the functions @xmath125 these equations take the form @xmath126 where @xmath127 considering the terms @xmath128 , @xmath129 in eqs . ( [ form1]),([form2 ] ) as small perturbations we put @xmath130 in the expressions for these quantities , following from eqs . ( [ sigman])-([sigmaar ] ) . then , solving eqs . ( [ form1]),([form2 ] ) , for the excitation energy we obtain @xmath131 , where @xmath132 _ { p\rightarrow ( \varepsilon _ { p},{\bf p})}\ ! , \label{cor}\ ] ] as @xmath133 and @xmath134 are complex , the correction to the excitation energy has both a real and an imaginary part : @xmath135 . the former gives the energy shift , and the latter is responsible for damping of the excitations . under the condition @xmath136 a straightforward calculation of eq.([cor ] ) on the basis of eqs . ( [ sigman])-([sigmaar ] ) and ( [ s+]),([s- ] ) , for the phonon branch of the spectrum ( @xmath137 ) yields @xmath138 it is important to emphasize that in this case the energy shift is determined by both resonant and non - resonant terms , whereas the damping rate @xmath139 is described solely by the resonant contributions . this type of damping originates from quasi - resonant scattering of thermal excitations from a given excitation ( landau damping ) and is absent at @xmath6 . both the energy shift and the damping rate are determined by the interaction of a given excitation with intermediate quasiparticles having energies @xmath140 . the damping rate @xmath139 ( [ reshomim ] ) coincides with that found in recent contributions @xcite and contains a slight numerical difference from the earlier szepfalusy - kondor result @xcite . the energy shift for the phonon branch of the spectrum was also calculated in @xcite . in the latter work the expansion of the self - energy functions near the point @xmath141 was used and formally divergent integrals were canceling each other in the final expression for the energy shift , which have led to the result by approximately factor 6 smaller then the shift ( [ reshomre ] ) obtained by the exact integration . eqs.([reshomre]),([reshomim ] ) clearly show that the small parameter of the theory is @xmath142 ( see eq.([small ] ) ) , whereas in the zero temperature approach @xcite the small parameter is @xmath143 . the presence of an additional large factor @xmath144 at finite temperatures @xmath145 originates from the bose enhancement diagrams containing one quasiparticle loop : compared to the zero - temperature case the contribution of each of these diagrams is multiplied by the bose factor @xmath146^{-1}$ ] ( or @xmath147 ) . as the most important is the contribution of intermediate quasiparticles with energies @xmath148 , for @xmath145 the bose factor @xmath149 . the criterion similar to eq.([small ] ) was found by popov @xcite as the condition which allows one to use the mean field approach at finite temperatures and to renormalize the theory for reaching beyond this approach . remarkably , the criterion ( [ small ] ) is fulfilled even at temperatures very close to @xmath150 . for @xmath151 we have @xmath152 , and eq.([small ] ) gives @xmath153 , which coincides with the well known ginzburg criterion @xcite for the absence of critical fluctuations . the criterion ( [ small ] ) also ensures that the main contribution to the damping rate originates from the interaction of a given excitation with thermal excitations through the condensate , i.e. , from the first term in @xmath154 ( [ hint ] ) . at energies @xmath155 ( but still @xmath156 ) the results for the energy shift and damping rate , following from eq.([cor ] ) , can be represented in the form @xmath157 the functions @xmath158 and @xmath159 have been calculated numerically . for @xmath160 we have @xmath161 , @xmath162 and eqs . ( [ homre ] ) , ( [ homim ] ) coincide with eqs . ( [ reshomre ] ) , ( [ reshomim ] ) . the dependence of @xmath163 and @xmath164 on the excitation energy is presented in fig.3 . for @xmath165 the damping rate increases with @xmath166 and reaches its maximum at @xmath167 . further increase of @xmath166 leads to slowly decreasing @xmath164 . for single - particle excitations ( @xmath168 ) we obtain @xmath169 . accordingly , the damping rate can be written as @xmath170 , where @xmath171 is the elastic cross section , and @xmath172 the particle velocity . this damping rate exceeds the beliaev temperature - independent term @xmath173 even at @xmath174 close to @xmath175 , if @xmath176 . in contrast to the phonon branch of the spectrum , for @xmath155 the damping is provided by both the szepfalusy - kondor ( @xmath17 ) and beliaev ( @xmath177 ) processes and , hence , can no longer be treated as landau damping . the small parameter of the theory is still given by eq.([small ] ) , since even at @xmath178 the energy of at least one of the thermal excitations is of order @xmath179 . the energy shift for @xmath180 is negative . the modulus of the shift increases with @xmath166 and reaches its maximum at @xmath181 . the further increase of @xmath166 decreases @xmath182 . the latter is equal to zero for @xmath183 , and becomes positive at larger @xmath166 . the above results for the damping rate and energy shift of a given excitation are obtained in the so called collisionless regime : we assume that the de broglie wavelength of the excitation , @xmath184 , is much larger than the mean free path of the thermal quasiparticles with energies @xmath185 , which are mostly responsible for the damping and shifts . it is also assumed that the excitation energy @xmath166 greatly exceeds the damping rate of these thermal excitations . the latter is of order @xmath186 ( see fig.3 ) , and for @xmath187 the two requirements of the collisionless regime are well satisfied under condition ( [ small ] ) . in the phonon branch of the excitation spectrum ( @xmath188 ) these requirements are equivalent to each other , and the collisionless criterion can be simply written as @xmath189 as clearly seen , in the phonon branch one can always find excitations which do not satisfy eq.([cols ] ) and , hence , require a hydrodynamic description with regard to their damping rates and energy shifts . the collisionless criterion ( [ cols ] ) provides an additional argument on support of the above used perturbative approach for solving eqs . ( [ form1 ] ) , ( [ form2 ] ) . under condition ( [ cols ] ) the term @xmath190 , the term @xmath191 , and the term @xmath192 . the non - mean - field shift @xmath163 is actually the shift of the excitation energy @xmath109 at a given condensate density @xmath5 . on the other hand , @xmath166 is determined by the bogolyubov dispersion law ( [ bogspectrum ] ) , with the temperature - dependent condensate density @xmath193 , and , hence , is temperature - dependent by itself . therefore , at a given @xmath174 one will also have the mean - field temperature - dependent energy shift @xmath194 . as the condensate density decreases with increasing temperature , @xmath195 is always negative . for @xmath145 it greatly exceeds the above calculated shift @xmath163 at any @xmath196 . the ratio @xmath197 decreases with temperature , but even for @xmath198 one has @xmath199 and @xmath200 . we now generalize the above obtained results for the case of elementary excitations in a spatially inhomogeneous ( trapped ) bose - condensed gas . as already mentioned in the introduction , a new ingredient here is related to the inhomogeneous density profile of the condensate and the discrete structure of the excitation spectrum . this requires us to develop a new method of calculating the self - energy functions in generalized bogolyubov - de gennes equations ( [ bdggenu ] ) , ( [ bdggenv ] ) . the self - energy operators in these equations are the sums of the zeroth and first order terms : @xmath201 and a similar relation for @xmath202 . at temperatures @xmath203 the zero order value of the above - condensate density in the condensate spatial region is coordinate independent and equal to the above - condensate density in the ideal gas approach : @xmath204 . on the contrary , the self - energies @xmath133 , @xmath205 depend explicitly on the condensate density . due to the discrete structure of the energy spectrum of excitations the expressions for these quantities should be written in the form of sums over the discrete states of intermediate quasiparticles @xmath8,@xmath9 . in the frequency - coordinate representation we have @xmath206 as we saw in the previous section , in the spatially homogeneous case all physical quantities are determined by the contribution to the self - energy functions @xmath133 , @xmath205 from intermediate quasiparticles with energies of order the mean field interaction between particles . the same holds for a spatially inhomogeneous ( trapped ) bose - condensed gas in the thomas - fermi regime , where the mean field interaction @xmath26 greatly exceeds the level spacing in the trapping potential . the intermediate quasiparticles with energies of order @xmath26 are essentially quasiclassical . with regard to the integral operator @xmath207 in the generalized gross - pitaevskii equation ( [ gpeq ] ) , which is solely determined by non - resonant contributions , this immediately allows one to replace the summation over the discrete intermediate states by integration . the kernel of this integral operator varies at distances @xmath208 of order the correlation length @xmath209 which is much smaller than the characteristic size of the condensate . therefore , the result of the operator action on the condensate wavefunction can be written in the local density approximation and , hence , should rely on eq.([mucorr ] ) with coordinate - dependent condensate density @xmath210 : @xmath211_{\varepsilon\rightarrow 0 } \psi_0({\bf r})= \int d^3r'\left[\sigma^{(1)}(\varepsilon,{\bf r},{\bf r'})- \sigma_a^{(1)}(\varepsilon,{\bf r},{\bf r'})\right]_{\varepsilon \rightarrow 0}\psi_0({\bf r'})=-\beta\psi_0 ^ 2({\bf r}).\ ] ] this result can be easily obtained from eqs . ( [ tnsn ] ) , ( [ trsa ] ) , where one should put @xmath212 , neglect the difference between @xmath213 and @xmath214 , and make a summation over @xmath9 . replacing the summation over @xmath8 by integration one should also take into account that for quasiclassical excitations the functions @xmath215 can be represented in the form @xmath216 where @xmath217 is the ratio of the local to total density of states for bogolyubov quasiparticles of a given symmetry , described by the classical hamiltonian @xmath218 on the basis of eq.([missed ] ) we obtain the generalized gross - pitaevskii equation in the form @xmath219 where @xmath220 is coordinate independent . compared to the ordinary gross - pitaevskii equation ( [ gpeq0 ] ) , eq.([gpeq1 ] ) contains an extra term @xmath221\psi_0 $ ] in the lhs . one can easily check that eq.([gpeq1 ] ) coincides with the equation @xmath222\psi_0=0\ ] ] obtained by averaging the non - linear schrdinger equation for the field operator . as mentioned above , for @xmath145 the above - condensate density @xmath223 in the condensate spatial region is mainly determined by the coordinate - independent ( ideal gas ) contribution @xmath224 . the coordinate - dependent correction to the above - condensate density , @xmath225 , turns out to be equal to the anomalous average @xmath226 : @xmath227 accordingly , the quantity @xmath228\psi_0= \widetilde{u}(2\langle\hat\psi^{\prime\dagger}\hat\psi^{\prime } \rangle + \langle\psi^{\prime}\psi^{\prime}\rangle)$ ] . taking advantage of eqs ( [ 0 ] ) , ( [ 00 ] ) and ( [ gpeq1 ] ) , the generalized bogolyubov - de gennes equations ( [ bdggenu ] ) , ( [ bdggenv ] ) are reduced to @xmath229 where the quantities @xmath230 , @xmath231 are given by eqs . ( [ s+ ] ) , ( [ s- ] ) , and @xmath25 is the exact value of the excitation energy . eqs . ( [ gpeq1 ] ) , ( [ fplus ] ) and ( [ fmin ] ) represent a complete set of equations for finding the energy shifts and damping rates of the elementary excitations . a precise calculation of the self - energy functions in eqs . ( [ fplus ] ) , ( [ fmin ] ) depends on the value of @xmath25 and on the trapping geometry . in this section we will make general statements on how the calculation can be performed . in most of the cases ( except the case of the lowest excitations with zero orbital angular momentum in spherically symmetric traps ) the characteristic time scale in the self - energy operators , @xmath232 , is much smaller than the inverse level spacing in the trap . therefore , the summation over the discrete intermediate states can be replaced by integration . this is a direct consequence of the general statement that the time - dependent discrete fourier sum can be replaced by its integral representation at times much smaller than the inverse frequency spacing ( see e.g. , @xcite ) . the kernels of the non - resonant parts of the self - energy operators , ( [ tnsn ] ) and ( [ tnsa ] ) , vary at distances @xmath208 which do not exceed the correlation length @xmath209 . as @xmath233 is much smaller than the characteristic size of the condensate , the non - resonant parts of the self - energies can be calculated in the local density approximation . one can easily find from eqs . ( [ tnsn ] ) , ( [ tnsa ] ) that they lead to the result of eqs . ( [ sigman ] ) , ( [ sigmaan ] ) ( with @xmath5 replaced by the coordinate - dependent density @xmath210 ) , multiplied by @xmath234 . for the lowest excitations @xmath7 one should put @xmath235 in eqs . ( [ sigman]),([sigmaan ] ) , and for quasiclassical excitations @xmath7 take @xmath196 from the bogolyubov dispersion law @xmath236 . the calculation of the resonant contributions to the self - energies is more subtle . using eq.([form ] ) for the functions @xmath237 in eqs . ( [ trsn ] ) , ( [ trsa ] ) , one can see that all resonant contributions contain the quantity @xmath238 writing @xmath239 as the integral over time @xmath240 , we obtain @xmath241 where the quantum - mechanical correlation function @xmath242 we will turn from the integration over the quantum states @xmath9 of the quasiclassical thermal excitations to the integration along the classical trajectories of motion of bogolyubov - type quasiparticles in the trap . following a general method ( see @xcite ) , employed in @xcite for the damping of low - energy excitations , we obtain @xmath243 where @xmath244 is the coordinate of the classical trajectory with initial momentum @xmath81 and coordinate @xmath245 . eq.([k2 ] ) will be used in the next sections where we demonstrate the facilities of the theory . concluding this section , we emphasize the key role of harmonicity of the trapping potential for temperature - dependent energy shifts of the excitations . as mentioned in the previous section , in the spatially homogeneous case at a given temperature the non - mean - field shift is much smaller than the shift @xmath246 appearing in the mean field approach simply due to the temperature dependence of the condensate density in the bogolyubov dispersion law ( [ bogspectrum ] ) . for the thomas - fermi condensate in a harmonic confining potential the situation is different . in this case the spectrum of low - energy ( @xmath77 ) excitations is independent of the mean field interparticle interaction @xmath26 ( chemical potential ) and the condensate density profile@xcite . hence , the temperature - dependent energy shifts can only appear due to non - thomas - fermi corrections . for finding these corrections one should use the mean field self - energies @xmath247 , @xmath248 , where the only difference from the @xmath6 case is related to the presence of above - condensate particles in the condensate spatial region at finite @xmath174 through the coordinate - independent term @xmath249 in @xmath250 . then eqs . ( [ fplus ] ) , ( [ fmin ] ) take the form of ordinary bogolyubov - de gennes equations ( [ fplus0]),([fmin0 ] ) , and eq.([gpeq1 ] ) becomes the ordinary gross - pitaevskii equation ( [ gpeq0 ] ) , with the chemical potential @xmath251 replaced by @xmath252 . the latter circumstance changes the condensate wavefunction compared to that at @xmath6 and ensures the temperature dependence of @xmath253 . accordingly , the excitation energies @xmath25 in eqs . ( [ fplus0]),([fmin0 ] ) also become temperature dependent . this type of approach , which for a spatially homogeneous gas would immediately lead to the result of lee and yang @xcite , has been used in recent numerical calculations of the energy shifts of the lowest quadrupole excitations in spherically symmetric @xcite and cylindrically symmetric @xcite harmonic traps . the presence of the coordinate - dependent part of the above - condensate density , @xmath254 , in these calculations is not adequate , since the anomalous average equal to this part was omitted and equations for the excitations did not contain the corrections to the self - energies , also proportional to @xmath4 . however , at @xmath145 , where @xmath255 , the coordinate - dependent part @xmath254 as itself should not significantly influence the result , and the calculations @xcite should actually demonstrate how important are the mean field non - thomas - fermi effects . the results of @xcite show the absence of energy shifts of the excitations at temperatures @xmath256 in the jila experiment @xcite and in this sense agree with the experimental data , but do not describe the upward and downward shifts of the excitation energies , observed experimentally at higher temperatures ( in this respect it is worth mentioning that the calculations @xcite performed for the thermal cloud in the hydrodynamic regime agree surprisingly well with the experiment @xcite ) . on the other hand , the calculation @xcite shows a downward shift of the energy of the lowest quadrupole excitation with increasing temperature in the conditions of the mit experiment @xcite . this is consistent with the experimental data and indicates that for not very small thomas - fermi parameter @xmath257 the mean field non - thomas - fermi effects can be important for temperature - dependent shifts of the lowest excitations . below we will assume a sufficiently small thomas - fermi parameter @xmath257 and demonstrate the use of the theory by the examples where the influence of non - thomas - fermi effects on the energy shifts of the excitations is not important . we will discuss the thomas - fermi condensates in a harmonic confining potential on the basis of eqs . ( [ gpeq1 ] ) , ( [ fplus ] ) , ( [ fmin ] ) . neglecting the kinetic energy term in eq.([gpeq1 ] ) , we arrive at a quadratic equation for @xmath253 . expanding the solution of this equation in powers of @xmath258 and retaining only the terms independent of @xmath258 and the terms linear in @xmath258 , for the condensate density we obtain @xmath259 where @xmath260 is the density of the thomas - fermi condensate in the ordinary mean field approach . we first consider the damping and energy shifts of quasiclassical ( @xmath261 ) low - energy excitations of a trapped thomas - fermi condensate , i.e. , the quasiclassical excitations with energies much smaller than the mean field interaction between particles @xmath26 . in this case the terms in eqs . ( [ fplus ] ) and ( [ fmin ] ) , originating from the kinetic energy of the condensate , can be omitted from the very beginning @xcite . then , using eq.([cd1 ] ) and treating the terms containing @xmath231 and @xmath230 as perturbations , we obtain @xmath262 , where @xmath263 is the excitation energy in the mean field approach , and the correction to the excitation energy @xmath264 is given by the relation @xmath265 here @xmath266 are the zero - order wavefunctions of the excitations , determined by the ordinary bogolyubov - de gennes equations ( [ fplus0 ] ) , ( [ fmin0 ] ) , with @xmath267 . in the case of quasiclassical excitations also the kernels of resonant parts of integral operators in eq.([pertenergy ] ) vary on a distance scale @xmath208 which does not exceed the correlation length @xmath233 . this can be already seen from eqs . ( [ q ] ) , ( [ k2 ] ) : the characteristic time scale @xmath232 in eq.([q ] ) is much shorter than @xmath268 and important is only a small part of the classical trajectory , where the condensate density is practically constant and @xmath269 , with @xmath270 . the correlation length @xmath233 is not only much smaller than the size of the condensate , but also smaller than the width of the boundary region of the condensate , where @xmath271 . therefore , the action of all integral operators on the functions @xmath266 in eq.([pertenergy ] ) can be calculated in the local density approximation . accordingly , for each of these operators one can use the quantity following from eqs . ( [ sigman])-([sigmaar ] ) , with @xmath272 and @xmath196 from the bogolyubov dispersion law @xmath236 . then , using eqs . ( [ form ] ) we can express the energy shift @xmath273 and the damping rate @xmath274 through the energy shift @xmath275 and damping rate @xmath276 of the excitation of energy @xmath25 in a spatially homogeneous bose - condensed gas with the condensate density equal to @xmath277 : @xmath278 the second term in the integrand of eq.([ecorghoms ] ) originates from the temperature dependence of the shape of the condensate wavefunction . for any ratio @xmath279 this positive term dominates over the negative term @xmath275 . the latter circumstance can be easily established from the results for @xmath280 in fig.3 . thus , for quasiclassical low - energy excitations the energy shift @xmath273 will be always positive , irrespective of the trapping geometry and the symmetry of the excitation . we confine ourselves to the case of cylindrical symmetry , where for the states with zero angular momentum one finds @xmath281 with @xmath282 , @xmath283 being the characteristic size of the condensate in the radial and axial direction , @xmath284 , @xmath285 the radial and axial frequencies , and @xmath286 the radial coordinate . the main contribution to the integral in eq.([ecorqhom ] ) comes from the boundary region of the condensate , where @xmath287 . from eqs . ( [ fplus]),([fmin ] ) one can easily see that in this region the possibility to omit the non - thomas - fermi effects originating from the kinetic energy of the condensate requires the condition @xmath288 . this condition ensures that the characteristic width of the boundary region greatly exceeds the excitation wavelength , and we arrive at the following relations for the energy shifts and damping rates of the excitations : @xmath289 it is important to emphasize that in the boundary region of the condensate , responsible for the energy shifts and damping rates of the quasiclassical excitations , the quantities @xmath290 , @xmath291 , and @xmath205 are determined by the contribution of intermediate quasiparticles which have energies comparable with @xmath25 . moreover , in this spatial region the quasiparticle energies are of order the local mean field interparticle interaction . as a consequence , the energy shift @xmath273 ( [ qcshift ] ) and the damping rate @xmath274 ( [ qcdamp ] ) are practically independent of the condensate density profile . for the same reason the damping rate is determined by both the szepfalusy - kondor and beliaev damping processes . therefore , similarly to the damping of excitations with energies @xmath292 in a spatially homogeneous gas , the damping of quasiclassical low - energy excitations of a trapped bose - condensed gas can no longer be treated as landau damping . the derivation of eqs . ( [ qcshift ] ) , ( [ qcdamp ] ) assumes that the motion of the excitation @xmath7 is quasiclassical for all degrees of freedom . we now turn to the condensate excitations in cigar - shaped cylindrical traps , which are quasiclassical only in the axial direction and correspond to the lowest modes of the radial motion . we will consider low - energy excitations ( @xmath77 ) , i.e. , the excitations with the axial wavelength much larger than the correlation length @xmath233 . in the recent mit experiment @xcite localized excitations of this type were created in the center of the trap by modifying the trapping potential using the dipole force of a focused off - resonant laser beam . then , a wave packet traveling along the axis of the cylindrical trap ( axially propagating sound wave ) was observed . in the mean field approach the sound waves propagating in an infinitely long ( axially homogeneous ) cylindrical bose condensate have been discussed in @xcite . for revealing the key features of the non - mean - field effects ( damping and the change of the sound velocity ) we confine ourselves to the same trapping geometry . with regard to realistic cylindrical traps this will be a good approach if the mean free path of sound waves is smaller than the characteristic axial size of the sample . as found in @xcite , for axially propagating sound waves radial oscillations of the condensate are absent , and the wavefunctions @xmath293 , with @xmath294 and @xmath295 being the axial momentum . the dispersion law @xmath296 is characterized by the sound velocity equal to @xmath297 , where @xmath298 is the maximum density of the thomas - fermi condensate in the ordinary mean field approach . it should be noted from the very beginning that , according to eq.([cd1 ] ) , @xmath299 is related to the corrected value of the maximum condensate density @xmath27 as @xmath300 . therefore , being interested in the sound velocity at a given value of the maximum condensate density , one should substitute this expression to eq.([disp ] ) . this immediately changes the sound velocity to @xmath301 in the leading term ( [ disp ] ) of the dispersion law and provides a contribution to the frequency shift of the sound wave @xmath302 the damping rate and other contributions to the frequency shift can be found directly from eq.([pertenergy ] ) by using the wavefunctions @xmath303 ( [ fk ] ) . the intermediate quasiparticles giving the main contribution to the damping rate and frequency shift have energies @xmath304 , i.e. much larger than the frequency of the considered sound wave , @xmath305 ( see below ) . therefore , similarly to the case of phonons in a spatially homogeneous condensate , the non - resonant terms ( [ tnsn ] ) , ( [ tnsa ] ) contribute only to the frequency shift . as already mentioned above , the characteristic distance scale @xmath306 in the kernels of the self - energies ( [ tnsn ] ) , ( [ tnsa ] ) is of order the correlation length @xmath233 , and the sum of their contributions to the frequency shift , @xmath307 , can be calculated by using the local density approximation for the action of the self - energy operators on the functions @xmath308 . as a result , we express @xmath307 through the non - resonant part of the energy shift @xmath309 in the spatially homogeneous condensate with the condensate density @xmath310 : @xmath311 for @xmath312 one can directly find from eqs . ( [ sigman ] ) , ( [ sigmaan ] ) that @xmath313 . then we obtain @xmath314 the resonant terms ( [ trsn ] ) , ( [ trsa ] ) contribute to both the frequency shift and damping rate . this means that the latter is determined by the szepfalusy - kondor scattering processes and , since the characteristic energies of intermediate quasiparticles are much larger than @xmath305 , can be treated as landau damping . the resonant contributions to the frequency shift and damping rate can not be found in the local density approximation , as the characteristic distance scale @xmath306 in the kernels of the self - energy operators in eq.([pertenergy ] ) is of order the radial size of the condensate . for finding these contributions one has to substitute the resonant parts of the self - energies , ( [ tnsn])-([trsa ] ) , to eq.([pertenergy ] ) and , by using eqs.([q])-([k2 ] ) , turn from summation over quasiclassical states @xmath8 , @xmath9 of intermediate quasiparticles to the integration along classical trajectories of their motion . then , a direct calculation yields ( cf . @xcite ) @xmath315 where @xmath316 is the classical trajectory starting at the phase space points @xmath317 on the ( hyper)surface of constant energy @xmath213 , @xmath318 , and @xmath319 generally speaking , the integration in eq.([deltak1 ] ) is a tedious task as it requires a full knowledge of the classical trajectories on a time scale @xmath320 . this is also the case in the idealized cylindrical trap , because of coupling between the radial and axial degrees of freedom . we will rely on the approach which assumes a fast radial motion of quasiparticles compared to their motion in the axial direction and , hence , requires the frequency of the sound wave , @xmath305 , significantly smaller than the radial frequency @xmath284 . then on a time scale @xmath320 the quasiparticles with energies @xmath321 ( which are the most important for the energy shifts and damping of the sound wave ) oscillate many times in the radial direction , whereas their axial variables @xmath322 , @xmath323 only slightly change and , hence , can be adiabatically separated from the fast radial variables @xmath324 , @xmath325 . in this case it is convenient to integrate eq,([deltak1 ] ) over @xmath326 and , using eq.([fk ] ) , represent it in the form @xmath327 where the integration is performed over the entire classically accessible region of the phase space . since @xmath322 is close to @xmath328 , in the exponent of the integrand we can write @xmath329 , where the axial velocity @xmath330 is obtained from the exact hamiltonian equations of motion by averaging over the fast radial variables : @xmath331 . for the classical radial motion ( @xmath332 ) the averaging procedure simply reduces to integration over @xmath333 under the condition @xmath334 at fixed values of @xmath213 , @xmath335 and @xmath328 , with the weight proportional to the local density of states for the radial motion : @xmath336 where @xmath337 . finally , averaging the function @xmath338 over the fast radial variables and integrating over @xmath339 in eq.([interm ] ) , we obtain @xmath340 the resonant contribution to the frequency shift , given by the real part of eq.([dkf ] ) , after the integration proves to be @xmath341 . the sum of this quantity with the non - resonant term ( [ dnonres ] ) and @xmath342 ( [ s1 ] ) leads to the frequency shift of the sound wave @xmath343 the imaginary part of eq.([dkf ] ) gives the damping rate @xmath344 except for the numerical coefficients , eqs,([sndshift ] ) , ( [ sd ] ) are similar to eq.([reshomre ] ) , ( [ reshomim ] ) for the damping rate and energy shift of phonons in a spatially homogeneous bose condensate . this a consequence of the fact that the condensate boundary region practically does not contribute to the damping rate and frequency shift of axially propagating sound waves , in contrast to the case of excitations quasiclassical for both axial and radial degrees of freedom . in the mit experiment @xcite the characteristic spatial size of created localized excitations was @xmath345 @xmath251 m and , accordingly , so was the initial wavelength of propagating sound . according to the experimental data , the propagating pulse died out during 25 ms , and after that only the lowest quadrupole excitation characterized by a much loner damping time ( @xmath346 ms ) was observed . we believe that the attenuation of axially propagating sound in the mit experiment @xcite on the time scale of 25 ms can be well explained as a consequence of damping . the characteristic frequency of the waves in the packet can be estimated as @xmath347 . then eq.([sd ] ) gives the damping rate independent of the condensate density @xmath299 . in the mit experiment the temperature @xmath3480.5 @xmath251k was roughly only twice as large as @xmath26 , which decreases the damping rate by approximately @xmath349 compared to that given by eq.([sd ] ) . in these conditions we obtain a characteristic damping time of @xmath350 ms , relatively close to the measured value . the relative change of the sound velocity , @xmath351 , increases with decreasing condensate density @xmath27 . however , even at the lowest densities of the mit experiment @xcite ( @xmath352 @xmath353 ) the quantity @xmath354 does not exceed @xmath355 and is practically invisible . we acknowledge fruitful discussions with m.w . reynolds , m.a . baranov and a. griffin . this work was supported by the dutch foundation fom and by the russian foundation for basic studies .

photosynthesis , arguably the most important photo - induced process on earth , converts the energy of light into its chemically / biologically useful form . it is often argued that this conversion is remarkably efficient . however , it has to be distinguished between the _ quantum _ efficiency , reaching almost unity@xcite , and light - to - chemical energy efficiency , which is significantly lower , in the order of several percent@xcite . this is given by an evolutionary pressure on the development of a robust photosynthetic machinery optimized for survival rather than energy conversion efficiency . energy relaxation processes are integral part of the photosynthetic function as they enable the energy transfer to proceed unidirectionally @xcite , and provide protection of the photosynthetic apparatus against harmful over - excitation . the photosynthetic machinery of plants has developed a complex hierarchy of self - regulatory mechanisms to avoid excess excitation or ( when unavoidable ) to dissipate it into heat @xcite . starting from processes controlled on the macroscopic level ( e. g. orientation of leaves ) , over spontaneous microscopic ( cellular ) events such as chloroplast movements , to truly nano- and sub - nanoscopic mechanisms such as reorganization of antenna complexes and direct regulation of energy transfer on the level of small groups of interacting chromophores , plants actively react to changing illumination conditions . the sub - nanoscopic processes , which are the focus of the present study , operate in response to the increase of @xmath0 across the thylakoid membrane . such an increase is an indicator of high illumination . nowadays it is generally accepted that carotenoids are involved in these energy dissipation processes . the precise molecular mechanism is , however , still subject of discussion@xcite . it is likely that different mechanisms evolved in different classes of organisms and/or that several mechanisms operate at once . most of our current knowledge about the early processes of photosynthesis was obtained by ultrafast spectroscopy . while conventional bulk spectroscopies are extremely useful in following ultrafast photo - induced events in photosynthesis , whenever structural inhomogeneity of the sample is involved , the information obtained from these spectroscopies becomes obscured by an inevitable ensemble averaging . some well established spectroscopic methods , such as hole - burning @xcite , and some modern multidimensional methods , such as coherent two - dimensional spectroscopy @xcite , provide certain access to the homogeneous properties . information on the behaviour of individual molecules has been , however , available only since the advent of single - molecule spectroscopy ( sms ) @xcite . the system studied in this work , the light harvesting complex ii ( lhcii ) of higher plants , is the plants major light - harvesting antenna containing almost half of all the chlorophyll in the chloroplast . correspondingly , most of the light absorption and subsequent energy transfer processes in plants and algae occur in this complex . the lhcii antenna occurs naturally in a trimeric form and its main function is to deliver excitation energy to the nearby photosystem ii ( psii ) . given the major role of lhcii in light - harvesting and energy transfer , it is not surprising that it is also implicated in participating in regulated energy dissipation , the so - called non - photochemical quenching ( npq ) @xcite . the crystal structure of the complex @xcite enables us efficient theoretical modeling of the complex s spectroscopic properties using its chromophores ( chlorophylls ( chl ) and carotenoids ) as the model units . the model parameters such as electronic coupling are greatly constrained by the known mutual orientation and distances . existence of a large body of previous measurements together with theoretical attempts to fit this whole body of data to a single model @xcite gives a great confidence in extending the modeling towards single molecular experiments . in recent years , single - molecule spectroscopy ( sms ) experiments on several photosynthetic antenna complexes including the lhcii were performed . fluorescence spectral peak distributions @xcite , spectral diffusion @xcite , fluorescence intensity distributions @xcite and time traces @xcite were obtained from these measurements . in many cases also fluorescence intermittency ( blinking ) was observed , and it was conjectured that in the case of lhcii the process behind the fluorescence intermittency plays a role in npq @xcite . as far as the theory of the sms of lhcii is concerned , the ensemble - averaged spectra and also the peak distributions of lhcii trimers can be successfully explained by the disordered excitonic model @xcite . in general , the excitonic model was successfully applied in the past on fs to ps timescale , and it represents an indispensable tool in analyzing ultrafast spectroscopic experiments on molecular aggregates and in particular on photosynthetic antennae@xcite . its application to longer time scales of seconds and minutes is conditioned by the assumption of a certain separation of time - scales . over the course of the excitation - emission cycle ( nanoseconds ) , individual chromophores of a complex are assumed to be found in fixed spatial arrangements and experiencing an environment described by a fixed set of parameters . the emission spectrum of a complex in such a fixed spatial arrangement is predicted by the excitonic model which gives the population distribution of the excited state manifold and the probability of emission at the corresponding transition frequencies . the spectrum of exciting light matters only to the extent to which the excited states reached at a given excitation wavelength are connected to the final state by some relaxation pathways . once the pathways are available , the final state is given irrespective of the initial state after absorption of light . because only the final distribution of excited state population matters , the changes ( switches ) of the spatial arrangement or environmental conditions occurring on the sub - nanosecond time scale are only observed as sudden changes ( with respect to the nanosecond fluorescence time scale ) of the fluorescence spectra . despite the fact that the experiment we analyze in this work does not provide more insight into the sub - nanosecond dynamics of individual complexes than previous works , we nevertheless formulate the theory consistently in such a way that it enables the description of such processes . this is done in order to highlight the existence of a less studied type of processes on an intermediate time scale and to stress the need to search for experimental techniques which can cover the range of timescales from femtoseconds to nanoseconds in single molecular spectroscopy . we thus provide theoretical techniques to treat these future experiments . we have recently reported a progress towards measuring processes on the order of hundreds of femtoseconds in single light - harvesting complexes @xcite . the paper is organized as follows : in section [ sec : excited - state - population ] we first discuss the frenkel exciton model as a basis for the formulation of equations of motion for the populations of the excited states of chromophore aggregates with strong interchromophore couplings . then , in section [ sub : excited - state - dynamics ] , we introduce equations of motion for excitonic populations valid over a broad ( from ps to min ) timescale range . we discuss their generality and the range of validity . it is argued that these equations provide an ideal means for the description of the sms experiments . in section [ sec : lhcii ] the results of section [ sub : excited - state - dynamics ] are applied to lhcii photosynthetic antenna complexes . all results are compared to the experiment . it is shown that our equations give correct fluorescence spectrum and peak statistics , i. e. appropriate steady - state population , under typical sms experiment condition . then we model the intensity traces , while the switching behaviour is included by incorporating one particular previously proposed npq mechanism , namely energy transfer to lutein lut 1 . the switching between on and off states is controlled by a 2-level model , where the switching causes a change of the chl a612 - lut 1 coupling . it is shown that using realistic parameters we are able to reproduce the experimental results . the details of our energy relaxation theory and of the stochastic model of switching are given in supporting information ( si ) . in the usual sms experiments , time - resolved ( on the times scale of 10 ms to s ) fluorescence of the studied molecules is observed . if the excited state life time of the studied chromophores is sufficiently long ( nanoseconds in the case of chlorophylls studied here ) , the expected fluorescence spectra can be calculated from the steady state , quasi - equilibriated populations of the excited electronic eigenstates of the molecular system , assuming canonical equilibrium . depending on the strengths of the chromophore - chromophore resonance interaction and the system - bath couplings , the eigenstates can be approximated either by the excited eigenstates of the electronic subsystem or the excited state of the individual chromophores forming the aggregate . in the present paper , we treat both cases within one formalism provided by the framework of the frenkel exciton model . the frenkel exciton model provides an excellent tool for the treatment of pigment - protein complexes on femto- and pico - second time scales @xcite . the basic notion of the frenkel exciton model is the one of localized excited states of the chromophores . these states have negligible wavefunction overlap with neighboring excited states , and they can therefore be assumed orthogonal to each other , forming a suitable basis for the aggregate hilbert space . in the treatment of linear absorption and fluorescence experiments , only singly excited collective states of the molecular aggregate have to be included . the localized excited states can thus be denoted as @xmath1where @xmath2 and @xmath3 are the electronic ground and excited states of the chromophore denoted by index @xmath4 , respectively . the states @xmath5 form a complete hilbert subspace for the case that exactly one molecule of the aggregate is excited . the molecular system hamiltonian , @xmath6 , is however rarely diagonal in the basis of the states @xmath5 and resonance couplings @xmath7 between excited electronic states @xmath5 and @xmath8 occur in all situations interesting for light - harvesting . in the absence of a protein environment , the light would resonantly excite eigenstates of the system hamiltonian . in all realistic cases , the interaction of the system with its environment co - determines the excitation frequencies , and prescribes thus a `` preferred basis '' of electronic states in which it is the most advantageous to formulate the energy transfer theory . in the case of the resonance interaction @xmath7 exceeding the typical reorganization energy @xmath9 of the protein bath of the chromophores @xmath4 and @xmath10 , ( @xmath11 , we include @xmath7 explicitly into the system hamiltonian . the hamiltonian is diagonalized to obtain electronic eigenstates , and the effect of the protein bath is included via perturbation theory yielding a redfield type relaxation tensor . in the opposite case @xmath12 we neglect the resonance coupling in the system hamiltonian , and its effects are included perturbatively yielding frster type energy transfer rates between localized excited states @xcite . let us first focus on deriving a closed set of equations for the excited state populations . extensive work was done in the last years on developing methods to accurately describe the system dynamics following an ultrafast excitation by external light . the traditional redfield and frster approaches were superseded by more accurate ( exact in some cases ) methods , such as heom @xcite , tedopa @xcite and other methods @xcite . these methods bring unprecedented accuracy at an increased numerical cost . it seems , however , that for analyzing many of the recent experiments , it is still possible to rely on the traditional tools , as they capture the physics of the problem ( and often even the quantitative aspects of the problem ) very well @xcite . commonly , equations of motion for some relevant degrees of freedom ( dof ) , electronic states in our case , are derived by reducing the liouville - von neumann equation for the total density matrix to an equation for the so - called reduced density matrix ( rdm ) . these equations describe time evolution of a molecular system for a fixed configuration of the protein environment ( assuming fast fluctuations around this fixed configuration ) , and they are therefore suitable for the description of ultrafast laser experiments . however , on the timescale relevant in sms ( up to tens of seconds ) , usage of these equations is actually not appropriate . first of all , some transient effect at the short times affect even the long time properties of the system and the steady state population dynamics , and second , slow changes in the protein environment can entirely change the energy landscape including the case that one has to change the theoretical limit ( localized states , delocalized states ) in which one works . the latter case is especially difficult to treat and requires to go beyond the traditional master equation approaches which we apply here . in this work we therefore concentrate on the extension of the validity of the master equation approach towards long times under the assumption that the dependence of the hamiltonian of the system on time is negligible within one absorption emission time scale ( nanoseconds ) . as for the transient effects at short time , when dealing with fast dephasing of optical coherences , short time transient effects are responsible for the absorption lineshape . this aspect of the transient effects will be taken into account in full . spontaneous emission of photons by chlorophylls occurs with a nanosecond life - time . another class of transient effects , namely dynamic electronic coherence due to excitation by light , is therefore also unimportant and its treatment can be avoided . we therefore derive approximative equations for the populations dynamics only , with the validity in the range from picoseconds to tens of seconds . initially , our hamiltonian consists of the system , bath and system - bath interaction terms , and we formally assume some total density matrix @xmath13 which follows the liouville von neumann equation . there are several methods how to arrive to a master equation ( in a convolution - less form ) rdm @xmath14 @xcite . in general , such equation reads : @xmath15-{\cal r}(t , t_{0})\rho(t),\label{eq : equation_of_motion}\ ] ] @xmath16 the total hamiltonian @xmath17 consists of the bath renormalized system hamiltonian @xmath18 and the system - light interaction , which is ( in the dipole approximation ) given by the dipole moment operator @xmath19 and electric field @xmath20 . the bath is completely eliminated in the reduced description , and its effects are represented by the relaxation tensor @xmath21 . it is important to note that in eq . ( [ eq : equation_of_motion ] ) , the relaxation tensor @xmath21 explicitly depends on some initial time @xmath22 , in which the initial condition for the propagation is set , or more precisely , the relaxation tensor depends on the quantity @xmath23 . an exact relaxation tensor would also depend on @xmath20 . in general , the choice of @xmath22 is arbitrary , and we should choose it such that the temporal profile of the excitation field is non - zero for @xmath24 only . however , in most practical theoretical approaches , the evolution of the bath due to excitation of the system is reflected in the relaxation tensor @xmath21 by the bath correlation function , which decays rather quickly . by choosing @xmath22 sufficiently small , we could always make such a relaxation tensor constant . this is obviously an artifact of the approximations used . for a smooth envelope of the external field changing on the same or slower timescale than the correlation function of the bath , there is no good choice of @xmath22 . the relaxation tensor always becomes constant too early . luckily , as we will see below , we do not need to account for the external field @xmath20 beyond first order perturbation theory , and in this regime , there is a natural choice of time @xmath22 which enables us to correctly account for the transient time dependence of the relaxation tensor even for a steady state externally driven by light . it is important to note , that the relaxation tensor @xmath21 is completely abstract up to now . it can represent some exact relaxation superoperator , or a result of perturbation theory with respect to some parameters , such as the redfield or frster tensors . let us now formulate the equations of motion for the excited state evolution in the linear regime of the system s interaction with the electric field . the validity of the linear regime has been discussed e.g. in refs . @xcite , and it is the same as the validity of the third order response theory for non - linear laser spectroscopy . we will also use the secular approximation ( equations of motion for populations @xmath25 and coherences @xmath26 , @xmath27 are assumed independent ) , although this we do only to simplify the resulting equations . secular approximation could be avoided if one so wishes , at the cost of treating the full dm . we write explicitly the elements of the rdm of eq . ( [ eq : equation_of_motion ] ) in which we keep the double time dependence of the hamiltonian and the relaxation tensor . we get the following set of coupled equations : @xmath28 @xmath29 here , the @xmath30 are transfer rates between populations ( from @xmath10 to @xmath4 ) , @xmath31 is the population relaxation rate of the state @xmath4 , including the rate of radiative depopulation @xmath32 ( we assume it is constant ) , and @xmath33 is the optical coherence dephasing rate . the equation of motion for the optical coherences , eq . ( [ eq : coherences eq ] ) , can be solved by introducing the interaction picture using the evolution operator element @xmath34 , and integrating eq . ( [ eq : coherences eq ] ) in the interaction picture . by returning back to the schrdinger picture , we obtain the actual field - induced and field driven optical coherence element in the form : @xmath35e(t - t^{\prime}).\label{eq : coherence_optical}\ ] ] we can see that eq . ( [ eq : coherence_optical ] ) is actually not a solution of eq . ( [ eq : coherences eq ] ) . in solving eq . ( [ eq : coherences eq ] ) numerically , the dephasing rate would quickly become constant for @xmath36 and the transient effects for @xmath37 would be lost completely . however , in eq . ( [ eq : coherence_optical ] ) these effects are properly taken into account . the discrepancy between eq . ( [ eq : coherences eq ] ) and eq . ( [ eq : coherence_optical ] ) is due to a different treatment of the bath . in a response function approach , of which the linearization of the full equations of motion with respect to the field is a variant , it is in general possible to account for the bath in a more consistent way than in master equations . in ordinary master equations , the bath is correctly described at @xmath22 , and the description of its subsequent evolution after @xmath22 is extremely limited . an example of such limitation is discussed e.g. in ref . @xcite . in most biological energy transferring systems , the pure dephasing is much faster ( hundreds of fs ) than the population relaxation ( units and tens of ps ) . we can therefore assume that the populations remain constant during the integration in eq . ( [ eq : coherence_optical ] ) . moreover , it is reasonable to set the upper limit of the integration in eq . ( [ eq : coherence_optical ] ) to infinity by sending @xmath38 . we can expect that the simultaneous action of pure dephasing and external driving by a field with a slowly varying envelop creates a steady state optical coherence . the dependence on @xmath22 in the upper limit of the integral describes a transient evolution of the optical coherence after switching on the interaction with the field . now that the short time time - dependent nature of the dephasing rates is taken into account correctly , we can set @xmath38 and write : @xmath39.\label{eq : coherence ft}\ ] ] it is important to note that now @xmath40 is a global time which can run through the whole minutes long sms experiment . the evolution operator changes on an ultrafast time scale , but this timescale is scanned in the integration over the variable @xmath41 . the properties of the hamiltonian @xmath42 change on a very slow time scale ( with respect to optical dephasing ) , and so does the evolution operator element @xmath43 . unlike the eq . ( [ eq : coherences eq ] ) , which is valid for @xmath23 small , eq . ( [ eq : coherence ft ] ) is valid for all times . the purpose of deriving eq . ( [ eq : coherence ft ] ) was to insert it eventually into eq . ( [ eq : populations eq ] ) . also here we face the problem of transient time dependence of the rates . however , because populations change much more slowly , these effects are not as important as in the case of coherences . they can be , however , treated rigorously , even including bath memory effects between evolutions by eq . ( [ eq : coherences eq ] ) and eq . ( [ eq : populations eq ] ) , as we have shown elsewhere @xcite . in this work , we will assume the energy transfer and relaxation rates not to depend on the difference @xmath23 , although they may depend weakly on the time @xmath40 , i. e. @xmath44 and @xmath45 . we are ready now to insert eq . ( [ eq : coherence ft ] ) for optical coherences into the equation for populations , eq . ( [ eq : populations eq ] ) . we obtain @xmath46 here , we defined @xmath47 , and we used the fact that @xmath48 and even @xmath34 are ordinary c - numbers . in eq . ( [ eq : driven_pop_evol ] ) , the populations are driven by a second order field term . we have treated the field classically so far . if we did that quantum mechanically , we would now have to trace over the field dof in order to obtain reduced equations of motion for the electronic state populations only . the term @xmath49 would thus be replaced by @xmath50 which can be interpreted as a quantum mechanical expectation value . the latter expression has the form of two - time correlation function of the electric fields of the light and its fourier transform is the power spectrum of the light @xcite , @xmath51 now , inserting eq . ( [ eq : power_spectrum ] ) into eq . ( [ eq : driven_pop_evol ] ) , switching the order of integrations , and using the definition of absorption lineshape of the i - th excitonic transitions ( see e.g. @xcite ) @xmath52 we arrive at @xmath53\int d\omega w(\omega)\chi_{i}(\omega;t).\label{eq : popevolution}\ ] ] our effort has yielded a closed set of equations for excitonic populations only . the population changes are given by the transfer rates between electronic levels , population quenching and a source terms expressed as an overlap of the excitonic spectra with the light spectrum . all quantities are in principle dependent on time , most importantly the excitonic absorption spectrum and all rates can weakly depend on time to simulate slow changes of the protein and chromophore configurations . also the light spectrum can be considered time - dependent by generalizing the wiener - khintchine theorem for the instaneous power spectrum @xmath54@xcite . the changes can be faster than the time resolution of the sms experiment , but they have to be slower than the dephasing , or even energy transfer dynamics . we should also note here that if we wish to stay strictly in the linear regime , the saturation term @xmath55 in eq . ( [ eq : popevolution ] ) should be very close to one . in fact , we can set it equal to one as a reasonable approximation with the same validity as the linear approximation for fields . according to crystalographic studies , lhcii complex consists of three monomeric units , each containing 14 chlorophylls and for carotenoids : two luteins , neoxanthin and a carotenoid of the xanthophyll cycle@xcite . the experiments described in this section were performed on monomeric lhcii complexes . in accord with the experiment , we focus on one such a monomeric unit in our model . we treat the lhcii monomer as a strongly coupled systems of chromophores , weakly coupled to the bath and weakly interacting with light . because only absorption and fluorescence are measured , we do not attempt to fit the site energies in our simulations from scratch , as the fitting of this limited set of experiments would not be unique . instead , we take the pigment transition energies from ref . @xcite , where both lhcii trimers and monomers where treated . the coupling energies between the pigments were calculated in the dipole - dipole approximation , and the dipole orientations were taken from the crystal structure using an effective dipole strengths of 3.4 d for chl b and 4.0 d for chl a. the bath is described by means of a spectral density obtained by fluorescence line narrowing experiment ( fln ) ( see ref . excitonic absorption and fluorescence lineshapes are calculated by means of the second order cumulant expansion ( see ref . @xcite ) , and the population transfer rates are calculated by the multilevel redfield theory ( see ref . @xcite ) . for comparison , we also calculated the rates by modified redfield theory , ref.@xcite , and we concluded that the results remain essentially the same . the population relaxation rates of chlorophylls were taken to be 3 ns in accord with the experiment @xcite . the equations of motion , eq . ( [ eq : popevolution ] ) , allow us to use light with any spectral composition . in the experiment described in this paper , we use spectrally narrow ( laser ) illumination at 630 nm . for details on the calculations see si . in figs . [ fig : excitonic - model]a and [ fig : excitonic - model]b we present the calculated bulk absorption and fluorescence spectra of lhcii monomers at 5 c compared to experimental values taken from @xcite . we note here that the bulk spectra of lhcii in monomeric and trimeric form are practically identical , see fig . s1 in si for comparison . in calculations , the spectra were averaged over a gaussian disorder of site energies with full width at half maximum ( fwhm ) of 110 @xmath56 . although the blue chl b shoulder is not perfectly reproduced , the agreement between measured and theoretical absorption is good in the region of our excitation , and the fluorescence ( fl ) spectrum shows a good agreement in general . we therefore conclude that our excitonic model captures correctly the features of the studied system that are the most relevant in the present study . the fl spectrum in fig . [ fig : excitonic - model]b is dominated by the lowest four excitons , which are the most populated ones . these excitons are formed by strongly coupled pigments chl 610 - 611 - 612 , chl 602 - 603 and chl 613 - 14 ( see ref . @xcite for nomenclature ) . this is in agreement with previous modeling results for the trimeric lhcii @xcite . bulk ( a ) absorption and ( b ) fluorescence spectrum of the qy band of lhcii monomers . the points are experimental values taken from ref . @xcite , the lines are calculated by our exciton model . the coloured lines are individual excitonic contributions , the black line is the overal spectrum . ] the calculated bulk spectra seem to be in a good agreement with the experiment . our model also reproduces successfully the statistics from the sms experiments . in fig . [ fig : fl - peak - distribution]a we present the fl spectral peak distribution compared to the experiment . we can see that the calculated distribution is a little broader than the experimental one , but the agreement is again reasonably good . the larger distribution width of the calculated spectra can be explained by a relatively long integration time in the experiment ( 1 s ) , during which the system samples several individual realizations of the disorder . in extreme cases the measured values get averaged towards the mean value . as a result the measured distribution is narrower . from our comparisons between experiments and calculations we conclude that the exciton model with gaussian disorder not only reproduces the averaged absorption spectra and equilibriated populations of excitons ( resulting in characteristic fl spectra ) , but also the individual realizations provided by this model are in a good agreement with the experiment . experimental ( red ) and calculated ( green ) fluorescence peak distribution . ( a ) peak position histogram , ( b ) fl peak position and intensity plot . theoretical points are calculated as individual realizations of energetic disorder . ] the next feature of the lhcii single molecular spectra that we need to address is the significant amount of blinking , i. e. reversible switching to the off state . since the measured dwell times in the off state are in the range of seconds , which is significantly longer than the lifetime of any long - living species such as triplet states @xcite , the off states must correspond to states with efficient excitation energy dissipation . correspondingly , our model has to be extended by including some fluorescence quenching mechanism . one of the possible mechanism of fl quenching in lhcii , proposed by ruban and coworkers @xcite , is an excitation energy transfer from the lowest chl a exciton states to a lutein molecule , lut 1 ( lut620 ) , see ref . @xcite for nomenclature ) . the lut 1 molecule resides in the vicinity of the so - called terminal emitter group of chlorophylls , composed of chls a610 , 611 and 612 , and it is supposed to be coupled mainly to chl a612 @xcite . the s1 state of the lut molecule is optically forbidden , and it has a short ( 10 ps ) lifetime due to a decay through a non - radiative channel @xcite . the transition energy from the s0@xmath57s1 of lut is in the vicinity of the transition to the chl @xmath58 state . due to its short excited state life - time , lut could in principle act as an excitation ( and fluorescence ) quencher . let us test this mechanism within our model to see if it can account for the observed blinking . the important parameters of the lutein in context of our model are its s1 state site energy relative to its groundstate and the coupling to chlorophylls , in particular to chl a612 . role of lut 1 parameters . dependence of the fluorescence yield on ( a ) lut 1 s1 energy and ( b ) lut 1 - chl a612 coupling . the energy dependence is calculated with redfield ( blue ) and modified redfield ( red ) theory for comparison . the dependence on the coupling strength depicts calculated points ( redfield theory ) fitted with exponential dependence . already a realistically small coupling around @xmath59 leads to significant fl quenching . ] in fig . [ fig : the - lut-1]a we present the dependence of the relative fl quantum yield on the lut energy for fixed value of the lut - chl coupling of 14 @xmath56 . the energy dependence agrees well with the one obtained by ruban @xcite . the quenching is only efficient when the lut energy is below one of the red chlorophylls ( around 15100 @xmath56 ) and the plateau enables lut to act as an efficient quencher even in disordered systems . in fig . [ fig : the - lut-1]b we show the dependency of the fl quantum yield on the lut - chl coupling for fixed lut energy 14500 @xmath56 . due to large reorganization energy , 14500 @xmath56 corresponds to the zero - phonon line at 13900 @xmath56 and thus agrees with experimental observations @xcite . from the coupling dependence of fluorescence in fig . [ fig : the - lut-1]b we can conclude that weak coupling is sufficient for lut 1 to act as a fluorescence quencher . very importantly , even small changes in the lut - chl coupling can result in a big difference in the fluorescence intensity . based on this analysis we decided to use lutein s1 site energy of 14500 @xmath56 in our simulations of blinking . we define the quenched state by the value of 12 @xmath56 for coupling of the lut 1 to chl 612 and the unquenched state by the zero coupling . our model allows any type of time dependency of the chl - lut coupling to be used , and it could in principle accommodate input from structure based md studies and quantum chemical treatment of the ( dexter type ) coupling of the chl @xmath58 and the lut s1 states . however , a much rougher phenomenological model of the chl - lut coupling changes enables a better discussion of the feasibility of the proposed quenching mechanism than the parameter free _ ab initio _ calculations suffering from uncertainties in the structural information . next we proceed to model the blinking statistics . as mentioned in the introduction , the blinking statistics alone can be well described by a two - level model proposed by valkunas et al . in ref . @xcite . by random fluctuations , the protein samples its potential energy surface ( pes ) performing thus a random walk ( rw ) . the model of ref . @xcite assumes that there are two stable conformations of the protein corresponding to two minima of the protein pes . these are approximated by two harmonic potentials . the protein undergoes a rw in this potential , and at every step it has a certain probability to switch from its current pes to the other pes . in our treatment we use a discrete rw description , which enables us to follow individual trajectories of the proteins . for the details of the approach taken in this study and the differences from the original model by valkunas et al . , see si and refs . @xcite . to connect the two pes model to our particular lut quenching model , we assume that the change of protein conformation somehow changes the lut - chl coupling . the lut s1 state does not have a dipole moment , and the resonance coupling similar to those between allowed states does not occur here . the two different protein conformations responsible for the quenched / unquenched states would then result in two slightly different orientations / positions of the pigments , leading to different strengths of the coupling . this mechanism is in accord with recent quantum - chemical study by duffy et al . @xcite , where small configurational changes were found to lead to substantial changes in chl - car couplings . the switching between the pes is controlled by the rw model with diffusion parameters adjusted to fit the experimental dwell time distributions . the comparison between the calculated and experimentally determined dwell - times is presented in figs . [ fig : dwell - times]a and [ fig : dwell - times]b , for the on and off times , respectively . the agreement is again fairly good letting us believe that our phenomenological model captures the most important features of the protein dynamics affecting the blinking behaviour . experimental ( red ) and calculated ( green ) probability distribution of dwell times in the ( a ) on and ( b ) off state , logarithmic scale . while the off state distribution follows a power law , the on state distribution has an exponential tail at longer times . ] finally , we can connect the two models described above and simulate the blinking behaviour . to this end we continuously model the fluorescence of the lhcii complexes , and output the intensity ( and spectrum ) every 10 ms corresponding to the experimental integration bins . simultaneously , we let the protein do the rw on its pes , and we adjust the chl / lut coupling , when the protein switches between pes . to obtain more realistic traces , either the site energies or couplings can be slightly varied after every jump , resulting in different energy levels . such a procedure , however , does not lead to qualitatively different conclusions and it can be in principle omitted . in the calculations presented here , we used gaussian disorder with the fwhm of 0.3 @xmath56 for couplings and 1 @xmath56 for energies . for every realization of the disorder a 60 s trace is modelled . this is repeated for 200 realizations of the disorder , reflecting the experimental conditions . the resulting statistics are presented in fig . [ fig : intensity - traces - statistics ] . total dwell times in fig . [ fig : intensity - traces - statistics]a represent the overall amount of time spent in a given intensity level . from this we can see the two - level character of the blinking and simultaneously also the presence of some intermediate levels , which result from particular disorder realizations . the dwell times are similar for the on and off states . [ fig : intensity - traces - statistics]b shows how often are the intensity levels visited per fixed amount of time . the modelled access frequency distribution is naturally very symmetrical , a direct result of the fact that in the model the complex switches only between the on and off state . the number of on / off states visited per minute therefore has to be the same . the experimentally analyzed intensity traces contain also jumps between levels within the on / off states in an amount which can be , to some extent , modified by adjusting the sensitivity of the level analysis . the presence of these intra - state jumps results in higher switching frequency and wider distribution in intensities , causing a moderate discrepancy between experiment and simulation . in order to include this kind of switching into the model , dynamic sampling of the disorder would have to be incorporated . work in this direction will be presented elsewhere . for the reasons stated above , the level access frequency distribution is not well suited for comparison of the model with the experiment . a more appropriate measure of the blinking would be the intensity - intensity correlation function defined as @xmath60 . this quantity is well - known from single - molecule measurements , where it is often used to characterize the blinking behaviour @xcite . in fig . [ fig : intensity - traces - statistics]c we present @xmath61 obtained from 50 measured long enough traces , compared with the model . we can see that the agreement between experiment and theory is good , indicating that our model gives reasonable switching between the intensity levels . the shape of the correlation function is given by the dwell time statistics , see also fig . [ fig : dwell - times ] . the initial fast drop implies the abundance of short blinking events . this results from the mechanism of the protein switching between its potential surfaces , where the short succesive blinking events are caused by the dynamics in the vicinity of the pes intersection . experimental ( red ) and calcuated ( green ) intensity blinking statistics . ( a ) the percentage of time the complexes dwelled on respective intensity levels . the two - state structure of the low - intensity off states and higher intensity on states is apparent . ( b ) how often per minute the complexes accessed the respective intensity levels . the experimental frequency is higher due to switching within the on / off states . ( c ) the intensity correlation function . ] in section [ sub : excited - state - dynamics ] of the results we derived approximate equations of motion for populations evolution on the timescale of the sms experiment . we have shown that the weak illumination regime , in which the sms experiment is performed , allows for an effective source term description of the light - matter interaction in which short time transient effects arising in the photoinduced evolution of molecular systems can be consistently accommodated side by side with the slow evolution of the protein bath observed in sms experiments . we demonstrated the validity and scope of application of our equations by simulating our single molecule experiment on lhcii monomers . based on the recent research in elucidating the npq mechanism in lhcii and the connection between fluorescence quenching and energy dissipation we implemented energy transfer to lut 1 as a blinking mechanism . based on our calculations we were able to confirm the findings of duffy et al . @xcite and chmeliov et al.@xcite that , within a reasonable range of parameters , lut 1 can indeed act as an efficient quencher . since our model extends the previous treatment by including realistic excitation conditions and population transfer rates , it is remarkable , how similar our fluorescence quenching dependence on the lut 1 energy is to the one in ref . moreover , we were able to confirm that lut 1 acts as an efficient quencher also under am1.5 illumination ( data not shown since the dependence is very similar ) . at the same time we can see that the amount of quenching is very sensitive to the change of coupling of the lut to the chlorophylls . since the coupling itself is very sensitive to the distance and orientation between the pigments , it provides a possible link to the protein conformation change working as a switching mechanism as proposed in ref . indeed , when using the 2-level switching model to control the change of coupling , we are able to reproduce the experimentally obtained blinking statistics . although far from being exclusive in any way , our argument strongly supports the notion of the protein acting as a conformational switch regulating the amount of quenching in the system . the agreement between the theory and experiment also serves as a good demonstration of the scope of our equations . they provide a description for controlling the energy transfer in the system by modulating the parameters of the excitonic model . we are aware of the remaining phenomenological nature of our connection of the 2-level switching model with the excitonic model . further improvements in the direction of introducing more parameters with a particular physical meaning , for example the relation to the actual pes shape , are needed and will be subject to further study . also recent experimental observations indicate the presence of more relevant timescales in the intensity traces suggesting the inadequacy of a 2-level model with one reaction coordinate . finally , although some connection between the fluorescence blinking and npq was already shown by krger et al . @xcite , their exact relation is yet to be elucidated . the equations derived here are a suitable tool for these future investigations . [ [ section ] ] trimeric lhcii complexes of spinach were isolated from freshly prepared thylakoid membranes as described earlier @xcite . monomeric complexes were obtained by incubating lhcii trimers with 1% ( w / v ) octyl glucoside and 10 @xmath62 phospholipase a2 ( sigma ) @xcite . subsequent fast protein liquid chromatography ( fplc ) ensured a homogeneous sample preparation . the ensemble fluorescence absorption and emission spectra were measured on a lambda40 spectro - photometer ( perkin - elmer ) and a fluorolog tau-3 ( jobin yvon ) , respectively . for sms experiments , the sample was diluted down to a concentration of ~ 10 pm in a measuring buffer ( 20 mm hepes , ph 8 and 0.03% ( w / v ) n - dodecyl @xmath63-d - maltoside ) and then immobilized on a pll ( poly - l - lysine , sigma ) coated cover glass . addition of an oxygen scavenging mix of 750 @xmath64 glucose oxidase , 100 @xmath65 catalase and 7.5 @xmath66 glucose ( all sigma ) to the closed sample chamber inhibited the formation of highly reactive singlet oxygen and improved the photostability of complexes . a confocal microscope was used to measure the fluorescence of single complexes at 5 @xmath67 as described previously @xcite . the sample was excited at 630 nm utilizing a ti : sapphire laser ( coherent mira 900f ) with a pulse width of 200 fs and a repetition rate of 76 mhz coupled to a tunable optical parametric oscillator ( coherent mira opo ) . near - circular polarized light was obtained by utilizing a berek polarization compensator ( 5540 new focus ) . a fluorescence beam splitter ( 70:30 , thorlabs ) allowed us to simultaneously measure the fluorescence spectrum _ via _ a ccd camera ( spec10:100br , roper scientific ) with an integration time of one second and the wavelength integrated fluorescence intensity _ via _ an avalanche diode ( spcm - aqr 16 , perkin elmer ) with a binning time of 10 milliseconds . the fluorescence of one complex was analyzed for either one minute or until it photo - bleached and a set of 200 complexes served as the basis for statistical analysis . the fluorescence peak distribution was obtained by fitting of a skewed gaussian to the fluorescence spectrum as shown in kruger et al . @xcite and the blinking analysis was performed equivalently to the algorithm described elsewhere @xcite . the equations of motion , eq . ( [ eq : popevolution ] ) , have quasi constant coefficients , and they can therefore be written in the form @xmath68 where @xmath69 is a matrix of relaxation and population transfer rates , and @xmath70 are the source terms . ( [ eq : matrix_eq-1 ] ) can be solved analytically : @xmath71 this expression enables us to find the populations at any time without the need to solve for all the previous times . if we aim at steady state , @xmath22 can be send to @xmath72 and the populations then depend only parametrically on time @xmath40 . the weak dependence of @xmath73 and @xmath69 on @xmath40 makes it possible to explain changes in the populations of the emitting states of a molecular system due to slow changes of the protein environment and the structure of the molecular system . 10 url # 1`#1`urlprefix[2]#2 [ 2][]#2 _ _ ( , ) . & . _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . , & _ _ ( , , ) . , & . _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . _ et al . _ . _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . , , , & _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . et al . _ . _ _ * * ( ) . & . _ _ * * , ( ) . , , , & . _ _ * * , ( ) . , , , & . _ _ * * , ( ) . _ _ * * , ( ) . & . _ _ * * , ( ) . _ et al . _ . _ _ ( ) . _ et al . _ . _ _ * * , ( ) . , , & . _ _ * * , ( ) . , , & _ _ * * , ( ) . , & ( ) . , , & _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . , , & _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . , & _ _ * * , ( ) . , , , & . _ _ ( ) . & _ _ ( , ) . , & _ _ ( , ) . & _ _ * * , ( ) . _ _ * * , ( ) . & . _ _ ( ) . , , & . _ _ * * , ( ) . & _ _ * * , ( ) . & _ _ * * , ( ) . , , , & . _ _ * * , ( ) . , , & . _ _ * * , ( ) . _ et al . _ . _ _ * * , ( ) . _ _ ( , , ) . & . _ _ * * , ( ) . & . _ _ * * , ( ) . & _ _ * * , ( ) . , & . _ _ * * , ( ) . & . _ _ * * , ( ) . & . _ _ * * , ( ) . _ _ ( , , ) . . _ _ * * , ( ) . , , & . _ _ * * , ( ) . & . _ _ * * , ( ) . , , & . _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . , & . _ _ * * , ( ) . _ _ * * , ( ) . , , , & _ _ * * , ( ) . _ et al . _ . _ _ * * , ( ) . , , & ( eds . ) _ _ ( , ) . , , , & . _ _ * * , ( ) . _ et al . _ _ _ * * , ( ) . _ _ ( , ) . p.m. , j.m.g . and r.v.g . were supported by the vu university and by an advanced investigator grant from the european research council ( no . 267333 , photprot ) to r.v.g . was also supported by the nederlandse organisatie voor wetenschappelijk onderzoek , council of chemical sciences ( nwo - cw ) via a top - grant ( 700.58.305 ) , and by the eu fp7 project papets ( ga 323901 ) . r.v.g . gratefully acknowledges his academy professor grant from the netherlands royal academy of sciences ( knaw ) . p.m. and t.m . received financial support from the czech science foundation ( gacr ) , grant no . 14 - 25752s . t. m. acknowledges the kind support of the nederlandse organisatie voor wetenschappelijk onderzoek ( nwo ) visitor grant nr . 040.11.423 during his stay at the vu amsterdam . the interaction of the pigments , i. e. the system , with the vibrations , i. e. its environment , is treated by second order perturbation theory . there the bath is completely described by its correlation function or , equivalently , in the frequency domain , by the spectral density of the bath vibrations . we use the spectral density from ref . it is constructed from one overdamped oscillator representing the slow protein vibrations and 48 high - frequency intrapigment modes ( for more context for the spectral density and excitonic model see ref . @xcite ) : the ( temperature dependent ) correlation function given by this spectral density is assumed to be uncorrelated between individual sites and differs only between chl a and chl b , while the difference is in the coupling strength @xmath75 : the time - dependent correlation function is then obtained by fourier transform of ( [ eq : correl.func . ] ) . we use the same values of the spectral density parameters as in ref . the difference between the vertical , franck - condon transition of the pigments , which are called site energies in this text , and their 0 - 0 transitions is given by the reorganization energy due to the interaction with the bath : because the pigments are strongly coupled , the preferred basis of calculations is the excitonic basis in which the system hamiltonian is diagonal with eigenvalues , exciton energies , @xmath78 . all quantities , including correlation functions @xmath79 , reorganization energies @xmath9 and transition dipole moments @xmath19 have to be transformed into the excitonic basis : the population relaxation rates of chlorophylls are a result of energy transfer and radiative and non - radiative decay , i.e. @xmath92 , where @xmath93 , @xmath94 is inverse lifetime @xmath95 of site @xmath96 excited state , which is taken to be 3 ns for all @xmath96 . to check for sample degradation a control bulk measurement of absorption and fluorescence was performed . in fig . [ fig : bulk - spectra - measured ] the measured spectra are given together with the spectra of lhcii trimers taken from @xcite . their excellent agreement confirms absence of degradation in course of the experiment and also justifies the usage of the trimer spectra for the bulk spectra modelling in the main text . in the original description by valkunas @xcite and , in more detail , in @xcite , the protein diffusive motion was described as a continuous - time random walk ( ctrw ) on a two - dimensional potential energy surface . here we somewhat simplify this description in the following way . the two generalized coordinates represent fast and slow degrees of freedom and thus , when following the slow dynamics , the fast fluctuations can be adiabatically eliminated . it can be shown that when we consider the potential dependence on the fast coordinate the same in both the on and off states for simplicity , i.e. setting @xmath97 in @xcite , the fast coordinate can be completely removed . this leaves us with effectively one - dimensional problem . @xmath104 is a rate of falling from the @xmath105 potential , the first exponential term reflects the energy gap law and the min term ensures detailed balance condition . the coefficient @xmath106 can be treated as a constant and @xmath107 is a characteristic frequency of the protein environment vibrations responsible for the tunneling . the protein diffusion under this conditions can be described either by the ctrw or by a discrete random walk ( rw ) . the former approach was employed by valkunas et al.@xcite . however , we believe that for our purpose it is better to solve this problem as a discrete rw for two reasons . first , the coupled smoluchovski equations for the ctrw on the two potentials can be decoupled only for conditional probabilities , i.e. assuming that the system was in the opposite state in the previous interval , and , in the same time , employing the same , equilibrium initial condition for each dwell time . when continuously modelling the trajectory of a single protein , we do not have to include the resetting after switching and also the conditioning will be inherent , as the system is observed being in a particular state . and second , when we want to simulate the individual intensity time traces , it is more natural to really follow the trajectories of the individual proteins on their pes . if we want to follow the time trace of every single molecule , we should follow its particular trajectory . the blinking statistics will then be recovered by averaging over a large number of molecules , exactly as in the experiment . to the purpose of following trajectories of individual proteins , we need to describe its discrete rw ( drw ) in the potential . we will denote probability of going right ( left ) as @xmath108 ( @xmath109 ) . in the symmetrical rw we have @xmath110 . if the protein is at a coordinate @xmath111 , in the next step it will move with probability @xmath108 to @xmath112 and with probability @xmath109 to @xmath113 , where @xmath114 is the length of the step . inspired by classical approach by van kampen @xcite , we augment the position dependent probabilities in the presence of the potential @xmath115 as we note that defined probabilities defined in eq . ( [ eq : pprobandqprob ] ) reflect the detailed balance condition @xmath117 . considering a small step @xmath114 , we can use taylor expansion in @xmath111 @xmath118 , obtaining where @xmath125 is the time duration of the step . during @xmath125 the protein walks either to the left or right with the respective probability and with the probability @xmath126 switches to the other surface . similarly to @xcite the position near the potential intersection can be chosen as an initial condition . however , as we do not include resetting after switching in our approach , this determines only the starting point of each trajectory and is therefore not of significant importance . in table [ tab : rw - model - paramaters ] we present the parameters of our rw model compared to the ones used by valkunas et al @xcite . considering the differences - lhcii monomers vs trimers , discrete vs continuous rw , no resetting vs resetting - the agreement is satisfactory .

the powerful infrared source and oh maser vy canis majoris is one of a few very luminous cool hypergiant stars that define the empirical upper luminosity boundary in the hr diagram @xcite . it is thought to represent a very rapid stage of evolution characterized by a high mass loss rate . the red supergiant stage is a few hundred thousand years long , or less for the most massive stars , but the transition time from main sequence to rsg and later from rsg back to a yellow hypergiant may be only a few thousand years @xcite . these evolved late - type stars are major sources of gas , dust and molecules that contribute to the replenishment of the interstellar medium . vy cma , in particular , is ejecting large amounts of gas and dust at a prodigious rate . with its extended and complex circumstellar nebula , it offers a unique opportunity to study its high mass loss episodes and the detailed kinematics of this process . vy cma is a special case even among the rare hypergiants . it is associated with a region of recent star formation , the young cluster ngc 2362 and h ii region sharpless 310 , and the dark cloud l1667 , and has been considered a pre main sequence star ( herbig 1970a , b ; lada & reid 1978 ) as well as an evolved massive star @xcite . the lack of li i @xmath26707 @xcite plus its powerful double peaked maser emission , like other evolved stars , is strong evidence that it is a massive post main sequence object . the measured @xmath3c/ @xmath4c ratio @xcite however is intermediate between what is measured for normal m supergiants and what is expected for the ism . at a distance of 1.5 kpc @xcite it is one of the most luminous red supergiants known ( @xmath5 ) . with its very visible asymmetric nebula 10@xmath6 across and high mass loss rate ( @xmath7 ) @xcite , vy cma is also one of the most important for understanding the high mass loss episodes in massive star evolution . interferometric maps of the oh and h@xmath8o emission reveal a complex structure interpreted as either an expanding disk or bipolar outflow ( bowers , johnston , & spencer 1983 ; richards , yates & cohen 1998 ) with expansion velocities of 30 40 km s@xmath1 . speckle imaging @xcite revealed an asymmetric resolved component 0@xmath91 across , elongated approximately north - south . a geometry consistent with the maser maps and the infrared images @xcite is an axially symmetric model with a slightly flattened optically thick envelope or an equatorial disk and a bipolar flow / rotational axis aligned roughly northeast - southwest and inclined 15 to 30@xmath10 to our line of sight @xcite . however , our multi - wavelength hst / wfpc2 images of vy cma revealed a complex asymmetric distribution of material with no clear axisymmetric structures ( see figures 1,2 , & 3 in smith et al . ( 2001 ) , and with no evident disk - like structure in the reflected optical light extending out to 2@xmath6 or more from the star as proposed by herbig ( 1970a , b ) , efstathious & rowan - robinson ( 1990 ) , and monnier et al ( 1999 ) . instead , the circumstellar nebula is dominated by a prominent , nebulous arc to the northwest and two bright filamentary loops or arcs to the southwest of the star , plus some relatively bright condensations or clumps of dusty knots near the star , all of which are evidence for multiple and asymmetric mass loss events . the apparent random orientations of the arcs suggest that they were produced by localized events not strongly aligned to the star s axis or equator . the models for disks are based on modelling the spectral energy distribution and therefore may be due reflected light from the prominent tail and the bright condensations near the star rather than a true disk . the arcs and loops also do not have the appearance of continuous outflow streams as suggested by monnier et al . ( 1999 ) , and are too massive to have been ejected by radiation pressure on dust . the initial ejection must have been induced by other processes . we therefore speculated that the arcs may be expanding loops caused by localized activity on the star s ill - defined surface , involving magnetic fields and convection . although this suggestion was conjectural , large starspots do occur on the surface of @xmath11 ori ( gilliland & dupree 1996 ; kluckers et al . 1997 ; uitenbroek , dupree & gilliland 1998 ; dupree , lobel & gilliland 1999 ) with accompanying outflows and chromospheric activity @xcite . non - radial pulsational instability is potentially an alternative ejection mechanism , but the distinction between pulsation and convection may be blurred in vy cma . to learn more about the morphology , kinematics , and possible origin of vy cma s complex ejecta , we have obtained long - slit spectra to map the emission and absorption lines across several structures , including the prominent nw arc and the two bright filamentary loops to the southwest . our results reveal the presence of a complex pattern of velocities in the the strong k i emission line and a significant velocity gradient across the dusty northwest arc . we find that ; + @xmath12 most of the gas and dust in the nebula , however , is nearly stationary or moving very slowly with respect to the star . this is not only surprising but may be difficult to explain for an evolved , massive star . + @xmath12 embedded in this nebula are several discrete structures that are kinematically distinct from the surrounding stationary nebulosity . the nw arc and the two prominence like arcs are expanding at 40 50 km s@xmath1 relative to the star . + @xmath12 they were ejected in separate events over several hundred years from different locations on the star . + @xmath12 a strong absorption line in the reflected spectrum in the ejecta is identified with ca i ] . although absorption from a semi - forbidden line seems unlikely , we demonstrate that the identification is probably correct . + the strong k i emission in the star and the ejecta has never been adequately explained . in the appendix , we discuss a possible mechanism for forming the k i emission and other rare emission lines , including ca i ] . in the next section we briefly describe the observations and the data reduction procedures . the spectrum of the central star and the origin of its peculiar emission lines are discussed in 3 and in the accompanying appendix . the kinematics of the gas and dust , the expansion of the arcs and loops , and the evidence for stationary ejecta are described in 4 . in the final two sections we summarize the evidence for localized high mass loss events and comment on the perplexing nature of vy cma . spectra were obtained at four slit positions across the nebula ( figures 1a and 1b ) with the hires echelle spectrograph on the keck1 telescope on mauna kea in december , 2002 . the resolution was 40000 with a dispersion of .06 per pixel and a spatial scale of 0@xmath919 per pixel . our spectral coverage from @xmath136410 to 8727 in 15 orders with inter order gaps was chosen to measure the velocities of known emission lines plus several absorption lines in the 8400 8700 region . we used the b3 decker which gives a 14@xmath6 long slit with a 0@xmath957 slit width . three of the slit positions are parallel across the nebula and cross clumps of knots , filaments , and the nw arc . one of the slit positions(i ) includes the star . the fourth slit ( v ) crosses the two prominence like arcs , 1 and 2 , in the outer parts of the ejecta to the southwest and the clump of knots south of the star . the seeing was estimated as 0@xmath98 or better throughout the night . the observations are summarized in table 1 . all of the data reductions and extractions were performed using iraf . the frames were trimmed , bias subtracted and flatfielded . calibration and object exposures were combined for each slit position . using the quartz exposures for reference , we extracted each echelle order from the combined frames using the ` strip ' option in the iraf task _ apsum_. the dispersion corrections were determined for each order separately . no sky subtraction was done because emission and absorption lines are still apparent in the most distant apertures , beyond the visible nebulosity . the spectra were then extracted in 1 long apertures deliberately placed across prominent features including the nw arc , the filamentary arcs , the knots and the embedded star . additional extractions were then made along the slit keeping 0@xmath94 separation between apertures . the positions of the apertures are summarized in table 2 ; the positions are given along each slit with respect to the point nearest the star , the radial distance of each aperture from the star , and the position angle from the star . figures 1a and 1b show the positions of the extraction apertures for slits i and ii , and for slits iii and v , respectively . four extractions were made along slit iii , across specific features listed in table 2 plus four additional apertures to the end of the slit . the wavelength calibration was then applied to each extracted spectrum . the separate exposures were averaged to remove cosmic rays , except for those along slit i which were summed because only two exposures were available . vy cma probably represents the most extreme known case of stellar activity , in the same sense as , e.g. , @xmath11 ori . its inner wind ( chromosphere ? ) has a density comparable to that of @xmath11 ori , but the volume , mass , and energy are larger by almost two orders of magnitude . the photosphere is appreciably cooler than betelgeuse and is about one order of magnitude more luminous . the observed emission lines particularly the strong k i and the ca i ] emission features differ from @xmath11 ori and have not been explained . in the next few paragraphs we review what one _ expects _ to see in this situation . the photospheric temperature , luminosity , and radius are thought to be , respectively , around 3000 k characteristic of an m5 type supergiant , @xmath14 , and @xmath15 @xmath0 12 au . ( if the wind is opaque , then @xmath16 is larger than the true stellar radius @xmath17 , the sonic point in the flow . ) the usually quoted mass - loss rate and wind speed , @xmath18 @xmath19 yr@xmath1 and 35 km s@xmath1 , imply that the wind density at @xmath20 au is @xmath21 @xmath22 or @xmath23 g @xmath22 . given those parameters , we expect dust grains to condense near radius @xmath24 au , at temperatures around 1000 k. if the resulting dust - to - gas ratio is `` normal '' , then , at red wavelengths @xmath25 , the opacity is @xmath26 @xmath27 g@xmath1 and the average optical thickness due to dust outside @xmath28 is @xmath29 . observations of scattered light at radii @xmath30 2000 au seem consistent with this amount of dust @xcite . with such a large average @xmath31 , one might naively expect the object to be seen only as a mid - ir source . in reality , however , the dust should be very inhomogeneous . even if the inner wind is smoothly isotropic ( which we doubt ) , radiative acceleration acting on the dust near @xmath28 is unstable in a rayleigh - taylor - like manner ; the ensuing flow must be clumpy with a fairly small size scale . = wind speed , @xmath32 = sound speed in the wind , and @xmath33 = acceleration due to radiation acting on the dust ; then the three most obvious characteristic size scales for instabilities near @xmath28 are @xmath34 , @xmath35 , and @xmath36 . for vy cma , the last of these is of the order of 510 au while the others are much less ; so the expected condensations in the region just outside @xmath28 should be , typically , no larger than 10 au , and very likely smaller . ] therefore we should not be surprised to see visual- or red - wavelength light from the star , escaping along low - density paths through a very inhomogeneous dusty shell . since the material is expanding , and a typical photon may be scattered one or more times before it escapes , the emergent spectrum is likely to be red - shifted by an amount comparable to the wind speed @xmath37 , and somewhat blurred in wavelength . van blerkom and van blerkom ( 1978 ) described this process for a constant wind speed , and the effect may be supplemented by radiative acceleration of material in the regions most favorable for photon escape , i.e. , in relatively low - density gas between condensations . after escaping from the inner dusty region , some photons can be scattered sideways by dust in the outer ejecta , causing us to see an extended , inhomogeneous halo around the bright central region . any photon that is scattered toward an appreciably non - radial direction is doppler shifted by a net amount @xmath38 , where @xmath37 is the outward wind speed and @xmath39 is ( @xmath40scattering angle ) . thus we expect the spectrum reflected in the halo to be more red - shifted than that seen when we observe the central source . since a line of sight through the halo typically samples a range of @xmath39-values , the reflected spectrum should be blurred more than the central spectrum . generally speaking , observations of vy cma confirm the above expectations . at wavelengths around 7000 , very roughly 2 percent of the star s light escapes through the dust ( see measurements reported in smith et al . 2001 ) , suggesting typical optical depths of the order of 4 or 5 , not 50 , along the photon escape routes . absorption lines in the central emergent spectrum are doppler - shifted by about + 25 km s@xmath1 relative to the supposed systemic velocity , and are somewhat broader than one expects for such a low - gravity star . when reflected in the halo , the same features have larger doppler shifts and are more blurred ( fig . 6 ) . throughout this paper we shall note additional details that do _ not _ fit the above picture in straightforward ways . the k i resonance lines ( @xmath13 7665 , 7699 ) are not only the strongest emission lines in the spectrum of vy cma @xcite , but these lines , rarely observed in emission , are stronger in vy cma than in any other known late - type star ( see bernat & lambert 1976 , guilain & mauron 1996 ) . in addition to k i , strong emission is also present in the na i `` d '' lines , ca i ] @xmath26572 , and rb i ( @xmath137800 , 7948)@xcite . narrow band heads of sco ( wallerstein in hyland et al 1969 and herbig 1974 ) , and band heads of tio and vo , plus lines of fe i , ti i and cr i @xcite are also in emission . our wavelength coverage includes k i @xmath27699 , ca i ] @xmath26572 and rb i @xmath27948 , but not the na i `` d '' lines . in addition to the emission features , there are strong absorption lines of ca ii , fe i , and ti i in the wavelength region from approximately 8400 to 8700 that are relatively free from molecular absorption bands so they can be confidently used for radial velocity measurements . in our data , the central star is in aperture 2 on slit i ( fig . 1a ) , and examples of the emission and absorption lines in the ejecta are shown in figures throughout the paper . the spectrum at the star does not show any significant differences from those published earlier and the velocity measurements for the star summarized in table 3 are consistent with previously published results . all velocities in this paper are heliocentric . in figure 2 we show the k i and na i line profiles on the star from a hires spectrum obtained by george herbig in 2000 that includes the d lines . both the k i and na i emission have peak fluxes several times as high as the continuum , and not surprisingly , their doublet ratios show that they are optically thick . the minima in the strong p cygni profiles have the same velocity at @xmath0 15 km s@xmath1 . for the k i lines , this is a shift of @xmath0 30 km s@xmath1 from the emission peaks which have a mean velocity of 45 km s@xmath1 , consistent with our measured k i and ca i ] emission velocities . the na i emission peaks however occur at @xmath41 60 km s@xmath1 which agrees with our mean absorption line velocity at the star ( table 3 ) . it has been recognized for some time @xcite that the absorption line velocities are not only different from the k i and ca i ] emission lines , but also deviate significantly from the velocity of the system inferred from the oh and h@xmath8o maser observations , v@xmath42 20 22 km s@xmath1 ( v@xmath43 @xmath41 35 37 km s@xmath1)(bowers , johnston & spencer 1983 ) . herbig ( 1970a ) first suggested that the stellar absorption lines are red - shifted and broadened due to scattering by an expanding dust shell or equatorial disk , and as mentioned above , this was modelled by kwok ( 1976 ) and van blerkom and van blerkom ( 1978 ) . note that our velocities for the k i and ca i ] emission lines at @xmath41 41 km s@xmath1 are significantly less than the mean absorption line velocity . this difference is usually attributed to the location where the emission lines are formed , outside the inner radius of the dusty shell or disk . the extremely strong k i emission and other rare emission lines such as semi forbidden ca i ] are difficult to explain . in the appendix we discuss problems associated with the formation of the k i lines in vy cma . some obvious excitation processes such as collisional and radiative recombination are eliminated , and while resonant scattering is the most promising explanation , it requires a special geometry to cause strong net emission . we propose a model for the production of the very strong k i emission observed at the star , but problems associated with resonant scattering remain , especially in the ejecta . the strong k i emission line is our best tracer of the gas in vy cma . it shows considerable variability with multiple emission peaks throughout the ejecta ( see also smith 2004 ) . examples of the line profiles along slits i , ii and iii are shown in figure 3 , with their velocities summarized in table 4 . the emission profile on the star is single ( slit i , ap 2 ) with a strong p cygni absorption feature formed in its expanding wind . the p cygni profile is present in the reflected spectrum near the star and is gradually red - shifted due to scattering by the dust expanding outwards . off the star , the k i profiles along the slits are much more complex , often with two or more recognizable emission peaks . in the outermost apertures the profiles become single and narrow . an emission peak with the expected systemic velocity is observed throughout much of the ejecta to the west and northwest of the star . it has a nearly constant velocity at increasing distances and is essentially motionless relative to the star . this and other evidence for stationary gas and dust in vy cma s ejecta are discussed in 4.3 . interestingly , a broad and shallow k i absorption feature to the red of the emission profile is present at positions where the slits cross the dusty condensations and nebulous arcs primarily to the west and nw of the star . this may be an inverse p cygni profile , but the broad absorption minimum has a red - shifted velocity , typically @xmath41 86 km s@xmath1 ( table 5 ) like the other reflected absorption lines at the same positions ( see 4.1 ) . note that on average , the k i emission greatly exceeds the strength of the absorption in our data . one would normally expect emission to balance absorption for a resonance scattered line ( see the appendix ) . the semi - forbidden ca i ] @xmath26572 emission line ( figs . 4 and 5 ) is visible only in the extractions on and near the star . the line is very narrow , does not show the multiple peaks observed in the k i line , and has essentially a constant doppler velocity at @xmath0 41 km s@xmath1 like that for the k i emission on the star ( table 4 ) . a strong , reflected absorption line which we attribute to ca i ] absorption is present in the ejecta . although absorption in a semi - forbidden line seems unlikely , we demonstrate later , 4.1 , that this identification is almost certainly correct . another rare emission line , rb i at @xmath2 7948 , is present in the reflected spectra , but is relatively weak and broad . h@xmath11 emission in vy cma is usually reported as very weak and variable . we find that the relative strength of the h@xmath11 emission depends on position in the nebula ( see figs . 4 and 5 ) . in the inner region of the ejecta it is weak and broad with an associated broad shallow absorption . the narrow h@xmath11 emission line visible in the outermost extractions along all of the slit positions is nebular from the nearby h ii region sharpless 310 . vy cma lies just to the west of an arc of nebular emission . weak nebular emission from [ n ii ] and [ s ii ] is also present in the spectra at the same positions . the velocity of this h@xmath11 line at @xmath0 35 37 km s@xmath1 agrees with the systemic velocity and the velocities of the k i lines measured at the same positions , thus confirming the close association of vy cma with the cluster and h ii region . the reflected and scattered spectrum in the circumstellar ejecta reveals a remarkably complex behavior of both the emission and absorption lines . the k i emission line profiles and their doppler velocities vary with position in the nebula as the slits cross recognizable knots , arcs , and filamentary features in the image . they also indicate apparently stationary gas in the same projected locations , which is difficult to understand for the reasons noted below . the reflected absorption lines show a strong velocity gradient across the prominent nebulous arc to the nw of the star , while the same absorption lines along slit v show virtually no variation in doppler velocity . these phenomena are discussed in the following subsections . the absorption lines from the obscured star and the dusty ejecta are reflected or scattered by the embedded dust grains and reveal both the relative motions of the dusty clouds and the velocity dispersions within them . the mean velocities of the strong absorption lines of fe i , ti i , and ca i in the far red spectrum are summarized in table 5 and examples of their profiles are shown in figure 6 . a few of the lines in apertures on and near the nw arc are double or slightly asymmetric . for example , four of the absorption lines in i ap 3 are double so we used the blue components , which agree with the velocities of the single lines , for the mean velocity in table 5 . in the extractions immediately on either side of the nw arc , a few of the absorption lines also have asymmetric profiles . only the single lines and well defined absorption minima of the some of the asymmetric lines were used for the mean velocity . the asymmetries may be due to additional dusty clouds along the line of sight or to the expected gradient along the nw arc . the absorption line profiles from the dusty arcs are broad and shallow compared to the narrower profiles on and near the star . along slit i , for example , the equivalent widths of the absorption lines show only a small decrease from the star to ap 3 and ap 4 ( nw arc ) , but the fwhm of the same profiles increase by a factor of 2 to 3 due to the scattering and motions of the grains . to determine the velocity dispersion of the particles responsible for the broadening , we smeared the stellar profiles of the fe i and ti i absorption lines with a gaussian . we used a range of values of @xmath44 from 0.5 to 1.6 and compared the smeared stellar lines with the observed profiles along the slits . the best fit varied slightly from line to line and of course for the different apertures , but for those profiles across the nw arc , the required smearing corresponds to a fwhm of 85 to 120 km s@xmath1 . the total velocity range will be even larger . the dusty scattering regions in this visible feature thus have a relatively high velocity dispersion in addition to their net motion with respect to the star . the mean velocities for the absorption lines are plotted in figure 7 with respect to the distance along each slit from the point nearest the star . they show a large redshift as the three parallel slits ( i , ii and iii ) cross the dusty arcs to the west and nw of the star , and then decrease in the outer nebulosity beyond the prominent nw arc . _ the profiles across the nw arc are significantly red - shifted with a velocity difference of @xmath41 + 50 km s@xmath1 relative to the star , suggesting a net motion of the expanding arc of this order . _ this difference is presumably the `` moving mirror '' redshift , @xmath38 , where u is the outward expansion of the reflecting material relative to the star and @xmath39 represents the direction of the flow relative to the plane of the sky ( @xmath45 for motion toward us ) . if the nw arc represents the material from a single ejection event we expect the expansion velocity to be proportional to distance along the arc . this expected trend is not immediately obvious in the observed velocities . it is probably blurred in our data because of the size of the extraction apertures , although some subtle evidence for this effect may be present in the velocities between slits ii and iii in apertures directly on the arc ; iii ap 4 is slightly further from the star and has a higher velocity than ii ap 5 . moreover , in figure 6 , one can see a velocity component on the short wavelength side of the lines at position ii ap 4 and on the long wavelength side at ii ap 6 . the velocity difference is consistent with the uniform expansion expected for the nw arc . unlike most other locations in the ejecta , appreciable k i _ absorption _ is observed in the extractions on and near the nw arc . it also shows an appreciable redshift at @xmath0 85 km s@xmath1 ( table 5 ) , and the line is also quite broad and asymmetric . unlike the other absorption lines , the k i velocity is approximately constant across the arc . this is probably due to the strong emission line covering or overlapping part of the absorption feature . the red wing of the k i absorption , however , does show a redshift as the slits cross the arc , from 135 to 151 km s@xmath1 along slit ii . using the velocities from the absorption lines at position ii ap 5 , directly on the nw arc , we apply its moving mirror " redshift of 54 km s@xmath1 to the k i velocity at this position , and get 31 km s@xmath1 , close to the 35 km s@xmath1 systemic velocity of the star . we therefore assume that the k i absorption in some sense represents the spectrum of the star as seen from the nw arc . in figure 8 we show the k i and ca i ] line profiles shifted by ( 35 + 54 ) km s@xmath1 , one of the ordinary absorption lines , and the k i emission at the star s position , i ap 2 . the k i and ca i ] absorption features are centered close to zero and are thus nearly in the rest frame of the star . the ti i absorption line is shifted an additional 25 km s@xmath1 in figure 8 , because as explained in 1 , it arises inside the circumstellar dust formation region . since the k i and ca i ] absorption features do not need the extra 25 km s@xmath1 shift , we conclude that they occur _ more than 100 au from the star , outside the ring or shell of dust . _ @xmath26573 is a semi - forbidden line , one would not have expected it to appear in absorption as observed . but the wavelength coincidence is essentially perfect , and we have found no other likely identification for the observed feature . thus we must conjecture that some atomic - physics trick makes the oscillator strength much larger than one would expect . ] the red wing of the k i absorption , at @xmath0 60 km s@xmath1 with respect to the absorption minimum , indicates the presence of a substantial dispersion within the expanding gas of 100 km s@xmath1 or more , about the same as observed for the other absorption lines at the same locations . this means that either the nw arc has a large disperison in its expansion velocity or a large extent along the line of sight , or both . note that the `` terminal velocity '' of the k i p cyg profile as seen at the star is also about 45 50 km s@xmath1 . thus , the moving - mirror redshift at the nw arc , the p cyg `` terminal velocity '' , the red wing of the k i absorption , and the reflected line widths _ all suggest flow speeds of about 50 km s@xmath1_. expansion velocities greater than 35 km s@xmath1 are required by the absorption line widths and there is no indication of velocities much greater than 70 km s@xmath1 relative to the star . an expansion speed of 50 km s@xmath1 implies that @xmath39 must be near zero so the nw arc is basically moving across our line of sight . this is not a surprising result given its orientation and appearance within the nebula . furthermore , the limits on the expected expansion velocity discussed above , 35 to 70 km s@xmath1 , would restrict @xmath39 to between -25@xmath10 and + 17@xmath10 . assuming mostly transverse motion for the arc , at its distance from the star of about 3@xmath6 or 4500 au at 1.5 kpc from the sun , the material in the _ nw arc was ejected about 400 years ago _ ( -100 , + 200 yrs . ) . we get a somewhat different result for slit i which crosses the nw arc at a section where it appears to be bending back toward the star . at i ap 4 the doppler velocities of the absorption lines are slightly lower and the moving mirror " redshift relative to the star is 40 km s@xmath1 . assuming 50 km s@xmath1 for @xmath37 , @xmath39 is roughly @xmath46 for this position . thus , this section of the nw arc may actually be the nearer side with a small component of motion toward us . the corresponding k i emission , positions ii ap 5 , ii ap 6 , iii ap 4 and iii ap 5 , has two definite velocity peaks ( figure 3 ) which can also be seen as two separate emission features in the image of slit iii in figure 9 . the shorter wavelength one presents some problems that are discussed in the next section . the red - shifted component with velocities of 60 to 70 km s@xmath1 exhibits a small velocity gradient across the arc as one would expect for an ejection . recognizing this trend , we can identify a corresponding bump in the line profiles at ii ap 4 and iii ap 3 at @xmath41 52 km s@xmath1 that probably corresponds to the same gas . since this feature spatially coincides with the nw arc and is expanding , we presume that this is the associated gas . however , the doppler velocity , at ii ap 5 , for example , corresponds to a shift of 25 km s@xmath1 relative to the star , half that for the absorption lines . the slope of the velocity gradient , 5 km s@xmath1 per 1000 au yields an expansion age of 950 years . if this emission is due to resonant scattering , then it is moving away from us at an angle of @xmath41 30@xmath10 which sems unlikely given the results from the absorption lines in the nw arc . if , on the other hand , it is reflected emission , it will be directed toward us at @xmath41 30@xmath10 since the redshift is smaller than for the absorption lines . the expansion age for the emitting gas is then about 550 years , comparable to the above results for the nw arc . slit positions iii ap 3 and ii ap 4 cross a filamentary , somewhat twisted arc just west of the star , between it and the nw arc . if we apply the same procedure described above for the nw arc with the k i and ca i ] absorption lines , we find that the moving mirror velocity for the dust is 45 km s@xmath1 and the flow speed is 55 km s@xmath1 with @xmath39 of @xmath47 . at its distance from the star , this material was ejected only 200 years ago . in contrast with the results for the ejecta to the west and nw of the star , the doppler velocities of the reflected absorption lines along slit v are nearly constant ( table 5 ) , show no gradient with position ( figure 7 ) , and no significant velocity shift with respect to the velocity at the star . this suggests that the reflecting material is basically stationary relative to the star . this and other evidence for surprisingly stationary gas and dust in the circumstellar nebula is discussed in 4.3 . in the next section we discuss the significant velocity shifts in the scattered light from some of the k i emission lines in the ejecta and additional evidence for ejection of discrete features in the nebula . the k i emission line often shows a complex profile with multiple peaks that vary with position in the ejecta . if the k i emission lines in the ejecta are due to resonant scattering , the doppler shifted and multiple peaks observed at different places must be due to gas moving at these different velocities thus permitting us to determine the expansion of the nebula and motions within it . however reflection by the dust can also contribute to the observed profiles . in this case we would expect to see either a single profile like that on the star , or an emission peak , with a velocity shift between the emission velocity at the star and the absorption lines at the same position . a few of the k i profiles closest to the star may be due to reflection or have a contribution from reflection . these appear to be either single or lack obvious multiple peaks even though the lines are quite broad . some possible examples are the k i emission at i ap 1 , v aps . 1 and 2 and iii ap 2 in figures 3 and 11 . for the multi - peaked profiles further from the star , it is difficult to determine if the emission is due to resonant scattering or reflection . both processes are probably occurring . in figure 10 we show the variation of the emission line velocities along the different slits as we did in figure 7 for the absorption lines . slit v is shown separately . along slits i , ii and iii we see a pattern to the west and northwest of the star in which most of the k i profiles have an emission peak at a velocity near 35 km s@xmath1 ( see table 5 ) . this is very apparent in figure 9 as the strong , straight emission feature in the two - dimensional image of slit iii . whether these emission peaks and single profiles are due to reflection or to resonant scattering , their velocities are consistent with stationary gas , near zero velocity with respect to the star . this possibility is discussed more fully in 4.3 . a second , somewhat weaker emission peak present in extractions on the nw arc was discussed in the preceding section . the changing profile and multiple peaks of the k i emission line along slit v ( figure 11 ) show the most dramatic shifts in doppler velocity with position in the nebula , and if these are due to resonant scattering in the gas , imply the existence of clouds or filaments of gas with a wide range of velocities in the ejecta . the k i line varies from a single profile in ap 1 with a strong p cygni absorption , similar to that viewed on the star , to multi - peaked profiles with large velocity shifts where the slit crosses knots and filaments in the outer parts of the nebula . the k i velocity to the north and east of the star , v ap 1 and ap 2 , is consistent with that observed on the star and may simply be the reflected profile from the star as we mentioned earlier . in apertures south of the star across the dusty knots and arcs , one of the strong emission peaks in apertures 3 to 6 has a velocity like that of the reflected absorption lines ( 60 65 km s@xmath1 ) . this emission component may be reflected by the dust or due to resonant scattering by gas moving @xmath41 30 km s@xmath1 relative to the star . likewise , the blueward peak near 25 km s@xmath1 in the same apertures , may be either reflected or scattered emission and represents a separate flow of emitting gas along the same line of sight . the most striking change occurs across the two outer filamentary arcs ( apertures 5 to 7 ) with the appearance of a strong blue - shifted emission feature . the sharp transition from an emission peak near 20 km s@xmath1 in aperture 5 on arc 2 to the blue - shifted emission across arc 1 at @xmath04 km s@xmath1 , suggests that this is resonant scattering in kinematically and spatially separate gaseous filaments or gas flows moving very differently from the reflecting dust . this shift can be clearly seen in the two - dimensional image in figure 12 where narrow emission appears in the p cyg absorption core and broadens out across arc 1 . the p cyg absorption is either no longer present or has been blocked by the strong blue - shifted emission . this is not a continuous outflow from a single ejection as suggested by smith ( 2004 ) , because these two emission features are resolved in the image of the slit . the velocities of the emission peaks associated with these three features are shown connected by a lines in figure 10 . we can trace this prominent blue - shifted emission feature in the two dimensional image over @xmath0 3@xmath6 in the ejecta and directly across arc 1 in aperture 7 . it corresponds to the strong emission peaks in apertures 6 and 7 and may also be weakly present in ap 8 and possibly in ap 5 as a small bump on the blue side of the profile . in ap 6 broad , weaker emission scattered or reflected in gas at a range of velocities up to 60 km s@xmath1 is also present , but is not observed in ap 7 across the arc . we therefore identify this strong blue - shifted emission feature with resonant scattering by gas in a stream or flow associated with the material in arc 1 . the doppler velocities at 4 8 km s@xmath1 indicate that this material is moving relative to the star at @xmath41 30 km s@xmath1 toward the observer . this of course contrasts sharply with the nearly constant 60 km s@xmath1 velocity of the absorption lines , and suggests a significant kinematic difference between the gas in the arc and the surrounding circumstellar material similar to what we found for the nw arc . although arc 1 has a significant velocity toward us relative to the star , its motion across our line of sight is not known . if we adopt 30 35 km s@xmath1 from the width of the broadened absorption lines ( see 4.3 ) at the same positions , as representative of the motions in the surrounding material for the transverse velocity , then the total motion relative to the star may be as much as 45 km s@xmath1 . the material associated with arc 1 would then have been ejected @xmath41 1000 years ago assuming it is moving in the plane of the sky . the two prominent emission peaks at 22 and 65 km s@xmath1 across arc 2 , in aperture 5 , are well separated kinematically in the two dimensional image of the slit , but a spatial separation or identification with distinct features is not obvious in the images at these positions . the 65 km s@xmath1 feature agrees with the velocity of the absorption lines in the same aperture , so we assume that this gas is associated with the reflecting dust , and since it is more red - shifted that it comes from background material . the second emission peak may then be due to reflection or resonant scattering by gas in arc 2 . this implies a motion of @xmath0 13 km s@xmath1 relative to the star for the emitting gas along the line of sight . the corresponding motion of the arc across our line of sight is not known and may be different from the assumed 30 35 km s@xmath1 , but with this motion , arc 2 is moving at @xmath41 37 km s@xmath1 and was ejected about 800 years ago . given the uncertainty in these estimates , it is possible that arcs 1 and 2 were ejected at about the same time . however the slit image shows that they are kinematically separate emission features which are also spatially separate in the wfpc2 image . therefore assuming comparable outflow velocities relative to the star , arc 1 would have been ejected earlier , but the ejection times also depend on the projection angle for these structures which is not known . other distinct features in the ejecta such as the two clumps of dusty knots , s and sw , are much closer to the star and presumably represent material ejected much more recently . the evidence for localized mass loss events and stellar activity are discussed in 5 . the presence of gas and dust throughout the nebula , apparently at the same velocity of the star , presents a serious problem because it implies that much of the circumstellar material is nearly stationary or moving very slowly relative to the star . this certainly conflicts with what one expects for a red supergiant with a high mass loss rate . in the ejecta to the west and nw of the star , the k i emission line has a velocity component consistent with relatively motionless gas , at the systemic velocity of the star , extending over @xmath0 6@xmath6 or @xmath41 9000 au in the nebula ( figure 10 ) . if this emission is due to reflection by dust , we should see a shift in velocity with respect to the rest frame of the star even if the motion is transverse from a uniform flow of gas outwards . this is not observed . if it is resonant scattered emission , and if the emission comes from widespread transparent gas , we would expect to see a broad profile centered at the average velocity of the system . however , much of the nebula where this k i emission is observed is not transparent and , more importantly , the emission components are quite narrow . the fwhm s range from 0.25 to 0.4 or only 10 15 km s@xmath1 wide , compared to @xmath41 1 for the single k i profiles on and near the star . this is much less than the expected expansion speed of the ejecta and suggests that there is very little doppler broadening in the lines . _ the emitting gas is therefore either not moving relative to the star , or all of the material must be moving in the plane of the sky _ i.e. in a narrow plane viewed perpendicularly . the latter seems highly unlikely over such a large area . furthermore , some of the profiles show two emission peaks , one near 35 km s@xmath1 and the other at 41 km s@xmath1 , the velocity at the star , suggesting that the emission is due to both reflection and resonant scattering . the narrow k i emission lines in the outer ejecta also have very broad emission wings that extend to 100 200 km s @xmath1 from the center of the emission line ( figure 13 ) . they are formed too far from the star to be electron scattering wings . it is possible that these emission wings are from more distant gas and dust , expanding outwards from vy cma perhaps in a more spherical distribution from a much earlier unstable period in the star s evolution . there is also some evidence for constant velocities , or zero motion relative to the star , in the reflected light to the northwest of the star . after crossing the near side of the nw arc , the absorption line velocities measured in the outermost apertures along slit i have a velocity like that observed at the star ( see figure 7 ) . along slits ii and iii , however , the absorption line velocities decrease , but stay red - shifted relative to the star by @xmath0 20 km s@xmath1 perhaps because the material , in the same plane , is being pushed by the expanding arc . the ca i ] emission has a nearly constant velocity at 40 41 km s@xmath1 throughout the ejecta . since this is the same as its velocity on the star , we assume that this is reflected emission . the cai ] emission profile does not show the variability with position and multiple peaks readily apparent in the k i emission line . as mentioned in the appendix , the presence of this semi - forbidden line in both emission and absorption is not understood . along slit v we observe absorption lines , including ca i ] , and some k i emission with nearly constant velocities extending over much of the length of the slit and across the nebula . all of these indicators have apparent doppler velocities near 60 65 km s@xmath1 which is also the red - shifted velocity of the absorption lines scattered by the dust envelope around the star . for the absorption lines , this implies that the reflecting dust is not only barely moving but also has little motion with respect to the star . unlike resonant scattered emission , reflected absorption lines are expected to show a redshift relative to the star even if the reflecting dust is moving perpendicular to our line of sight . admittedly , the problem is reduced if the material is moving toward us ; in that case , the `` moving - mirror '' effect is less , but there should still be a measurable redshift . two selection effects , forward scattering by the dust plus extinction , can also serve to reduce the measured redshft because we preferentially see the side of the ejecta toward us . but these considerations do not explain the near constancy of the velocities with distance from the star and the lack of any additional shift in the apertures across arcs 1 and 2 , which are quite red and dusty in the images . the absorption line profiles along slit v are also broadened compared to those on the star , although not as much as across the nw arc ( see figure 6 ) . we compared the gaussian smeared stellar profile with the observed profiles and obtained a range of results which varied with position along the slit . the reflected profiles from apertures 2 and 3 closest to the star , showed only marginal broadening with respect to the star , even though aperture 3 crosses the relatively bright dusty knot or condensation just south of the star . the smeared profiles that best matched those observed from aperture 4 , which crosses several small arcs , have fwhm s corresponding to 20 - 40 km s@xmath1 , and for apertures 5 through 7 across the outer filamentary arcs 1 and 2 , the fwhm s are 65 km s@xmath1 . although this is smaller than measured for the nw arc , it represents significant motions along the line of sight of @xmath48 30 35 km s@xmath1 for the material within these arcs , in contrast with the apparent lack of motion with respect to the star . consequently , we would expect to see a redshift relative to the star of at least 10 km s@xmath1 or so , especially across the arcs even with the above considerations for a reduced redshift . instead , the motion relative to the star is at most @xmath41 5 km s@xmath1 assuming that the material along slit v sees the same reflected and red - shifted spectrum of the star that we see , i ap 2 . it is possible that at different positions in the nebula , we may be observing radiation that escaped from different regions on the star , perhaps through holes or gaps in the dusty envelope . consequently , the gas and dust at these locations may be getting a different view or spectrum of the star . for example , if there is a ne - sw axis of symmetry with a bipolar flow , as has been suggested , then along slit v we may be observing gas and dust that views the star from near the polar direction which would be less obscured if there is a disk or flattened dusty envelope around the star . if gas is escaping through a low density region in the dust envelope , then the reference velocity for the reflected absorption lines would more likely be closer to the systemic velocity than the red - shifted velocity observed at the star . if this is the case , the reflected material may be moving at 30 km s@xmath1 relative to the star . this is just a hypothesis , but it might solve the problem of the zero - velocity gas and dust along slit v. however , it would also require that along most of slit v , we see the star from exactly the same direction or angle which then poses problems for the assumed geometry . a person unfamiliar with the nature of vy cma would conclude from the above description that a substantial amount of nearly stationary or very slow moving gas and dust exists in its ejecta . this is obviously a paradox given the star s strong wind and probable mass loss history . at this time we have no satisfactory explanation . it is possible that the general flow of the diffuse gas in the extended circumstellar material is extremely slow , but the escape velocity for vy cma is @xmath41 60 km s@xmath1 at its photosphere and 20 km s@xmath1 at the dust formation radius at 100 a.u . conceivably this material has been slowed by some process or it did not have quite the escape velocity and is now falling back . but there is no evident reason for this to happen , other than it would help solve our paradox . the results of our high resolution spectroscopy demonstrate that the three primary arcs in the visual images of vy cma are kinematically distinct from the surrounding stationary circumstellar material . the nw arc and arcs 1 and 2 have expansion speeds of 50 km s@xmath1 to @xmath41 40 km s@xmath1 , respectively . the gas and dust associated with these structures was thus ejected in specific mass loss episodes ; @xmath41 400 years ago for the nw arc and possibly 1000 to 800 years ago for arcs 1 and 2 . for the `` western arc '' we found an expansion age of only slightly more than 200 years . thus we have evidence for a period of considerable activity during the past 1000 years , including some recent ejections . in addition to the arcs , numerous clumps of small knots or condensations are observed in the images relatively close to the star . the dusty clumps to the south(s ) and southwest(sw ) of the star are the brightest , but several very small knots are observed just to the west and northwest of the star . these are clustered quite close , within 0@xmath95 of the star , and must have been ejected quite recently , only @xmath41 70 years years ago assuming an outflow speed of 50 km s@xmath1 . with the possibility of ejection episodes this recently , there may be some hint of activity in the star s observational record . numerous ground - based observers in the past have often described the appearance of knots of nebulosity that come and go and may vary in position , but the timescales for the motions and variablility seem too short to be related to the expanding ejecta . ( see smith et al 2001 for a discussion and the references ) these variations could be due to changing illumination of the dusty ejecta by radiation escaping from gaps or low density regions in the envelope @xcite . vy cma s historical light curve @xcite shows some very intriguing variability . the most significant change is its decline from m@xmath49 @xmath41 6.5 mag . beginning in 1872 to @xmath41 8 mag . by 1880 with some brief brightening episodes superposed . whether this 1.5 mag decrease was due to a change in bolometric correction or to the formation of dust , is not known , but it may correspond to the appearance of several knots and small arcs about 100 to 200 years ago . the star apparently then slowly faded to about 8.5 9th magnitude , and although the record is spotty , it may have had periods of greater variability in the 1910 s and again in the 1930 s . the more modern record based on the aavso light curve and published photometry , shows that vy cma displayed considerable variability of 1 to 2 magnitudes from about 1984 to 1995 with oscillations lasting several 100 to a thousand days that resemble the pulsations in irregular variables . since then it has been relatively steady at about 8.5 mag . the nw arc is the most clearly defined structure or mass concentration within the ejecta and is therefore the best candidate for a mass estimate . in our imaging paper , we estimated its total mass at @xmath50 m@xmath51 from the surface brightness in a well resolved 2 arcsec@xmath52 section . in this estimate we assumed an optical depth of unity . this assumption plus the evidence from the velocity gradient across the arc that more material may be participating in the outflow than is apparent in the images , suggests that this may be an underestimate for the arc s total mass . the dynamical timescale for vy cma is on the order of three years as are the episodes of large variability in its light curve . so whether the observed variability and the ejection events are due to non radial pulsations or large scale convection and activity on the star , the timescales will be comparable . therefore the short term mass loss rate associated with the ejection of the nw arc was at least 10@xmath53 m@xmath51 per year . the corresponding kinetic energy of the nw arc is @xmath41 10@xmath54 ergs which is not unrealistic for a star like vy cma with a total energy output of 6@xmath55 ergs per year . given the evidence we have for a relatively stationary nebula , most of vy cma s mass loss may be occurring in these separate high mass loss events perhaps due to activity on its surface . if we associate the ejections and surface activity with vy cma s photometric variability then they may occur approximately every 50 years or so . the nw arc may be an especially massive outflow , but if we assume @xmath0 10@xmath53 m@xmath51 per ejection , then over a thousand years the mass loss rate would average @xmath56 m@xmath51 per year which is close to what we estimated in smith et al ( 2001 ) for the entire circumstellar nebula and to previous independent mass loss measurements @xcite . the random orientations of the three primary arcs relative to the star together with their different ages , suggest that they were ejected not only at different times , but by localized events or disturbances at different positions on the star . our results show that the nw arc is moving primarily across our line of sight with little projection with respect to the plane of the sky . vy cma s geometry is unknown , but if we assume the ne - sw rotation / symmetry axis from the maser model , then the nw arc is very likely within 10@xmath10 to 15@xmath10 of the expected equatorial plane . given the relative positions of arcs 1 and 2 within the nebula and with respect to the presumed axis , each has likely been ejected from different locations on the star , at higher latitudes relative to the nw arc , and possibly from the polar region , but that depends on the unknown projection angle . one of the uncertainties in the above discussion is our lack of information on the transverse motions of the ejecta and the geometry of the nebula . we will soon be obtaining second epoch images and imaging polarimetry of vy cma with hst to measure its expansion and map the morphology and relative distribution of the embedded structures . our observations confirm that vy cma has experienced several mass loss events , presumably from localized active regions . these clouds or loops of gas and dust are now expanding outward into what appears to be a surrounding nebula of surprisingly stationary gas and dust . this stationary or slow moving material material is difficult to explain around a star that has had a long history of high mass loss . we can only speculate that the more diffuse , uniformly distributed gas has been created by a continuous , but slow general outflow of material from the overall surface . it therefore seems likely that the high mass loss for rate for vy cma is really a measure of the mass carried out by these specific ejections perhaps accompanied by streams or flows of emitting gas through holes or gaps in the dust envelope . vy cma is the only m type supergiant known to have a large , visible reflection nebula . schuster & humphreys ( 2005 ) have obtained hst / wfpc2 images of several cool hypergiant stars including the oh / ir m supergiants nml cyg , s per and vx sgr to search for associated nebulosity . the results are basically negative or yield only marginal detections , except for nml cyg which has a small bean - shaped nebula that has probably been shaped by the strong ionizing winds from the nearby hot stars in the cygnus x complex @xcite . in many ways vy cma is most like irc + 10420 , another high mass loss , strong oh / ir source believed to be in a post red supergiant stage and with a complex circumstellar environment @xcite . its extended reflection nebula has a variety of structures including knots and condensations , ray - like features and numerous arcs and scalloped loops which in the outer parts closely resemble the arcs observed in vy cma s ejecta . spectroscopy of its ejecta though showed a uniformly expanding outflow of gas @xcite ; there was no evidence for stationary gas or dust in its ejecta . the referee has suggested that the slow moving gas is due to gravitationally bound material in circular orbits ( jura and kahane 1999 ) . this may be the case but the examples of orbiting material about evolved stars come from a very different class of objects ; carbon - rich red giants and post agb stars which are binaries or presumed binaries ( jura , bahm , and kahane 1995 , bujarrabel et al . 2003 ) with `` reservoirs '' of long lasting molecular material . at the distances where the motionless k i emission is observed ( 4000 - 10000 au from the star ) , the orbital period will be several hundred thousand to a million years , exceeding the expected duration of the red supergiant stage for a star this luminous . furthermore , the oh and h@xmath8o maser observations show that the clouds are expanding at 30 40 km s@xmath1 , greater than the expected escape velocity from the dust shell . therefore we think this explanation is less likely for vy cma if it is an evolved red supergiant . vy cma is a very perplexing star experiencing discrete mass loss episodes that may be the primary mechanism by which it is losing so much mass . we have also uncovered some puzzling questions about the origin of its extended diffuse nebula with the slow moving gas and dust . if it is an evolved massive star , it is apparently in a unique stage , with no obvious counterparts . alternatively , the apparently `` stationary '' ejecta with embedded streams of gas and massive outflows is also consistent with a pre - main sequence state for vy cma . we are especially grateful to george herbig for his continuing interest in vy cma , his comments and advice on this study , and for obtaining the telescope time for these observations . it is also a pleasure to thank michael koppelman for assistance with the data reduction . this work was supported in part by the university of minnesota . the k i emission line is difficult to assess , because this atom has an extremely small ionization potential ( 4.34 ev ) and because some critical parameters are inaccessible for vy cma . since k i @xmath577665,7699 is a low - excitation resonance doublet , one is tempted to think `` it s merely resonance scattering '' which , however , does not answer the most important questions . note , for instance , the strength of @xmath27699 emission relative to the continuum when we look directly toward the star , position i ap 2 in fig . 3 . in order to achieve the observed equivalent width of @xmath41 5 in emission simply by resonance scattering , a substantial amount of radiation must be removed from the continuum along some ( indeed most ) lines of sight . on average , _ resonance scattering in itself neither augments nor diminishes the net emergent flux of @xmath27699 photons , _ excepting the small fraction that are scattered back into the photosphere . however , there is no hint of sufficient `` absorption '' to balance the observed `` emission '' anywhere along our slit positions , including reflected spectra that represent the star as seen from various directions in space . other puzzles involve the apparent velocity and the line profiles seen in the extended halo , 4.2 above . as a preliminary detail , note the optical thickness for k i @xmath27699 . ( @xmath27665 has a larger oscillator strength but is not included in our data . ) assuming a normal potassium abundance @xmath58 and a local line width of 4 km s@xmath1 , the line - center opacity is @xmath59 @xmath27 g@xmath1 . at least in the dusty zones , compare this to the continuum opacity for scattering by dust grains , @xmath60 @xmath27 g@xmath1 at the same wavelength . if k i resonance scattering is to be at least comparable to reflection by dust , we require @xmath61 , i.e. , @xmath62 . given potassium s low ionization potential , this is a stronger constraint than one might guess at first sight . can k i emission arise within @xmath63 au , before dust has formed ? first consider the @xmath64 ionization ratio , which , in a simple model , depends on the star s photon supply at energies above 4.3 ev and on the electron density in the wind . suppose that free electrons are provided by the elements with low ionization potentials , giving @xmath65 . at @xmath20 au , for example , in a simple model we expect @xmath66 @xmath22 . if the continuum spectrum of vy cma were a 3000 k planckian , then it would produce about 10@xmath67 potassium - ionizing photons per second ; but this is embarrassingly sensitive to temperature and other parameters , and the star s uv brightness may considerably exceed a simple black - body value ( compare @xmath11 ori , glassgold & huggins 1986 ) . anyway , using reasonable estimates of the photoionization cross - section and recombination coefficient , the values quoted above imply @xmath68 which applies throughout the inner wind , 12 au @xmath69 @xmath70 @xmath69 100 au . this ratio would be decreased by any additional supply of free electrons , e.g. , by some process that can ionize a tiny fraction of the hydrogen and/or an appreciable fraction of carbon ; but it is difficult to reduce the k@xmath71/k@xmath72 ratio by a factor much smaller than 0.2 in this way . most other likely complications tend to _ increase _ the ionization ratio , see glassgold & huggins ( 1986 ) . altogether , at this level of reasoning , one expects the line - center opacity in the inner wind to be rather small for a resonance line , @xmath73 @xmath27 g@xmath1 . the total line - center optical depth along a radial path may exceed 10 , but if so the effect on the total emergent spectrum is merely a fractional - wavelength redistribution of photons near 7699 i.e. , quite different from the observed strong emission feature which is not accompanied by a comparable amount of absorption at nearby wavelengths . if there are large - scale inhomogeneities in the local scattering coefficient @xmath74 , the emergent spectrum can depend on viewing direction ; a net surplus of 7699 photons may escape in some directions at the cost of a net deficit elsewhere . in order to obtain the observed strong feature in this way , however , one must view the star from a statistically unusual direction . a fundamentally non - spherical model in this vein for instance , imagine a circumstellar torus seen from a near - axis viewpoint is difficult to construct for the following reason : in order to obtain the large observed ratio ( @xmath27699 peak height)/(continuum level ) @xmath41 5 , the solid angle of `` resonance photon escape '' directions must be far smaller than that of the `` opaque '' directions . ( admittedly this may be the case south of the star , along slit v. ) in summary , this general type of scenario can not be rejected outright but we hope to find alternatives that are quantitatively more plausible . can the observed k i emission be excited by collisional processes in the inner wind ? in such a model the required luminosity in the @xmath27699 line is of the order of 150 @xmath75 or @xmath76 erg s@xmath1 . electron collisions are rather ineffective because @xmath77 is insufficient . if the inner - wind temperature is , say , 4000 k , and we adopt the relative electron density mentioned above , then collisional excitation of k i @xmath137665,7699 proves too weak by a factor of @xmath78 , i.e. , by 5 to 9 orders of magnitude . radiative recombination is too weak by about 8 orders of magnitude . unusual effects such as collisional excitation by fast non - thermal electrons , collisional excitation by hydrogen atoms , etc . , are likewise unpromising , though we omit details here . should we consider fluorescence , radiative excitation to higher levels of k i ? the upper level for @xmath137665,7699 is 4p @xmath79p@xmath80 , and the next resonance level , 5p @xmath79p@xmath80 , can be excited from the ground level by photons near 4044 . an ensuing cascade , 5p @xmath81 ( 3d or 5s ) @xmath81 4p , would lead to @xmath27699 emission . since the photosphere of vy cma is relatively faint at violet wavelengths , this process seems worthwhile only if it is pumped by some other strong emission line near 4044 . so far as we know , there is no likely candidate for this role , nor for pumping k i levels above 5p . in summary , we can almost eliminate all the obvious excitation processes other than resonance scattering , and even the latter does not seem promising in the inner wind . therefore let us focus on resonance scattering farther out , @xmath82 au , where inhomogeneous dust may help to solve the puzzle . once dust has begun to form near @xmath83 au in the outward mass flow , the k@xmath71/k@xmath72 ionization ratio becomes even more difficult to predict . the uv flux is greatly reduced by even a small column density of dust , while , on the other hand , some ion species that help provide @xmath77 become depleted . meanwhile the recombination time is several years , long enough for the gas to move a distance comparable to @xmath28 . since the data indicate that the k i column density is indeed substantial in dust farther out , let us _ assume _ that @xmath84 @xmath27 g@xmath1 , i.e. , at least comparable to scattering by dust , in the inner parts of the dusty region . this assumption is questionable but it is not implausible . figure 14 shows , in an idealized way , how resonance scattering can result in a net emission line if large - scale dusty condensations are present . suppose , for the sake of illustration , that + @xmath12 each localized dense blob in the figure contains enough dust to be practically opaque at all relevant wavelengths ; + @xmath12 most photons from the star encounter such an obstacle ; + @xmath12 along some escape paths not necessarily straight the optical thickness of the lower - density medium is less than 3 ; + @xmath12 the parameters allow about one or two percent of the red - wavelength continuum radiation to escape , as observed for vy cma . often the escape process involves an encounter with one of the dense blobs , scattering by a dust grain , and perhaps further scattering in the less - dense medium . the average dust - grain albedo is obviously relevant . in fig . 14 , a 7699 photon from the star encounters a dusty condensation , is resonance - scattered by a potassium atom , and then escapes after one more scattering event which may be due to either a k i atom or a dust grain . this type of behavior , averaged over all photons , leads to _ an apparent emission line in the emergent spectrum . _ there are at least two equivalent ways to see why : + ( 1 ) temporarily imagine , for simplicity , that the dust absorbs but does not reflect radiation . in that case , if the photon in fig . 14 had not been scattered by k i , then it would have been absorbed by dust in the dense condensation .. this answers a question we asked above , namely , why does the observed spectrum not show k i @xmath27699 absorption lines to balance the apparent emission ? the answer is that most @xmath27699 scattering events occur in locations where the continuum radiation would not otherwise escape anyway , being absorbed by dust if it is nt scattered by k i. + ( 2 ) of course , real dust grains reflect radiation as well as absorbing it . in effect , resonance scattering by k i enhances the overall albedo at a wavelength of 7699 . in other words , resonance scattering provides an extra way for 7699 photons to escape from a dusty condensation . once they escape into a less - dense channel , they have an appreciable probability of escaping from the configuration . ( moreover , k i scattering may be weak in the low - density regions , so it does not necessarily impede the final escape process . ) the main point is that the k i emission line may be `` formed '' by an indirect process at radii beyond 100 au or so . our observations indicate that expansion of the dusty medium does not red - shift the emergent k i emission line as much as it does the stellar absorption spectrum ( see 3 ) ; but a quantitative model is necessary to judge whether this is to be expected . unfortunately , a model of this type has many adjustable parameters and probably requires monte - carlo - style radiative transfer calculations , far beyond the scope of this paper . similar reasoning can probably be applied to the ca i ] emission . however , as we noted in 4 , the existence of ca i ] in both emission and absorption seems to require an oscillator strength far larger than one would expect for a semi - forbidden transition . if this statement is wrong due to some geometrical or radiative - transfer fallacy , we have not been able to identify the error . bernat , a. p. & lambert , d. l. 1976 , , 210 , 395 bowers , p. f. , johnston , k. j. , & spencer , j. h. 1983 , , 274 , 733 bujarrabel , v. , neri , r. , alcolea , j. & kahane , c. 2003 , , 409 , 573 danchi , w.c . , bester , m. , degiacomi , c.g . , greenhill , l.j . , & townes , c.h . 1994 , , 107 , 1469 de jager , c. 1998 , , 8 , 145 dupree , a.k . , lobel , a. & gilliland , r.l . 1999 , , 194 , 6605 efstathiou , a. & rowan - robinson , m. 1990 , , 245 , 275 . gilliland , r.l . & dupree , a.k . 1996 , , 463 , l29 glassgold , a. e. & huggins , p. j. 1986 , , 306 , 605 guilain , g. & mauron , n. 1996 , , 314 , 585 herbig , g.h . 1970a , mem . soc . liege , 19 , 13 herbig , g.h . 1970b , , 162 , 557 herbig , g.h . 1972 , , 172 , 375 herbig , g.h . 1974 , , 188 , 533 humphreys , r. m. 1970 , , 82 , 1158 humphreys , r. m. 1975 , , 87 , 433 humphreys , r.m . & davidson , k. 1994 , , 106 , 1025 humphreys , r.m . , et al . 1997 , , 114 , 2778 humphreys , r.m . , davidson , k. , & smith , n. 2002 , , 124 , 1026 hyland , a.r . , becklin , e.e . , neugebauer , g. & wallerstein , g. 1969 , , 15 8 , 619 jones , t. j. et al . 1993 , , 411 , 323 jura , m. & kahane , c. 1999 , , 521 , 302 jura , m. , balm , s. p. & kahane , c. 1995 , , 453 , 721 kluckers , v.a . , edmunds , m.g . , morris , r.h . & wooder , n. 1997 , , 284 , 711 lada , c.j . & reid , m.j . 1978 , , 219 , 95 lobel , a. & dupree , a. k. 2000 , , 545 , 454 lobel , a. & dupree , a. k. 2001 , , 558 , 815 marvel , k.b . 1997 , , 109 , 1286 monnier , j.d . , tuthill , p.g . , lopez , b. , cruzalebes , p. , danchi , w.c . , & haniff , c.a . 1999 , , 512 , 351 morris , m. & bowers , p. f. , 1980 , , 85 , 724 morris , m. & jura , m. 1983 , , 267 , 179 nercessian , e. , guilloteau , s. , omont , a. & benayoun , j. j. 1989 , , 210 , 225 richards , a. m. s. , yates , j. a. , & cohen , r. j. 1998 , , 299 , 319 robinson , l. j. 1971 , ibvs , 599 schaller , g. , schaerer , d. , meynet , g. & maeder , a. 1992 , , 96 , 269 schaerer , d. , meynet , g. , maeder , a. & schaller , g. 1993 , , 98 , 523 schuster , m. t. & humphreys , r. m. 2005 , in preparation smith , n. 2004 , , smith , n. , humphreys , r. m. , davidson , k. , gehrz , r. d. , and schuster , m. t. & krautter , j. 2001 , , 121 , 1111 uitenbroek , h. , dupree , a.k . , & gilliland , r.l . , 1998 , , 116 , 2501 wallerstein , g. 1958 , , 70 , 479 wallerstein , g. 1971 , , 169 , 195 wallerstein , g. 1977 , , 211 , 170 wallerstein , g. 1978 , observatory , 98 , 224 wallerstein , g. 1986 , , 164 , 101 wallerstein , g. & gonzalez , g. 2001 , , 113 , 954 wittkowski , m. , langer , n. & weigelt , g. 1998 , . , 340 , l39 lcccc i & 1 & 1.40 & 1.40 & 122 + & 2 & n.a . & 0.00 & 0 + & 3 & 1.75 & 1.75 & -58 + & 4 & 3.50 & 3.50 & -58 + & 5 & 5.00 & 5.00 & -58 + & 6 & 6.50 & 6.50 & -58 + & 7 & 8.00 & 8.00 & -58 + & 8 & 9.50 & 9.50 & -58 + ii & 1 & -2.80 & 3.23 & 145 + & 2 & -1.40 & 2.02 & 161 + & 3 & 0.00 & 1.28 & -156 + & 4 & 1.40 & 1.77 & -104 + & 5 & 2.80 & 2.93 & -84 + & 6 & 4.20 & 4.23 & -76 + & 7 & 5.60 & 5.58 & -71 + & 8 & 7.00 & 6.95 & -69 + & 9 & 8.40 & 8.34 & -67 + iii & 1 & -0.60 & 1.21 & 178 + & 2 & 0.60 & 1.12 & -121 + & 3 & 1.60 & 1.81 & -91 + & 4 & 3.00 & 3.13 & -76 + & 5 & 4.50 & 4.48 & -71 + & 6 & 5.90 & 5.58 & -68 + & 7 & 7.30 & 7.23 & -66 + & 8 & 8.70 & 8.62 & -65 + v & 1 & -2.10 & 2.28 & 54 + & 2 & -0.70 & 0.28 & 18 + & 3 & 0.70 & 0.81 & 177 + & 4 & 2.10 & 2.07 & -155 + & 5 & 3.50 & 3.44 & -149 + & 6 & 4.90 & 4.82 & -147 + & 7 & 6.30 & 6.22 & -145 + & 8 & 7.70 & 7.61 & -144 + & 9 & 8.10 & 9.01 & -143 + lcccccc 1 & 31.6 , 40.2 & 6.7 & 16.8 , 29.3 & & 25.4 , 60.4 & 6.3 + 2 & 41.0 & 7.8 & 25.0 & 6.3 & 27.3 , 44.5 & 8.2 + 3 & 23.4 , 45.6 , 52.6 & 10.2 & 25.8 , 42.5 & 6.3 & 26.9 , 41.7 & 8.2 + 4 & 30.8 , 56.5 & 10.2 & 42.1 & 8.2 & 34.3 , 60.4 & 13.3 + 5 & 33.9 & & 33.9 , 59.7 & 12.9 & 35.0 , 70.0 & 12.0 + 6 & 35.1 & & 34.3 , 67.8 & 12.5 & 34.0 , 41.0 & + 7 & 38.6 & & 33.5 , 41.0 & & 36.0 & + 8 & 35.1 & & 34.7 , 41.0 & & 32 , 39 & + 9 & & & 32.0 & & & + 1 & 41.5 & & 42.9 & 24.6 & 13.2 , 40.1 & + 2 & 41.9 & & 41.9 & 31.0 & 39.7 & + 3 & 41.0 & & 41.0 & & 41.0 & + 4 & & & 40.6 & & 38.8 & + 5 & & & 41.5 & & 39 & + 6 & & & & & 41 & + 7 & & & & & 42 & + 1 & -14 & & 34.0 & & -8.1 & + 2 & & & -7.6 & & -5.3 & + 3 & 10 & & -4.4 & & & + 4 & & & 8.4 & & 34.0 & + 5 & & & 35.3 & & 36 & + 6 & 37 & & 34.0 & & 36 & + 7 & 37 & & 37.2 & & 36 & + 8 & 37 & & 38.5 & & 36 & + 9 & & & 37.6 & & & + lcccccccc 1 & 63.2 & 9 & 63.2 & 9 & 69.8 & 8 & 62.6 & 9 + 2 & 59.9 & 9 & 69.0 & 8 & 75.9 & 8 & 59.3 & 9 + 3 & 76.7 & 8 & 73.6 & 8 & 107.9 & 5 & 64.7 & 9 + 4 & 100.4 & 8 & 101.8 & 4 & 116.0 & 8 & 64.4 & 8 + 5 & 76.7 & 5 & 113.6 & 8 & 102.3 & 7 & 64.6 & 8 + 6 & 68.1 & 3 & 107.4 & 6 & 82.4 & 8 & 64.8 & 8 + 7 & 58.5 & 2 & 85.9 & 7 & 78.0 & 6 & 62.7 & 8 + 8 & 56.5 & 1 & 78.9 & 6 & 81.3 & 3 & 64.2 & 4 + 9 & & & 83.9 & 3 & & & 59.5 & 3 + 1 & & & 63.4 & & 57.5 & & & + 2 & & & 58.8 & & 85.3 & & 64.8 & + 3 & 89.9 & & 84.4 & & 93.1 & & 57.9 & + 4 & 103.1 & & 92.6 & & 92.6 & & 59.7 & + 5 & 80.8 & & 91.2 & & 99.0 & & 62.9 & + 6 & & & 89.0 & & 81.0 & & 68.9 & + 7 & & & 79.8 & & & & 64.8 & + 8 & & & 76.2 & & & & & + 1 & & & & & & & 92.0 + 2 & & & & & 88.1 & & + 3 & & & 89.7 & & 86.5 & & + 4 & 86.9 & & 89.7 & & 84.6 & & + 5 & & & 84.6 & & 87.0 & & + 6 & & & 84.6 & & & & + 1 & 48 , 74 & & 76 & & 73.7 & & 92.0 & + 2 & 48 & & 73 & & 86.5 & & 71.9 & + 3 & 114 & & 76 & & 121.7 & & 73.7 & + 4 & 132 & & 118 & & 118.1 & & 68.2 & + 5 & & & 117 & & 126 & & 79.7 & + 6 & & & 119 & & 109 & & 77.8 & + 7 & & & 111 & & & & 68.7 & + lccccc 1 & 43.3 & 4.3 & 41.9 & 20.0 & -42.8 , -2.1 , 31.7 + 2 & 37.1 & 6.3 & 41.9 & 20.5 & -3.5 , 30.3 + 3 & 24.1 , 60.4 & 6.7 & 40.6 & & -3.5 + 4 & 28.1 , 52.6 , 67.1 & 7.8 & & & -14.5 + 5 & ( 8.6),21.9 , 65.5 & & 41.5 & & + 6 & 8.6 , 42.5 , 56.2 & & 40.6 & & 35.3 + 7 & 3.9 & & 38.3 & & 34.9 + 8 & ( 3.9 ) , 39.0 & & & & 35.8 + 9 & 40.2 & & & & 37.6 +

electronic transport through dna is crucial for its biological functions such as the repair mechanism after radiation damage and biosynthesis . as proposed in @xcite electron transport through dna proceeds along a one - dimensional pathway constituted by the overlap between @xmath1orbitals in neighboring base pairs . with the perspective to possible applications in molecular electronics biomaterials are supposed to play an important role because they may offer the ultimate way to miniaturization and their attributed electrical transport properties are expected to be markedly different from those of traditional macroscopic conductors @xcite . in this context enormous effort has been directed towards investigations of conductivity of dna viewed as a promising conduit for charge transport on nanoscales @xcite . so far the findings concerning conductivity of dna have been controversial . there are the results of recent measurements indicating that dna behaves as a well conducting one - dimensional molecular wire @xcite-@xcite . in contrast it was reported that dna is insulating @xcite and for short oligomers built up from base pairs of the same type semiconductivity was observed @xcite . this qualitative discrepancy is shared also by theoretical findings so that it remains unclear whether dna is conductive or not . for example a theoretical investigation based on first principle calculations suggests the absence of dc - conductivity in dna @xcite . on the other hand there stand the studies using various assumptions for modeling the dna structure which focus on different aspects such as the influence of aperiodicity , temperature driven fluctuations and aggregation effects on the energetic control of charge migration which have all come to the conclusion that dna is a conductor @xcite-@xcite . the detailed understanding of the conduction mechanism of dna remains still an unsolved task . different attempts to model the charge transport of dna were based on transport via coherent tunneling @xcite , classical diffusion under the conditions of temperature - driven fluctuations @xcite , incoherent phonon - assisted hopping @xcite,@xcite , variable range hopping between localized states @xcite and charge carriages mediated by polarons @xcite,@xcite and solitons @xcite . in the present work we use for the theoretical description of charge transport in dna a nonlinear approach facilitating the concept of polaron respectively breather solutions . the structure of the bent double helix of @xmath0dna is modeled by a sterical network of oscillators in the frame of the base pair picture @xcite,@xcite,@xcite taking into account deformations of the hydrogen bonds within a base pair and twist motions between adjacent base pairs . the electron motion is described by a tight - binding system . it is assumed that the electron motion is predominantly influenced by vibrational modes of the double helix . the nonlinear interaction between the electron and the vibrational modes causes the formation of polarons respectively electron - vibron breathers supporting charge localization and transport . the aim of the current investigation is to elucidate the role played by static parametrical disorder and complex structural effects due to irregular deviations from the ordered helical shape of the double helix ( reflected in random arrangements of the positions of the bases ) with regard to the charge transport in dna . there arises at least one interesting question , namely : is long - lived and stable charge transport in dna achievable at all under the impact of disorder ? to tackle this problem in our paper we exploit analytical and numerical methods from the theory of ( nonlinear ) dynamical systems in combination with the theory of ( linear ) disordered systems . the paper is organized as follows : in the first section we describe the model for the charge transport along the steric structure of the bent dna double helix . the second section is concerned with the construction of polaron respectively static electron - vibron breather solutions of the nonlinear lattice system when parametrical and structural disorder effects are taken into account . in section [ section : moving ] we discuss the mobility properties of the polarons respectively electron - vibron breathers and their transport efficiency in dependence on the degree of disorder contained in the on - site energies and/or in structural arrangements of the helix . in a subsequent section we invoke the floquet analysis to explain also the mobility of the polarons and breathers in terms of their spectral properties . finally , we present a summary and discussion of the results . dna chains are quite flexible and exhibit structural fluctuations influencing the transport properties . our model is designed to cover the basic features of the dna double helix structure needed for a proper description of the charge transport dynamics under the impact of fluctuating base motions . despite the fact that the charge transport proceeds across a one - dimensional channel of base stacks spanning along the sugar - phosphate backbone so that dna can effectively be viewed as a one - dimensional molecular wire @xcite some significant details of the three - dimensional dna structure have to be incorporated in the model . for the physical description of dna molecules their detailed chemical structure can be neglected . they can be simply considered as bent double - stranded systems with an adequate number of parameters sufficient for modeling the considered physical process . in the frame of simple dna models such as the base pair picture the bases have been treated as single nondeformable entities . the helicoidal structure of dna is then conveniently described in a cylindrical reference system where each base pair possesses two degrees of freedom , namely a radial variable measuring the transversal displacements of the base pair ( that is deformations of the h - bond ) and its angle with a reference axis in a plane perpendicular to the helix backbone which defines the twist of the helix @xcite . in fig . [ fig : sketch ] we depict schematically the structure of the dna model . , @xmath2 , @xmath3 and the radial and angular variables @xmath4 and @xmath5 , respectively are indicated . ] with relevance to charge migration the most influential fluctuating motions of dna originate from tranverse vibrations of the bases relative to each other within each base plane , i.e. the base pair distance stretchings / compressions and the torsional variations of the helicoidal twist dominating longitudinal ( acoustic strand ) phonons @xcite because the sugar - phosphate chains are rather stretch proof @xcite . therefore it is justified to discard the impact of longitudinal vibrations and the considered motion is restricted to the base pair planes @xcite . furthermore , the angular twist and radial vibrational motions evolve independently on two different time scales and can be regarded as decoupled degrees of freedom in the harmonic treatment of the normal mode vibrations of dna @xcite . the hamiltonian for the electron migration along a strand in dna consists of three parts @xmath6 with @xmath7 representing the part corresponding to the particle charge transport over the base pairs , @xmath8 describes the dynamics of the h - bond vibrations and @xmath9 models the dynamics of the relative twist angle between two consecutive base pairs . the electronic part is given by a tight - binding system @xmath10 the index @xmath11 denotes the site of the @xmath12th base on one of the two strands and @xmath13 determines the probability that the electron occupies this site . @xmath14 is the local electronic energy and @xmath15 stands for the transfer matrix element which is responsible for the transport of the electron along the stacked base pairs . the vibronic part @xmath8 takes into account radial displacements of the base units from their equilibrium positions within the base pair plane connected with deformations of the hydrogen bonds linking two bases . the corresponding part of the hamiltonian treating the bond vibrations classically and harmonically is given by @xmath16\,.\ ] ] with the momenta @xmath17 conjugate to the radial coordinates @xmath18 measuring the deviation from the equilibrium length of a hydrogen bond along the line bridging two bases of a base pair , @xmath19 is the frequency and @xmath20 is the reduced mass of a base pair . the twist motion part reads as @xmath21\,,\label{eq : htwist}\ ] ] where @xmath5 is the relative angle between two adjacent base pairs quantifying displacements from its equilibrium twist angle @xmath2 and @xmath22 is the reduced moment of inertia . we emphasize that in the context of charge transport in dna the nonlinear dynamics of large amplitude structural transitions are not of interest because charge transport processes are thought to be mainly facilitated by soft fluctuational vibrational modes allowing the harmonic approximation of their dynamics . the interaction between the electronic variable and the structure variables @xmath18 and @xmath5 stems from the parameter dependence of the electronic parameters @xmath14 and @xmath23 . the diagonal term expressing the most efficient coupling is of the form @xmath24 and takes into account the modulation of the on - site electronic energy @xmath25 by the radial vibrations of the base pairs . in turn the actual charge occupation has its impact on the local radial distortion of the helix . as the transfer matrix elements @xmath23 are concerned we assume that they depend on the three - dimensional distance between two consecutive bases along a strand in the following fashion @xmath26 the quantity @xmath27 regulates how strong @xmath23 is influenced by the distance and the latter is determined by @xmath28 with @xmath29 expanding the expression ( [ eq : dist ] ) up to first order around the equilibrium positions gives @xmath30\,.\label{eq : distance}\ ] ] realistic parameters for dna molecules are given by @xcite,@xcite : @xmath31 , @xmath32 , @xmath33 , @xmath34 , @xmath35\,\times10^{12}\,s^{-1}$ ] , @xmath36 , @xmath37 and @xmath38 . we scale the time according to @xmath39 and introduce the dimensionless quantities : @xmath40 subsequently we omit the tildes . with the use of the expression ( [ eq : distance ] ) the equations of motion derived from the hamiltonian ( [ eq : hel])-([eq : htwist ] ) read as @xmath41+[c_{n}^*c_{n-1}+c_{n}c_{n-1}^*]\,\right\}\label{eq : dotr}\\ \ddot{\theta}_{n\,n-1}&=&-\omega^2\,\theta_{n\,n-1}\,-\ , \alpha\,v\,\frac{r_0 ^ 2}{l_0}\ , \sin \theta_0\,[c_{n}^*\,c_{n-1}\,+\,c_{n}\,c_{n-1}^*]\ , , \label{eq : dottheta}\end{aligned}\ ] ] and the quantity @xmath42 appearing in eq . ( [ eq : dotc ] ) determines the time scale separation between the fast electron motion and the slow bond vibrations . we remark that in the limit case of @xmath43 and constant @xmath44 the set of coupled equations represents the holstein system widely used in studies of polaron dynamics in one - dimensional lattices @xcite . furthermore in the linear limit case emerging for @xmath45 the anderson model is obtained for random @xmath46 @xcite,@xcite . regarding the dynamical energy exchange between the various degrees of freedom we state that due to the large differences between the electronic and vibronic frequencies we are able to prove rigorously that the energy exchange is suppressed on time scales growing exponentially with the ratio of the frequencies . corresponding bounds on the energy exchange suppression are derived with the help of nekhoroshev - like arguments @xcite,@xcite . furthermore applying the concept of the anti - integrable limit @xcite we are able to prove the existence of breather solutions for the coupled system ( [ eq : dotc])-([eq : dottheta ] ) . the mathematical details are delegated to a forthcoming article @xcite . the values of the scaled parameters are given by @xmath47 , @xmath48\,\times 10^{-2}$ ] , @xmath49 , @xmath50 and @xmath51 . in our model study we treat the electron - mode coupling strengths @xmath52 and @xmath27 for which no reliable data are available as adjustable parameters . in the forthcoming investigations we adopt the values of these free parameters such that not too strong deformations of the helix result which is essential for our used harmonic treatment of the dynamics of the structural coordinates . being interested in the simulation of a nonlinear charge transport mechanism in dna we embark as a first step on the construction of localized stationary solutions of the coupled system ( [ eq : dotc])-([eq : dottheta ] ) . exploiting the fact that the adiabaticity parameter @xmath53 is small , i.e. the fast charge transport along the base stacks and the slow bond vibrations and the even slower rotational twist motions evolve on distinct time scales , the inertia in eqs . ( [ eq : dotr ] ) and ( [ eq : dottheta ] ) are negligible . therefore , one can solve in the adiabatic limit the resulting static equations which results in the instantaneous displacements @xmath54 + [ c_{n}^*c_{n-1}+c_{n}c_{n-1}^*]\,\right\}\ , , \label{eq : rinst}\\ \theta_{n\,n-1}&=&-\frac{\alpha\,v}{\omega^2}\ , \frac{r_0 ^ 2}{l_0}\,\sin \theta_0\ , [ c_n^*c_{n-1}+c_nc_{n-1}^ * ] \label{eq : thetainst}\,.\end{aligned}\ ] ] upon insertion of ( [ eq : rinst ] ) and ( [ eq : thetainst ] ) into eq . ( [ eq : dotc ] ) one obtains a nonlinear discrete schrdinger equation for the electronic amplitude @xmath55 + [ c_n^*c_{n-1}+c_n c_{n-1}^*]\,\right\}\,\right]\,c_n \nonumber\\ & -&(\,1+b_{n+1\,n}\,)\,c_{n+1}\ , -\,(\,1+b_{n\,n-1}\,)\,c_{n-1 } \label{eq : dnls}\,,\end{aligned}\ ] ] with @xmath56\right.\right.\nonumber\\ & + & \left.\alpha\,(1-\cos \theta_0)\ , \left(\,[c_{n+1}^*c_{n}+c_{n+1}c_n^*]\right.\right.\nonumber\\ & + & \left.\left.2\,[c_n^*c_{n-1}+c_n c_{n-1}^ * ] + [ c_{n-1}^*c_{n-2}+c_{n-1 } c_{n-2}^*]\,\right)\right]\nonumber\\ & + & \left . \frac{\alpha\,v}{\omega^2}\,r_{0}^2\,\sin^2\theta_0\ , [ c_n^*c_{n-1}+c_n c_{n}c_{n-1}^*]\right\}\end{aligned}\ ] ] in order to search for stationary localized solutions ( polaron states ) we substitute @xmath57 $ ] in eq . ( [ eq : dnls ] ) and obtain the difference system @xmath58 + [ \phi_n^*\phi_{n-1}+\phi_n \phi_{n-1}^*]\,\right\}\,\right ] \,\phi_n \nonumber\\ & -&(\,1+b_{n+1\,n}\,)\,\phi_{n+1}\ , -\,(\,1+b_{n\,n-1}\,)\,\phi_{n-1 } \ , , \label{eq : statthree}\end{aligned}\ ] ] where @xmath59 is given by @xmath60\right.\right.\nonumber\\ & + & \left.\alpha\,(1-\cos \theta_0)\ , \left(\,[\phi_{n+1}^*\phi_{n}+\phi_{n+1}\phi_n^ * ] \right.\right.\nonumber\\ & + & \left.\left.2\,[\phi_n^*\phi_{n-1}+\phi_n \phi_{n-1}^ * ] + [ \phi_{n-1}^*\phi_{n-2}+\phi_{n-1 } \phi_{n-2}^*]\,\right ) \right ] \nonumber\\ & + & \left . \frac{\alpha\,v}{\omega^2}\,r_{0}^2\,\sin^2\theta_0\ , [ \phi_n^*\phi_{n-1}+\phi_n \phi_{n-1}^*]\right\}\end{aligned}\ ] ] the ground state of the system ( [ eq : statthree ] ) is computed with the numerical map method outlined in @xcite,@xcite . . ( b ) the static radial displacements @xmath4 . ( c ) the static angular twists @xmath61 . ] in fig . [ fig : profile ] we depict typical patterns of polaron states . first we discuss the ordered ( periodic ) case arising e.g. for synthetically produced dna molecules consisting of a single type of base pairs ( e.g. poly(g)-poly(c ) dna polymers ) surrounded by vacuum which provides realistic ideal conditions for the consideration of dna conductivity in experimental and theoretical studies @xcite . the electronic contribution is shown in fig . [ fig : profile ] ( a ) . in the ordered case the electronic wave function is localized at the central lattice site and the amplitudes decay monotonically and exponentially with growing distance from this central site . we remark that the spatial extension of the polaron decreases with increased coupling strength @xmath52 and there is a smooth transition from large to small polarons . analogously , enlarging the off - diagonal coupling strength @xmath27 leads also to enhanced degree of localization . the size of the polarons plays a crucial role for their mobility in the sense that for large up to medium polarons coherent motion can be activated whereas small polarons are immobile due to their pinning to the discrete lattice @xcite,@xcite . the associated patterns of the static radial and angular displacements of the bases are shown in fig . [ fig : profile ] ( b ) and ( c ) , respectively . like the electronic wave function the radial and angular displacements are exponentially localized at the central lattice site ( base pair ) . notice that due to the overall minus sign of the radial and angular excitation patterns the h - bridges get compressed while the helix experiences a local unwinding around the occupation peak of the localized electron . directed along the corresponding pinning mode component . the kick amplitude is @xmath62 . ( a ) uniformly moving electronic breather . ( b ) the associated moving breather in the radial displacements . ( c ) the angular breather contribution consisting of a small amplitude breather pinned at the starting site and a relatively large amplitude moving breather component propagating in alliance with the electron and radial breathers . , title="fig : " ] directed along the corresponding pinning mode component . the kick amplitude is @xmath62 . ( a ) uniformly moving electronic breather . ( b ) the associated moving breather in the radial displacements . ( c ) the angular breather contribution consisting of a small amplitude breather pinned at the starting site and a relatively large amplitude moving breather component propagating in alliance with the electron and radial breathers . , title="fig : " ] directed along the corresponding pinning mode component . the kick amplitude is @xmath62 . ( a ) uniformly moving electronic breather . ( b ) the associated moving breather in the radial displacements . ( c ) the angular breather contribution consisting of a small amplitude breather pinned at the starting site and a relatively large amplitude moving breather component propagating in alliance with the electron and radial breathers . , title="fig : " ] for a more realistic study we take static diagonal disorder in the on - site electronic energy @xmath46 into account . the random values caused for example by the inhomogeneous broadening of the sites of distinct ion pairs with different energies were simulated by random potentials @xmath63 $ ] with mean value @xmath64 and different mean standard deviations @xmath65 . ( note that any @xmath66 term on the r.h.s . ( [ eq : dotc ] ) with constant @xmath67 can be eliminated by a gauge transformation @xmath68 . ) generally the localized excitation patterns do not change qualitatively when diagonal disorder is taken into account . compared with the ordered case one merely observes a shift of the excitation peak away from the central lattice site depending on the respective realization of the random potential . however , since in the presence of diagonal disorder the translational invariance of the lattice system is broken the localized amplitude pattern is no longer reflection symmetric with respect to a distinguished lattice site as opposed to the corresponding symmetric localized state of the ordered system . moreover we find that the combined effect of the two localization mechanisms , viz . nonlinear polaron and linear anderson - mode formation respectively , leads to enhancement of the degree of localization compared to the case when only one of the two mechanisms acts . furthermore , real dna molecules exhibit random structural imperfections of their double helix caused e.g. by the random base sequence , the deforming impact of the chemical surroundings when dna gets buffeted by water molecules and/or the varying hydrophobic potential of the base pair interactions depending on the ambient aqueous solvent that may leave the involved helix structures in irregularly distorted shapes . accordingly , the structural disorder is incorporated in our model by randomly distributed equilibrium positions of the bases so that the resulting irregular helical dna matrix deviates from the perfectly regular helix structure . to be precise , we consider randomly distributed torsional coordinates @xmath5 and radii @xmath4 . since the positions ( actually , the distances ) govern the value of the corresponding transfer matrix elements @xmath69 nondiagonal disorder is induced in the latter at the same time . random transfer integrals along the helix were used to represent random base pair sequences for which the charge transport along dna was discussed recently in a model relying on localized electron states @xcite . in this section we study the charge transport supported by mobile polarons respectively breathers propagating along dna . to activate polaron motion an established method is provided by proper stimulation of localized internal normal modes of certain shape and frequency which are also called pinning modes @xcite . the identification of possibly existing pinning modes is related to the linear stability analysis of the polarons . in order to investigate the stability of the polarons with respect to linear perturbations we derive the linearized system for which we impose small perturbations on the stationary polaron solution @xmath70 , @xmath71 in the form : @xmath72\,\exp(-i\,e t/\tau ) \label{eq : cperturb}\\ r_{n}(t)&=&r_{n}^{(0)}+ \delta r_{n}(t)\\ \theta_{n}(t)&=&\theta_{n}^{(0)}+ \delta \theta_{n}(t)\,,\label{eq : thetaperturb}\end{aligned}\ ] ] where @xmath73 , @xmath74 and @xmath75 are small time - dependent variables . substitution of ( [ eq : cperturb])-([eq : thetaperturb ] ) into eqs . ( [ eq : dotc])-([eq : dottheta ] ) and retaining only the terms linear in the perturbations results in the system of tangent equations . the stability analysis is then performed using floquet theory . to this end we integrate the linear system of the tangent equations over one period @xmath76 yielding the floquet map @xmath77 with the floquet matrix f determining the evolution of the initial set of variables @xmath78 after times @xmath79 . to ensure linear stability all eigenvalues @xmath80 of the floquet matrix @xmath81 have to lie on the unit circle , i.e. the floquet arguments @xmath82 have to be real - valued . intensive numerical investigations in a wide parameter range have proved that all the polarons gained from the map method are linearly stable . moreover the floquet analysis supplies the frequencies of the normal modes via the relation @xmath83 $ ] with @xmath84 expressed in @xmath85 and @xmath86 is an arbitrary integer number . we identified the pinning modes for given sets of parameters . according to @xcite we initiate the motion of polarons through suitable perturbations of the velocity variables @xmath87 and/or @xmath88 targeted in the direction of the pinning mode , that is we use for the numerical integration of the system ( [ eq : dotc])-([eq : dottheta ] ) the following initial conditions @xmath89 with the normalized momentum parts @xmath90 of the pinning mode and @xmath91 are the perturbation strengths ( for details see @xcite ) . in fig . [ fig : fig3 ] we present results for the activation of breather motion along dna in the ordered case . we integrated the set of coupled nonlinear equations ( [ eq : dotc])-([eq : dottheta ] ) with a fourth - order runge - kutta method and the accuracy of the computation was checked through monitoring the conservation of the total energy as well as the norm @xmath92 . the applied kicking strength is @xmath93 corresponding to an amount of radial kinetic energy of the order of @xmath94 . interestingly , we found that there exists no angular pinning mode component @xmath95 ( see further below ) so that alone the excitation of the radial component @xmath96 produces long - lived stable polaron and breather motions . hence , our findings suggest that higher frequency radial fluctuational modes of dna instigate the propagation of localized structures along dna rather than low frequency twist fluctuational modes . the fig . [ fig : fig3 ] ( a ) demonstrates that the electronic component of the polaron moves with uniform velocity along the lattice maintaining its localized shape throughout the journey . in the same manner the radial polaron part travels as a localized wave along the strand so that the electron propagation is accompanied by a local and temporal contraction of the radii . however , the angular static polaron component splits up into two parts when the radial pinning mode is initially imposed . at the starting position there remains a standing breather of comparatively large amplitudes which reverse periodically their sign in the course of time . thus the helix experiences alternately a local winding and unwinding around its central base pair . in addition a second breather of relatively small and exclusively negative amplitude propagates in unison with the electronic and radial polaron components . the result is that the region of the helix which is traversed by the electron gets locally untwisted . as the quantitative assessment of the charge transport is concerned we note that the breather travels across seventy sites ( base pairs ) in a thousand ( dimensionless ) time units meaning that physically the charge conduction across the base pairs of dna over a distance of @xmath97 takes @xmath98 leading to a higher transport rate than a typical one found for protein conductivity @xcite . this underlines the status ascribed to dna molecules as promising candidates establishing fast and efficient charge transport in ( albeit only theorized yet ) molecular electronical engineering utilizing them as molecular wires . in the presence of diagonal static disorder , that is when the on - site electronic energies @xmath46 are represented by random numbers distributed in the interval @xmath99 with width @xmath65 , we find that mobile breathers exist up to a critical degree of disorder @xmath100 and conductivity persists . ( let us stress that it was indeed experimentally found that the random sequence @xmath0dna is an electrical conductor @xcite . ) beyond this critical value continuous propagation of the breathers along the lattice is prevented . this behavior is illustrated in fig . [ fig : fig4 ] displaying the spatio - temporal evolution of the electronic amplitude @xmath101 . while for a relatively small amount of disorder @xmath102 the electronic breather ( together with its radial and torsional counterparts ) is able to travel uniformly along the lattice for a ten times increased @xmath103 the breather motion is restricted to a neighborhood of the initially excited position around which it wanders in an oscillatory manner . of the random distribution of the on - site electronic energy @xmath104 . ( a ) width @xmath102 . the electron breather moves with uniform velocity along a strand . ( b ) width @xmath103 . confined electron breather with positions oscillating itinerantly around the initial lattice position . , title="fig : " ] of the random distribution of the on - site electronic energy @xmath104 . ( a ) width @xmath102 . the electron breather moves with uniform velocity along a strand . ( b ) width @xmath103 . confined electron breather with positions oscillating itinerantly around the initial lattice position . , title="fig : " ] regarding mobility there act two competing mechanisms in the combined disordered nonlinear system , namely a linear anderson - type one related with the effects of disorder which attempts to pin the modes ( bear in mind that the linear anderson - modes are supposed to be immobile ) and opposed to this there operates the nonlinear mechanism utilizing the pinning mode to activate motion . the relative strength of these two mechanisms @xmath105 determines then breather mobility . in order to summarize the results for the breather mobility in the presence of diagonal disorder we display in [ fig : fig5 ] the temporal behavior of the first momentum of the electronic occupation probability defined as @xmath106 ) , dashed - dotted line ( @xmath107 ) , dotted line ( @xmath108 ) and long dashed line ( @xmath109 ) . ] which determines the position of the breather center as a function of time . the breathers are mobile and propagate unidirectionally with ( virtually ) constant velocity along the lattice up to an amount of disorder @xmath110 . generally , the larger the width @xmath65 the lower becomes this velocity . finally , for an overcritical amount of disorder the breather performs oscillatory motions around its starting lattice position . the frequency of these oscillations diminishes gradually with more enlarged @xmath65 and simultaneously the excursions from the starting site get continuously stronger confined . despite the non - directed character of its motion even for a moderate amount of disorder @xmath102 the breather is able to travel as far as twenty lattice sites away from its starting position . eventually linear anderson - type disorder effects overwhelm more and more the nonlinear activation mechanism for breather motion through the excitation of the nonlinear pinning mode . concerning the impact of structural disorder we considered random distributions of the base pair spacings @xmath111 $ ] and the twist angles @xmath112 $ ] with different mean standard deviations @xmath113 . for representative results we averaged over several realizations of structural disorder . the results for the temporal evolution of the electron breather center are depicted in fig . [ fig : fig6 ] . we conclude that structural disorder does not significantly affect the mobility of the breathers . for growing degree of disorder merely the propagation velocity gets reduced . even in the case of fairly strong disorder @xmath114 ( dashed line ) the breather is still able to propagate steadily . and angles @xmath61 . assignment of the line types to the increasing mean standard deviations ranging up to @xmath115 of the respective mean value : full line ( @xmath116 ) , dashed dotted line ( @xmath117 ) and dashed line ( @xmath115 ) . ] we studied also the polaron mobility under the impact of combined static diagonal and structural disorder and found that the strong pinning effect generated by static diagonal disorder prevails over the more harmless and merely velocity - reducing influence of structural irregularities . particularly the last result is in compliance with the findings in @xcite where it was reported that hopping conductivity in proteins and dna is influenced first of all by the molecular conformations and not by their sequences . it is instructive to consider also the linear limit case of the coupled nonlinear electron - vibration system even though the results obtained throughout this paper bear essentially nonlinear character . the study of the linear limit system itself and its floquet analysis can be applied as a reference and as a tool for obtaining useful information for the interpretation of the floquet analysis of the fully nonlinear system . the linear limit system is derived from the eqns . ( [ eq : dotc])-([eq : dottheta ] ) for @xmath45 giving @xmath118 in the first equation the terms @xmath119 , which in a homogeneous system reduce to @xmath120 , have been eliminated through the gauge transformation @xmath121 . in the linear limit the electronic and vibrational degrees of freedom are decoupled . to construct the electronic state we substitute in eq . ( [ eq : dotlinealc ] ) , @xmath122 with time - independent @xmath123 yielding the linear stationary discrete schrdinger equation : @xmath124 . with the bloch mode solutions @xmath125 we get then the dispersion relation : @xmath126 . the linear electronic energy spectrum runs from @xmath127 to @xmath128 . the polaron in the full system displayed in fig . [ fig : profile ] possesses an energy value @xmath129 which is slightly below the lower band edge @xmath130 assigned to a wave number @xmath131 and in the linear regime the frequencies are in the range @xmath132 $ ] . conclusively , the time scale for the evolution of the angular ( radial ) variables @xmath5 ( @xmath4 ) represented by independent harmonic oscillators with frequencies @xmath133 and @xmath134 , respectively , is two orders ( one order ) of magnitude larger than the one of the electron . the linear analysis is also useful to understand the origin of many of the numerically calculated floquet arguments for the fully nonlinear system . the values of the variables @xmath4 and @xmath5 after a period @xmath135 are : @xmath136 and @xmath137 . therefore , their corresponding floquet arguments are given by @xmath138 and @xmath139 , for @xmath4 and @xmath5 , respectively , which are attributed to angular positions in the complex plane with values around @xmath140 and @xmath141 for polarons with energies slightly below @xmath130 and with increased @xmath142 these angular positions shift to lower values on the unit circle . the numerical integration of the system of tangent equations over one period of the polaron s oscillation @xmath143 allows us to obtain the actual floquet matrix . its eigenvalues can be written as @xmath144 with @xmath82 being the floquet arguments . the system is linearly stable provided all the floquet arguments are real . as noted earlier in this paper within the range of parameters studied in this paper linear stability is assured , but this is not the only information that can be gained from the floquet analysis . in order to relate the positions of the numerous eigenvalues of the floquet matrix in the complex plane to the various degrees of freedom contained in the system it is elucidating to study first an anti - coupling limit in the electronic subsystem . introducing an auxiliary tunable coupling parameter @xmath145 , which corresponds to the transfer integral in the scaled system , the anti - coupling limit is obtained with @xmath146 and the actually studied coupled system with @xmath147 . for the decoupled electron system ( @xmath146 ) the value of the energy @xmath148 can be easily derived with the help of the eqs . ( [ eq : rinst ] ) and ( [ eq : statthree ] ) , giving @xmath149 and @xmath150 , which leads to @xmath151 for the localized state @xmath152 and @xmath153 as a typical value . therefore , the @xmath4-multipliers are in the vicinity of @xmath154 , the @xmath5 s lie at small angles at both sides of @xmath155 and the @xmath13 s are located at @xmath155 on the unit circle , in accordance with the analysis performed in [ sub : linear ] . as we switch on the coupling @xmath156 and increase it up to the actual value @xmath147 , we can follow the evolution of the multipliers . eventually , for @xmath147 the @xmath4 s and @xmath5 s have moved to positions with about half their initial angle values while spreading a little on the unit circle ( the @xmath5 s to lesser extent though ) . the @xmath13 s separate from @xmath155 and invade the whole circle , as a consequence of the nonlinear coupling between the electron amplitudes and the vibrations . the exceptions are the pair at @xmath155 , corresponding to the phase mode , and another isolated pair close to this point . the latter pair is associated with the spatially antisymmetric pinning mode , responsible for the mobility of the polaron . note that the linear approximation described in [ sub : linear ] is an approximation for the radial and angular variables . the floquet multipliers for the decoupled and full systems , respectively are shown in fig . [ fig : floquet1 ] . ) -([eq : dotlinealth ] ) at the anti - coupling limit @xmath146 . right figure : the fully coupled nonlinear system . ] in order to illuminate further the role played by the vibrational modes with respect to the initiation of mobility of the polaron we cast the floquet matrix @xmath81 in the form of a supermatrix formed by submatrices : @xmath157 and the indices of the submatrices refer to the origin of their entries . from the arguments of the eigenvalues of the submatrices an / or combinations of them , it is possible to deduce additional information about the spectral features attributed to the various degrees of freedom contained in the coupled nonlinear system . the matrix @xmath158 appears in the born - oppenheimer approximation . to recover the pinning mode of the complete system it suffices to consider only the subsystem consisting of the electronic part coupled to the radial motions for which the floquet matrix reduces to the submatrix @xmath159 , which means that the mobility of the polaron is mainly a consequence of the coupling between the electron amplitudes and the radial variables and that the electron will not move within the born - oppenheimer approximation . the figs . [ fig : floquet2 ] and [ fig : floquet3 ] illustrate these statements . the left ( right ) part of the fig . [ fig : floquet2 ] shows the multipliers of @xmath160 ( @xmath161 ) . in comparison with the linear case studied in the preceding section we note that while the radial components of the multipliers shifted their positions to lower angles and exhibit some spread the angular multipliers hardly alter their position . hence , perturbations of the angular twist motions in the context of the coupled system resemble the behavior of the linear limit system . the left part of fig . [ fig : floquet3 ] depicts the eigenvalues of @xmath162 in the vicinity of @xmath155 . the pair of pinning mode eigenvalues is recognizable . in the right part of the figure the normalized pinning mode is portrayed . ( left figure ) and @xmath161 ( right figure ) . ] in the vicinity of @xmath155 . the pinning mode eigenvalues are recognizable . the right part of the figure shows the normalized pinning mode . ] moreover , it should be pointed out that for larger elongations of @xmath5 , beyond the range where the linear expansion of @xmath163 is valid , there exists also an angular pinning mode as a consequence of the coupling between the electronic amplitudes and the angular variables . the fact that the disorder eventually prevents the polaron from being moved can also be explained with the help of the floquet multipliers . when the degree of disorder is increased the pinning mode eigenvalues move progressively away from @xmath155 towards larger arguments @xmath84 and approach the rest of the @xmath164 eigenvalues . [ fig : fdisorder ] displays the growth of the arguments of the pinning mode as a function of the disorder strength @xmath65 . eventually the system is too far away from the instability due to the marginal mode , which appears when the pair of pinning mode eigenvalues collapses at @xmath155 . this is easy to understand , as the disorder brings about a decoupling of the variables , favoring localized , stationary , isolated modes . demonstrating the growth of the argument of the pinning mode with increased amount of disorder so that the pinning mode eventually merges in the rest of the modes . ] in the present work we have studied an electron transfer mechanism in dna relying on the coupling of the charge carrying unit to the vibrational modes of dna responsible for the formation of polarons and electron - vibron breathers . the steric structure of the double helix of @xmath0dna has been described in the context of the base - pair picture considering angular and radial vibrational motions of the base pairs . we have focused interest on the initiation of long - range and stable polaron and breather motion along the dna structure when static and/or structural disorder inherent to any real dna molecule is taken into account . we remark that the proposed polaron - like transport mechanism applies to electron hole conduction too . in the first part of the paper we have constructed exact stationary localized solutions ( polarons and static electron - vibron breathers ) with the help of a nonlinear map approach . we have found that the interplay of disorder and nonlinearity amplifies the degree of localization in comparison with cases when either disorder or nonlinearity alone is present . with concern to mobile electron - vibron breather solutions , whose motion is activated through a kick mechanism utilizing the pinning mode , with the current work we have demonstrated that they are fairly robust when the electron - vibration system is subjected to randomness . generally , even for a moderate amount of disorder the breathers retain their localized shapes and support directed , long - ranged coherent electron transfer along strands of the bent double helix . in particular the last result suggests that dna seems suitable for the design of ( one - dimensional ) functional nanostructures as ingredients of molecular nanoelectronical devices . furthermore , we have observed that the electron breather is accompanied by vibrational breathers in the radial as well as angular components , respectively causing local and temporal deformations of the traversed region of the double helix . conclusively , efficient charge transport in dna proceeds on the condition that the double helix undergoes structural changes which is exemplary for the interplay of structure and function in flexible and adaptive biomolecules . one of the authors ( d.h . ) acknowledges support by the deutsche forschungsgemeinschaft via a heisenberg fellowship ( he 3049/1 - 1 ) . j.f.r.a . acknowledges d.h . and the institut fr theoretische physik for their warm hospitality . j.a . is grateful for a scholarship supplied by the studienstiftung des deutschen volkes . the authors are also grateful to the support under the locnet eu network hprn - ct-1999 - 00163 . 99 d.d . eley and d.i . spivey , trans . faraday soc . * 58 * , 411 ( 1962 ) . mirkin , r.l . letsinger r.c . mucic and j.j . storhoff , nature * 382 * , 607 ( 1996 ) ; a.a . voityuk , n. rsch , m. bixon , and j. jortner , j. phys . chem . * 104 * , 9740 ( 2000 ) ; f.c . grozema , y.a . berlin and l.d.a . siebbels , j. am . soc . * 122 * 10903 ( 2000 ) ; y.a . berlin , a.l . burin and m.a . ratner , j. am . soc . * 123 * , 260 ( 2001 ) . m. ratner , nature * 397 * , 480 ( 1999 ) . . hall , r.e . holmkin and j.k . barton , nature * 382 * , 731 ( 1996 ) . d.b . hall and j.k . barton , j. am . * 119 * , 5045 ( 1997 ) . arkin , e.d.a . stemp , s.c . pulver and j.k . barton , chem . biol . * 4 * , 369 ( 1997 ) . y. okahata , t. kobayashi , k. tanaka , and m. shimomura , j. am . 120 * , 6165 ( 1998 ) . e. meggers , m.e . michel - beyerle and b. giese , j. am . soc . * 120 * , 12950 ( 1998 ) . fink and c. schnenberger , nature * 398 * , 407 ( 1999 ) . p. tran , b. alavi and g. gruner , phys . * 85 * , 1564 ( 2000 ) . e. braun , y. eichen , u. sivan and g. ben - yoseph , nature * 391 * , 775 ( 1998 ) . d. porath , a. bezryadin , s. de vries and c. dekker , nature * 403 * , 635 ( 2000 ) . de pablo , f. moreno - herrero , j. colchero , j. g@xmath165mez herrero , p. herrero , a.m. bar@xmath166 , p. ordej@xmath166n , j.m . soler and e. artacho , phys . * 85 * , 4992 ( 2000 ) . j. jortner , m. bixon , t. langenbacher and m.e . michel - beyerle , proc . usa * 95 * , 12759 ( 1998 ) . y .- j . ye , r .- s . chen , a. martinez , p. otto and j. ladik , sol . . comm . * 112 * , 139 ( 1999 ) . r. bruinsma , g. grner , m.r . dorsogna and j. rudnick , phys . lett . * 85 * , 4393 ( 2000 ) . yu and xueyu song , phys . lett . * 86 * , 6018 ( 2001 ) . m. hjort and s. stafstrm , phys . lett . * 87 * , 228101 - 1 ( 2001 ) . d. ly et al , j. am . * 118 * , 8747 ( 1996 ) . j. jortner , proc . usa , * 95 * , 12759 ( 1998 ) . e. conwell and s.v . rakhmanova , proc . usa , * 97 * , 4556 ( 2000 ) . d. ly , l. sanii and g.b . schuster , j. am . soc . , * 121 * , 9400 ( 1999 ) . z. hermon , s. caspi and e. ben - jacob , europhys . lett . * 43 * , 482 ( 1998 ) . m. peyrard and a.r . bishop , phys . lett . * 62 * , 2755 ( 1989 ) . yakushevich , quart . * 26 * , 201 ( 1993 ) . m. barbi , s. cocco and m. peyrard , phys . lett . a * 253 * , 358 ( 1999 ) . l. stryer _ biochemistry _ , ( freeman , new york , 1995 ) . s. cocco and r. monasson , j. chem . phys . * 112 * , 10017 ( 2000 ) . holstein , ann . ny * 8 * , 325 , 343 ( 1959 ) . p.w . anderson , phys . * 109 * , 1492 ( 1958 ) . mott , adv . phys . * 10 * , 107 ( 1961 ) . d. bambusi and a. giorgilli , j. stat . phys . * 71 * , 569 ( 1993 ) . d. hennig , j. math . phys . * 39 * , 3568 ( 1998 ) . mackay and s. aubry , nonlinearity * 7 * , 1623 ( 1994 ) . d. hennig , in preparation . g. kalosakas , s. aubry and g.p . tsironis , phys . b * 58 * , 3094 ( 1998 ) . n.k . voulgarakis and g.p . tsironis , phys . rev . b * 63 * , 14302 ( 2001 ) . m. peyrard and m.d . kruskal , physica d * 14 * , 88 ( 1984 ) . s. flach and c.r . willis , phys . 295 * , 181 ( 1998 ) . d. chen , s. aubry , and g.p . tsironis , phys . lett . * 77 * , 4776 ( 1996 ) . y .- j . ye and j. ladik , int . . chem . * 52 * , 491 ( 1994 ) .

understanding strong correlation effect away from equilibrium has been one of the most interesting yet challenging problems in condensed matter physics . a prominent realization of such phenomena is embodied in quantum transport through a nanostructure under finite bias voltage . to understand the interplay of the correlation effect and nonequilibrium nature , nonequilibrium version of the single impurity anderson model ( siam ) and its extensions have been serving and continue to do so as a central theoretical model . the siam is indeed considered to be one of the best studied strongly correlated model , and despite its apparent simplicity , it exhibits rich physics already in equilibrium , such as the coulomb blockade and the kondo physics that have been observed in experiments . equilibrium properties of the siam have been well understood thanks to concerted efforts of several theoretical approaches over years by perturbative treatment , fermi liquid description , as well as exact results by the bethe ansatz method and numerical renormalization group ( nrg ) calculations ( see for instance , . ) in contrast , the situation of the nonequilibrium siam is not so satisfactory . each of the above approaches has met some difficulty in treating nonequilibrium phenomena . a theoretical approach that can deal with the strong correlation effect in nonequilibrium is still called for . notwithstanding , a number of analytical and numerical methods have been devised to investigate nonequilibrium stationary phenomena : nonequilibrium perturbation approaches @xcite and its modifications @xcite , the non - crossing approximation @xcite , the functional renormalization group treatment @xcite , quantum monte carlo calculations on the keldysh contour @xcite , the iterative real - time path integral method @xcite and so on . unfortunately those approaches fail to give a consistent picture concerning the finite bias effect on the dot spectral function , particularly regarding a possible splitting of the kondo resonance . as for the equilibrium siam , the second - order perturbation regarding the coulomb interaction @xmath0 on the dot @xcite is known to capture essential features of kondo physics and agrees qualitatively well with exact results obtained by the bethe ansatz and nrg approaches @xcite . such good agreement seems to persist in nonequilibrium stationary state at finite bias voltage . for the two - terminal particle - hole ( ph ) symmetric siam , a recent study by @xcite showed that the nonequilibrium second order perturbation calculation of the spectral function agrees with that calculated by the diagrammatic quantum monte carlo simulation , excellently up to interaction strength @xmath1 ( where @xmath2 is the total relaxation rate due to leads ) ; pretty well even for @xmath3 at bias voltage @xmath4 . a typical magnitude of @xmath5 of a semiconductor quantum dot is roughly @xmath6 depending on the size and the configuration of the dot . therefore , there is a good chance of describing a realistic system within the validity of nonequilibrium perturbation approach . the great advantage of semiconductor dot systems is to allow us to control several physical parameters . those include changing gate voltage as well as configuring a more involved structure such as a multi - terminal dot @xcite or an interferometer embedding a quantum dot . theoretical treatments often limit themselves to a system with the ph symmetry where the dot occupation number is fixed to be one half per spin . although assuming the ph symmetry makes sense and comes in handy in extracting the essence of the kondo resonance , we should bear in mind that such symmetry is not intrinsic and can be broken easily in realistic systems , by gate voltage , asymmetry of the coupling with the leads , or asymmetric drops of bias voltage @xcite . the ph asymmetry commonly appears in a multi - terminal dot or in an interferometer embedding a quantum dot . it is also argued that the effect of the ph asymmetry might be responsible for the deviation observed in nonequilibrium transport experiments from the `` universal '' behavior of the ph symmetric siam @xcite . to work on realistic systems , it is imperative to understand how the ph asymmetry affects nonequilibrium transport . in this paper , we examine the second - order nonequilibrium perturbation regarding the coulomb interaction @xmath0 of the multi - terminal siam . the ph symmetry is not assumed , and miscellaneous types of asymmetry of couplings to the leads and/or voltage drops are incorporated as a generic multi - terminal configuration . our main focus is to provide solid analytical results of the behavior of the nonequilibrium self - energy and hence the dot spectral function for the full range of frequency and bias voltage , within the validity of the second - order perturbation theory of @xmath0 . the result encompasses fermi liquid behavior as well as incoherent non - fermi liquid contribution , showing analytically that increasing finite bias voltage leads to a crossover from the kondo resonance to the coulomb blockade behaviors . the present work contrasts preceding perturbative studies @xcite whose evaluations relied on either numerical means or a small - parameter expansion of bias voltage and frequency . the only notable exception , to the author s knowledge , is a recent work by @xcite , which succeeded in evaluating analytically the second - order self - energy for the two - terminal ph symmetric dot . intending to apply such analysis to a wider range of realistic systems and examine the effect that the two - terminal ph symmetric siam can not capture , we extend their approach to a generic multi - terminal dot where the ph symmetry may not necessarily be present . an embarrassing drawback of using the nonequilibrium perturbation theory is that when one has it naively apply to the ph asymmetric siam , it may disrespect the preservation of the steady current . @xcite as a result , one then needs some current - preserving prescription , and different self - consistent schemes have been proposed and adopted . @xcite as will be seen , the current preserving condition involves all the frequency range , not only of low frequency region that validates fermi liquid description but also of the incoherent non - fermi liquid part ( see eq . below ) . therefore , an approximation based on the low energy physics , particularly the fermi liquid picture , should be used with care . the self - energy we will construct analytically is checked to satisfy the spectral sum rule at finite bias voltage , so that we regard it as giving a consistent description for the full range of frequency in nonequilibrium . by taking its advantage , we also demonstrate a self - consistent , current - preserving calculation of the nonequilibrium spectral function for a system where the ph symmetry is not present . the paper is organized as follows . in sec . [ sec : model ] , we introduce the multi - terminal siam in nonequilibrium . we review briefly how to obtain the exact current formula by clarifying the role of the current conservation at finite bias voltage . section [ sec : self - energy ] presents analytical expression of the retarded self - energy for a general multi - terminal dot up to the second - order of the interaction strength . subsequently in sec . [ sec : discussion ] , we examine and discuss its various analytical behaviors including high- and low - bias voltage limits . sec . [ sec : spectral - fn ] is devoted to constructing a non - equilibrium spectral function using the self - energy obtained in the previous section . we focus our attention on two particular situations : ( 1 ) self - consistent , current - preserving evaluation of the nonequilibrium spectral function for the two - terminal ph asymmetric siam , and ( 2 ) multi - terminal effect of the ph symmetric siam . finally we conclude in sec . [ sec : conclusion ] . mathematical details leading to our main analytical result eq . as well as other necessary material regarding dilogarithm is summarized in appendices . the model we consider is the single impurity anderson model connecting with multiple leads @xmath7 whose chemical potentials are sustained by @xmath8 . the total hamiltonian of the system consists of @xmath9 , where @xmath10 , @xmath11 and @xmath12 represent the dot hamiltonian with the coulomb interaction , the hopping term between the dot and the leads , and the hamiltonian of a noninteracting lead @xmath13 , respectively . they are specified by @xmath14 where @xmath15 is the dot electron number operator with spin @xmath16 and @xmath17 are electron operators at the lead @xmath13 . in the following , we consider the spin - independent transport case , but an extension to the spin - dependence situation such as in the presence of magnetic field or ferromagnetic leads is straightforward . when applying the wide - band limit , all the effect of the lead @xmath13 is encoded in terms of its chemical potential @xmath8 and relaxation rate @xmath18 , where @xmath19 is the density of states of the lead @xmath13 . the dot level @xmath20 controls the average occupation number on the dot ; it corresponds roughly to @xmath21 , @xmath22 , @xmath23 for @xmath24 , @xmath25 , and @xmath26 , respectively . the ph symmetry is realized when @xmath27 and @xmath28 ( see eqs . and below ) . we here briefly summarize how the current through the dot is determined in a multi - terminal setting . special attention is paid to the role of the current conservation because it has been known that nonequilibrium perturbation calculation does not respect it in general . @xcite we illustrate how to ensure the current conservation by requiring minimally . the argument below is valid regardless of a specific approximation scheme chosen , whether nonequilibrium perturbation or any other approach . following the standard protocol of the keldysh formulation @xcite , we start with writing the current @xmath29 flowing from the lead @xmath13 to the dot in terms of the dot s lesser green function @xmath30 and the retarded one @xmath31 : @xmath32,\end{aligned}\ ] ] where @xmath33 is the fermi distribution function at the lead @xmath13 . as the present model preserves the total spin as well as charge , the net spin current flowing to the dot should vanish in the steady state , which imposes the integral relation between @xmath30 and @xmath31 , @xmath34 = 0 . \label{eq : gless - gr}\end{aligned}\ ] ] here we have introduced the total relaxation rates @xmath35 and the effective fermi distribution @xmath36 weighted by the leads , @xmath37 when we ignore the energy dependence of the relaxation rates @xmath38 , we can recast eq . into a more familiar form @xmath39 because @xmath40 is the definition of the exact dot occupation number . note the quantity @xmath41 is nothing but the exact dot spectral function out of equilibrium . we emphasize that eq . ( or equivalently eq . ) is the minimum , exact requirement that ensures the current preservation . it constrains the exact @xmath42 and @xmath43 that depend on the interaction as well as bias voltage in a nontrivial way . one can accordingly eliminate @xmath44 in @xmath29 , to reach the landauer - buttiker type current formula at the lead @xmath13 , @xmath45 or , the current conservation allows us to write it as @xmath46,\end{aligned}\ ] ] where @xmath47 is the exact number of states with spin @xmath16 at finite temperature in general , defined by @xmath48 it tells us that differential conductance @xmath49 with fixing all other @xmath50 s is proportional to the nonequilibrium dot spectral function , provided changing @xmath8 does not affect the occupation number @xcite . such situation is realized , for instance , when a probe lead couples weakly to the dot . the case of a noninteracting dot always satisfies the current preserving condition eq . as @xmath51 holds for any @xmath52 ; the distribution function of dot electrons @xmath53 is equal to @xmath54 . this is not the case for an interacting dot , however . as for the interacting case , not so much can be said . we only see the special case with the two - terminal ph symmetric dot satisfy eq . by choosing @xmath55 irrespective of interaction strength . except for this ph symmetric case , a general connection between @xmath42 and @xmath43 is not known so far . it is remarked that , based on the quasiparticle picture , a noninteracting relation @xmath56 is sometimes used to deduce an approximate form of @xmath42 out of @xmath43 for an interacting dot . such approximation is called the ng s ansatz @xcite . although it might be be simple and handy , its validity is far from clear . we will not rely on such additional approximation below . it is also important to distinguish in eq . the electron occupation number @xmath57 from the quasiparticle occupation number @xmath58 , as the two quantities are different at finite bias voltage since the luttinger relation holds only in equilibrium @xcite . contribution to the dot occupation number comes from all range of frequency , including the incoherent part . one sees fulfilling the spectral weight sum rule @xmath59 crucial to have the dot occupation number @xmath57 normalized correctly . in general , one needs to determine @xmath57 appropriately to satisfy eq . as a function of interaction and chemical potentials of the leads . the applicability of quasi - particle approaches that ignores the incoherent part is unclear . in this section , we evaluate analytically the nonequilibrium retarded self - energy up to the second order of interaction strength @xmath0 for the multi - terminal siam . we first examine the contribution at the first - order and the role of current preservation . then we present the analytical result of the second - order self - energy in terms of dilogarithm . following the standard treatment of the keldysh formulation , @xcite the nonequilibrium green function and the self - energy take a matrix structure @xmath60 satisfying symmetry relations @xmath61 and @xmath62 . the retarded green function is defined by @xmath63 ; the retarded self - energy , by @xmath64 . to proceed evaluation , it is convenient to classify self - energy diagrams into the two types : the hartree - type diagram ( fig . [ fig : hartree - type ] ) that can be disconnected by cutting a single interaction line , and the rest which we call the correlation part and reassign the symbol @xmath65 to . the latter starts at the second - order . the resulting green function ( matrix ) takes a form of @xmath66^{-1},\end{aligned}\ ] ] where @xmath67 represents a pauli matrix of the keldysh structure , and @xmath68 refers to the exact occupation number of the dot electron with the opposite spin . accordingly , the corresponding retarded green function becomes @xmath69 where @xmath70 is the hartree level of the dot . . the double line refers to the exact green function . ] before starting evaluating the correlation part @xmath71 that starts contributing at the second order , it is worthwhile to examine the current preserving condition eq . at the first order . at this order , it reduces to the self - consistent hartree - fock equation for the dot occupation number @xmath72 : @xmath73 . \label{eq : hf - self - consistency}\end{aligned}\ ] ] it shows how the two - terminal ph symmetric siam is special by choosing @xmath74 , @xmath75 and @xmath76 ; the second term of the right - hand side vanishes by having the solution @xmath77 even at finite bias voltage . it also indicates that the current preservation necessarily has the occupation number depend on asymmetry of the lead couplings as well as interaction strength for the ph asymmetric siam . indeed , for a small deviation from the ph symmetry and bias voltage , we see the hartree - fock occupation number behave as @xmath78 where @xmath79 is the average chemical potential weighted by leads , @xmath80 note @xmath79 vanishes when no bias voltage applies , as we incorporate the overall net offset by leads into @xmath20 . following the standard perturbation treatment of the keldysh formulation , we see there is only one diagram contributing to @xmath81 at the second order ( fig . [ fig : sigma2 ] ) after eliminating the hartree - type contribution . the contribution is written as @xmath82 where @xmath83 refer to to the unperturbed green functions ( including the hartree term ) , whose concrete expressions are found in appendix a. the polarization matrix @xmath84 is defined by @xmath85 ( fig . [ fig : pi ] ) .. ] as was shown by the current formula in the previous section , we need only the dot spectral function to study quantum transport , hence @xmath71 suffices . therefore it is more advantageous to work on the rak - representation , where the polarization parts become @xmath86 , \\ & \pi^{a}_{12 } = \frac{i\hbar}{2 } \left [ g^{a}_{12 } g^{k}_{21 } + g^{k}_{12 } g^{r}_{21 } \right ] , \\ & \pi^{k}_{12 } = \frac{i\hbar}{2 } \left [ g^{k}_{12 } g^{k}_{21 } + g^{r}_{12 } g^{a}_{21 } + g^{a}_{12 } g^{r}_{21 } \right ] , \end{aligned}\ ] ] [ eq : def - pi ] and their fourier transformations are given in appendix b. accordingly , we can express the retarded self - energy @xmath71 as @xmath87,\end{aligned}\ ] ] where @xmath88 the above second - order expression of @xmath71 is standard , but it has so far been mainly used for numerical evaluation , quite often restricted for the two - terminal ph symmetric siam . we intend to evaluate eqs . ( [ eq : i1],[eq : i2 ] ) analytically for the generic multi - terminal siam , along the line employed in ref . . delegating all the mathematical details to appendices c and d , we summarize our result of the analytical evaluation of @xmath71 as follows : @xmath89 , \label{eq : sigmar}\end{aligned}\ ] ] here , functions @xmath90 ( @xmath91 ) are found to be ( using @xmath92 ) , [ eq : xi123 ] @xmath93 \nonumber \\ & \qquad \qquad + \sum_{a , b,\beta } \tfrac{4\gamma_{a } \gamma_{b}}{\gamma^{2 } } \big [ { \mathop{\mathrm{li}_{2}}\nolimits}(\tfrac{-\varepsilon+ \beta \zeta_{a\bar{\sigma } } } { \beta \zeta_{b \bar{\sigma } } + i\gamma } ) + \tfrac{1}{4 } { \mathop{\mathrm{log}}\nolimits}^{2 } ( \tfrac{-\zeta_{a\bar{\sigma}}+i\gamma}{\zeta_{b\bar{\sigma } } + i\gamma } ) \big ] , \\ \xi_{2}(\varepsilon ) & = 6\pi^{2 } - \sum_{a , b,\beta } \tfrac{4\gamma_{a } \gamma_{b}}{\gamma^{2 } } \big [ \lambda(\tfrac{\varepsilon -\zeta_{a\sigma } + 2i\gamma}{\beta \zeta_{b\bar{\sigma } } + i\gamma } ) + \lambda(\tfrac{\varepsilon - \beta \zeta_{b\bar{\sigma } } + 2i\gamma}{\zeta_{a\sigma } + i\gamma } ) + \tfrac{1}{2 } { \mathop{\mathrm{log}}\nolimits}^{2 } ( \tfrac{\zeta_{a\sigma } + i\gamma}{\beta \zeta_{b \bar{\sigma } } + i\gamma } ) \big ] \nonumber \\ & \qquad \qquad - \sum_{a , b,\beta } \tfrac{4\gamma_{a } \gamma_{b}}{\gamma^{2 } } \big [ \lambda ( \tfrac{\varepsilon+ \beta \zeta_{a\bar{\sigma } } + 2i\gamma } { \beta \zeta_{b\bar{\sigma } } + i\gamma } ) + \tfrac{1}{4 } { \mathop{\mathrm{log}}\nolimits}^{2 } ( \tfrac{-\zeta_{a\bar{\sigma } } + i\gamma}{\zeta_{b\bar{\sigma } } + i\gamma } ) \big ] , \\ \xi_{3 } & = \left[\sum_{a } \tfrac{2\gamma_{a}}{\gamma } { \mathop{\mathrm{log}}\nolimits}\left ( \tfrac{-\zeta_{a\bar{\sigma } } + i\gamma}{\zeta_{a\bar{\sigma } } + i\gamma}\right ) \right]^{2 } , \end{aligned}\ ] ] [ eq : xi-123 ] where the summations over @xmath94 as well as terminals @xmath95 are understood . function @xmath96 is dilogarithm , whose definition as well as various useful properties are summarized in appendix c ; @xmath97 is defined by is equivalent to that given in ref . , but we prefer writing it in this form because its analyticity is more transparent . ] @xmath98 { \mathop{\mathrm{log}}\nolimits}z . \label{def : lambda}\end{aligned}\ ] ] analytical formula @xmath71 given in eqs . and constitutes the main result of the present paper . consequently , the nonequilibrium spectral function of the multi - terminal siam is given analytically for full range of frequency and bias voltage , once one chooses @xmath57 to satisfy eq . . the result also applies to a more involved structured system , such as an interferometer embedding a quantum dot , by simply replacing @xmath20 and @xmath38 to take account of those geometric effect . having obtained an explicit analytical form of the second - order self - energy @xmath99 at arbitrary frequency and bias voltage , we now examine its various limiting behaviors . most of those limiting behaviors have been known for the two - terminal ph symmetric siam , so it s assuring to reproduce those expressions in such a case . simultaneously , our results below provide multi - terminal , ph asymmetric extensions of those asymptotic results . we can reproduce the equilibrium result by setting all the chemical potentials equal , @xmath100 . then we immediately see @xmath101 and @xmath102 , \\ & \xi_{2 } = 8\cdot \big[\frac{3\pi^{2}}{4 } - 3\lambda(\tfrac{\varepsilon_{\sigma } + 2i\gamma}{i\gamma } ) \big],\end{aligned}\ ] ] where @xmath103 . as a result , the correlation part of the self - energy in equilibrium becomes @xmath104 . \label{eq : sigma - equilibrium}\end{aligned}\ ] ] the ph symmetric case in particular corresponds to @xmath105 . it reproduces the perturbation results by yamada and yoshida @xcite up to the second - order of @xmath0 , when we expand the above for small @xmath52 . the ph symmetric result is indeed identical with the one obtained in ref . for arbitrary frequency ( see also eq . below ) . @xcite have evaluated analytically the self - energy and the spectral function for the two - terminal ph symmetric siam . in our notation , it corresponds to the case @xmath106 , and @xmath107 . when we parametrize the two chemical potentials by @xmath108 with @xmath109 in eqs . , the self - energy can be written as @xmath110 , \label{eq : ph - symmetric - sigmar}\end{aligned}\ ] ] where @xmath111 , \\ & \xi_{2 } = 6\pi^{2 } - 6\sum_{a , b } \big [ \lambda(\tfrac{\omega -a ev/2 + 2i\gamma}{b ev/2 + i\gamma } ) + \frac{1}{4 } { \mathop{\mathrm{log}}\nolimits}^{2 } ( \tfrac{a ev/2 + i\gamma}{b ev/2 + i\gamma } ) \big].\end{aligned}\ ] ] the above result are identical with what ref . obtained . we now employ the small - parameter expansion of @xmath71 around the half - filled equilibrium system . here we assume parameters @xmath112 and @xmath113 are much smaller than the total relaxation rate @xmath2 . the expansion of @xmath114 is found to contain the first- and second - order terms regarding @xmath115 and @xmath116 , while @xmath117 do only the second - order terms . therefore the result of the expansion up to the second - order of these small parameters is presented as @xmath118.\end{aligned}\ ] ] functions @xmath90 can be expanded straightforwardly by using the taylor expansion of dilogarithm in appendix c. they are found to behave @xmath119 , \\ & \xi_{2}(\varepsilon ) \approx - 8 \ , \big [ \frac{3 ( -\varepsilon^{2 } + 2 \varepsilon\bar{\mu } + 2 \bar{\mu}^{2 } - \overline{\mu^{2 } } - 2 \overline{\mu^{2}})}{4(i\gamma)^{2 } } \big ] \\ & \xi_{3 } \approx \frac{16 \bar{\mu}^{2}}{(i\gamma)^{2}},\end{aligned}\ ] ] where @xmath79 is defined in eq . and we have introduced @xmath120 combining all of these , we reach the small bias / frequency behavior of the self - energy @xmath81 as @xmath121 \nonumber \\ & \qquad \qquad - \frac{iu^{2}}{2\pi^{2}\gamma^{3 } } \big [ ( \omega - \bar{\mu})^{2 } + 3\ , ( \delta \mu)^{2 } \big ] . \label{eq : small - parameter - expansion}\end{aligned}\ ] ] small bias expansion of @xmath122 for the two - terminal system has been discussed and determined by the argument based on the ward identity . @xcite the dependence appearing in eq . fully conforms to it ( except for the presence of the bare interaction instead of the renormalized one ) . indeed , correspondence is made clear by noting the parameters @xmath79 and @xmath123 for the two - terminal case become @xmath124 the presence of linear term in @xmath52 and @xmath125 for the two - terminal ph asymmetric siam was also emphasized recently . @xcite one expects naively that the limit of large bias voltage @xmath126 corresponds to the high temperature limit @xmath127 in equilibrium ; it was shown to be so for the two - terminal ph symmetric siam . @xcite we now show that the same applies to the multi - terminal siam where bias voltages of the leads are pairwisely large , _ i.e. _ , half of them are positively large , and the others are negatively large . in the large bias voltage limit , all the arguments of dilogarithm functions appearing in eqs . become @xmath128 , where the values of dilogarithm are known ( see appendix c ) . accordingly , the pairwisely large bias limit of @xmath90 is found to be @xmath129 correspondingly , the retarded self - energy becomes @xmath130 it shows that the result of the multi - terminal siam is the same with that of the two - terminal ph symmetric siam except for a frequency shift . accordingly , the retarded green function @xmath131 in this limit is given by @xmath132 the form indicates that for sufficiently strong interaction @xmath133 , the dot spectral function has two peaks at @xmath134 with broadening @xmath135 , so the system is driven into the the coulomb blockade regime . on the other hand , for weak interaction @xmath136 , it has only one peak with two different values of broadening that reduce to @xmath2 and @xmath137 in the @xmath138 limit . what is the role of the current preservation condition eq . in this limit ? it just determines the dot occupation number explicitly . in fact , the condition becomes @xmath139 with @xmath140 , and @xmath57 is independent of the interaction strength because bias voltage sets the largest scale . one can evaluate the above integral exactly to have @xmath141 where @xmath142 . as a result , expanding it up to the second order of @xmath143 leads to @xmath144 the first term corresponds to the occupation number that one expects naturally from the effective distribution @xmath36 ; it corresponds , for instance , @xmath145 for the two - terminal dot with @xmath146 with @xmath147 . the second term is the deviation from it , which is proportional to the average of the inverse chemical potential weighted by the leads . we now turn our attention to the behavior of the nonequilibrium dot spectral function , using our analytical expression of the self - energy eqs . ( [ eq : sigmar],[eq : xi123 ] ) . below we particularly focus our attention on the two cases : the two - terminal ph asymmetric siam where current preservation has been an issue , and the multi - terminal ph symmetric siam where the role of multiple leads has been raising questions . in all of the calculations below , we have checked numerically the validity of the spectral weight sum rule at each configuration of bias voltages . as was emphasized in sec . [ sec : current - formula ] , when a dot system does not retain the ph symmetry , the stationary current is not automatically conserved and one must impose the current preservation condition eq . explicitly . as the right - hand side of eq . also depends on the dot occupation number @xmath57 , this requires us to determine @xmath57 self - consistently by using the retarded green function in a certain approximation the second - order perturbation theory in the present case . ) at finite bias voltage @xmath148 . the interaction strength is chosen as @xmath149 . the dotted line represents the result of @xmath150 and @xmath151 . ] figure [ fig : ph - symmetric ] shows the result of nonequilibrium dot spectral function of the two - terminal ph symmetric siam at bias voltage @xmath152 , and @xmath153 , which is essentially the same result with ref . ( of a different set of parameters ) . the occupation number is fixed to be @xmath154 in this case , so its self - consistent determination is unnecessary . the results were compared favorably with those obtained by diagrammatic quantum monte carlo calculations @xcite ; a relatively good quantitative agreement was observed up to @xmath155 ( where the bethe ansatz kondo temperature @xmath156 @xcite while the estimated half width of the kondo resonance @xmath157 ) and bias voltage @xmath158 . applying bias voltage gradually suppresses the kondo resonance without splitting it , and the two peaks at @xmath159 are developed at larger bias voltages , which corresponds to the discussion in the previous section . figure [ fig : ph - asymmetric ] shows the result of our self - consistent calculation of the nonequilibrium spectral function for the two - terminal ph asymmetric siam at ( a ) @xmath160 and ( b ) @xmath161 . a paramagnetic - type solution is assumed in determining @xmath57 . as in the ph symmetric siam , one sees increasing bias voltage not split but suppress the kondo resonance while it develops a peak around @xmath162 . the kondo resonance peak is suppressed more significantly at @xmath160 than at @xmath163 , because the kondo temperature of the former ( @xmath164 ; @xmath165 ) is smaller than that of the latter ( @xmath166 ; @xmath167 ) . an interesting feature of the ph asymmetric siam is that spectral weight of the kondo resonance seems shifting gradually toward @xmath168 with increasing bias voltage , without exhibiting a three - peak structure in the ph symmetric case . this suggests a strong spectral mixing between the kondo resonance and a coulomb peak at finite bias voltage . because of it , the interval of the two peaks at finite bias is observed as roughly @xmath169 and gets widened up to @xmath0 for larger @xmath170 . the bias dependence somehow looks similar to what was obtained by assuming equilibrium noninteracting effective distribution for @xmath57 @xcite ( which is hard to justify in our opinion ) , though we emphasize our present calculation only relies on the current - preservation condition without using any further assumption . it is remarked that the effect shown by bias voltage is quite reminiscent of finite temperature effect that was observed in the ph asymmetric siam in equilibrium @xcite . , and ( b ) @xmath171 . all the other parameters are the same with fig . [ fig : ph - symmetric ] . as an eye guide , the ph symmetric result of @xmath150 and @xmath151 is shown as a dotted line.,title="fig : " ] , and ( b ) @xmath171 . all the other parameters are the same with fig . [ fig : ph - symmetric ] . as an eye guide , the ph symmetric result of @xmath150 and @xmath151 is shown as a dotted line.,title="fig : " ] . results of the interaction strengths @xmath172 , @xmath173 , and @xmath174 are shown , while dotted lines refer to the noninteracting case as an eye guide . ( a ) spectral function of the ph symmetric siam at @xmath175 . ( b ) spectral function of the ph asymmetric siam at @xmath176.,title="fig : " ] . results of the interaction strengths @xmath172 , @xmath173 , and @xmath174 are shown , while dotted lines refer to the noninteracting case as an eye guide . ( a ) spectral function of the ph symmetric siam at @xmath175 . ( b ) spectral function of the ph asymmetric siam at @xmath176.,title="fig : " ] more insight can be gained by examining how the spectral structure depends on the interaction strength at finite bias voltage . figures [ fig : ldos - u ] ( a , b ) show a structural crossover from a noninteracting resonant peak ( the dotted line ) to correlation peaks , for ( a ) the ph symmetric siam @xmath175 , and ( b ) the ph asymmetric siam @xmath177 . the ph symmetric siam shows introducing @xmath0 leads to developing the correlation two peaks as well as the kondo peak that is suppressed by finite bias voltage . in contrast , the bias voltage effect on the ph asymmetric siam is more involved , because the kondo resonance is apparently shifted and mixed with one of the correlation peaks , eventually showing the two - peak structure at @xmath178 in large bias voltage limit . , and @xmath179 ) . other parameters are chosen as the same as in fig . [ fig : ph - symmetric ] . the dotted line correspond to the two - terminal noninteracting unbiased case , while the dashed line , to the two - terminal interacting unbiased case . ] to examine finite bias affects further and see particularly how the presence of multi - terminals affects the nonequilibrium spectral function , we configure a special setup of the multi - terminal siam that preserves the ph symmetry : the dot is connected with @xmath180 identical terminals , with bias levels being distributed equidistantly between @xmath181 and @xmath182 , and each of relaxation rates is set to be @xmath183 . the latter ensures that the unbiased spectral function is the same , hence the kondo temperature . results of the nonequilibrium spectral function are shown in fig . [ fig : nldos - by - v ] . again , we confirm that no splitting of the kondo resonance is observed in this multi - terminal setting . one sees further that the increasing the number of terminals _ enhances _ the kondo resonance . this can be understood by weakening the bias suppression effect on the kondo resonance for a larger @xmath180 . more precisely , one may estimate the suppressing effect from small bias behavior eq . . hence @xmath184 is a control parameter . in the present multi - terminal ph symmetric setting , the quantity @xmath185 is found to be @xmath186 therefore @xmath185 decreases with increasing @xmath180 , which results in weakening the suppression and enhancing the kondo resonance for a larger @xmath180 . ( @xmath187 ) . other parameters are the same as in fig . [ fig : nldos - by - v ] . ( b ) the effective fermi distribution @xmath188 at zero temperature for the ph symmetric multi - terminal dot ( @xmath189 ) . the inset shows relative locations of bias levels with fixed @xmath190 as a function of the number of terminals.,title="fig : " ] ( @xmath187 ) . other parameters are the same as in fig . [ fig : nldos - by - v ] . ( b ) the effective fermi distribution @xmath188 at zero temperature for the ph symmetric multi - terminal dot ( @xmath189 ) . the inset shows relative locations of bias levels with fixed @xmath190 as a function of the number of terminals.,title="fig : " ] the preceding argument also tells us that if the spectral function bears any multi - terminal signatures at all , they would be more conspicuous by examining it with fixing @xmath185 rather than @xmath125 . this is done in fig . [ fig : nldos - by - dmu ] ( a ) ; figure [ fig : nldos - by - dmu ] ( b ) shows how the effective dot distribution @xmath188 and the relative locations of bias levels ( in the inset ) evolve for a fixed @xmath185 when @xmath180 increases . no multi - terminal signature in the nonequilibrium spectral function is seen in fig . [ fig : nldos - by - dmu ] ( a ) ; results of different @xmath180 actually collapse , not only around zero frequency but in the entire frequency range . it suggests that the suppression of the kondo resonance deeply correlates the development of coulomb peaks , and a mixing between those spectral weight is important . the parameter @xmath185 controls a crossover from the kondo resonance to the coulomb blockade structure . we may also understand the similarity between bias effect and temperature effect by the connection through the large @xmath180 limit of the effective fermi distribution @xmath188 , as shown in fig . [ fig : nldos - by - dmu ] ( b ) . although there is a consensus that bias voltage starts suppressing the kondo peak , and eventually destroys it with developing the two coulomb peaks when bias voltage is much larger than the kondo temperature , there is a controversy as to whether the kondo resonance peak will be split or not in the intermediate range of bias voltage . all the results obtained by the second - order perturbation consistently indicate that there is no split of the kondo resonance ; finite bias voltage starts to suppress the kondo resonance , and develops the two coulomb blockade peaks by shifting the spectral weight from the kondo resonance . we should mention that some other approximations draw a different conclusion . here we make a few remarks on apparent discrepancy seen in various theoretical approaches as well as experiments , as well as some speculation based on our results . typically , several approaches that rely on the infinite @xmath0 limit , notably non - crossing approximation , equation of motion method , and other approaches investigating the kondo hamiltonian , observed the splitting of the kondo resonance under finite bias voltage @xcite . those results , however , have to be interpreted with great care , in our view . generally speaking , the spectral function obtained by those approaches does not obey the spectral weight sum rule : ignoring the doubly occupied state typically leads to the sum rule @xmath191 @xcite , rather than the correct value . therefore only half of the spectral weight can be accounted for in those methods . simultaneously , such ( false ) sum rule in conjugation with the bias - suppression of the kondo resonance can not hep but lead to a two - peak structure of the spectrum within the range of attention . splitting of the kondo resonance might be an artifact of the approximation . not fulfilling the correct sum rule , those approaches may not be able to distinguish whether finite bias will split the kondo resonance or simply suppresses it with developing the coulomb peaks . as for the two - terminal ph symmetric siam , fourth - order contribution regarding the coulomb interaction @xmath0 has been evaluated numerically @xcite . the results seem unsettled , though . while @xcite suggested the fourth - order term may yield the splitting of the kondo resonance in the intermediate bias region @xmath192 for sufficiently large interaction @xmath193 , which the second - order calculation fails to report , another numerical study indicates that the spectral function remains qualitatively the same with the second - order result @xcite . experimental situation is not so transparent , either . while the splitting of the kondo resonance was reported in a three - terminal conductance measurement in a quantum ring system @xcite , a similar spitting observed in the differential conductivity was attributed to being caused by a spontaneous formation of ferromagnetic contacts , not purely to bias effect @xcite . it is also pointed out that it has been recently recognized that the rashba spin - orbit coupling induces spin polarization nonmagnetically in a quantum ring system with a dot when applying finite bias voltage @xcite ; hence such spin magnetization might possibly lead to the splitting of the kondo resonance . the kondo resonance is a manifestation of singlet formation between the dot and the lead electrons . one may naively think that when several chemical potentials are connected with the dot , such singlet formation would take place at each lead _ separately _ , causing multiple kondo resonances . the results of the multi - terminal ph symmetric siam presented in the previous section tempt us to speculate a different picture . let us suppose that ( almost ) the same dot distribution function @xmath194 is realized for a fixed @xmath185 with a different terminal number @xmath180 , as fig . [ fig : nldos - by - dmu ] ( a ) suggests . note the assumption is fully consistent to the ward identity for low bias , but it invalidates a quasiparticle ansatz @xmath195 that explicitly depends on @xmath180 . in the large @xmath180 limit with a fixed @xmath185 , the effective fermi distribution @xmath188 resembles the fermi distribution at finite temperature @xmath196 . accordingly bias voltage may well give effect similar to finite temperature . it is seen in low and large bias limit for a dot with or without the ph symmetry . it implies that a dot electron can not separately form a singlet with the lead at each chemical potential , because it needs to implicates states at different chemical potentials through coupling with other leads . our second - order perturbation results seem to support this view . in summary , we have evaluated analytically the second - order self - energy and green function for a generic multi - terminal single impurity anderson model in nonequilibrium . various limiting behaviors have been examined analytically . nonequilibrium spectral function that preserves the current is constructed and is checked to satisfy the spectral weight sum rule . the multi - terminal effect is examined for the ph - symmetric siam , particularly . within the validity of the present approach , it is shown that the kondo peak is not split due to bias voltage . it is found that most of the finite bias effect is similar to that of finite temperature in low- and high - bias limits with and without the ph symmetry . such nature could be understood by help of the ward identity and the connection through the @xmath197 terminal limit . the present analysis serves as a viable tool that can cover a wide range of experimental situations . although there is still a chance that high - order contributions might generate a new effect such as split kondo resonances in a limited range of parameters , it is believed that the second - order perturbation theory can capture the essence of the kondo physics in most of realistic situations . moreover , having a concrete analytical form that satisfies both the current conservation and the sum rule , the present work provides a good , solid , workable result that more sophisticated future treatment can base on . the author gratefully acknowledges a. sunou for fruitful collaboration that delivered some preliminary results in this work . the author also appreciates r. sakano and a. oguri for helpful discussion at the early stage of the work . the work was partially supported by grant - in - aid for scientific research ( c ) no . 22540324 and 26400382 from mext , japan . [ sec : appendix ] we start with the nonequilibrium green function @xmath198 without the coulomb interaction on the dot . its keldysh structure is specified by @xmath199 where @xmath36 is the effective dot distribution defined in eq . . we incorporate the hartree - type diagram into the unperturbed part by replacing @xmath20 to @xmath200 . note @xmath68 is the exact dot occupation that needs to be determined consistently later . after mapping to the rak representation , those components are given by @xmath201 \left [ g^{r}_{\sigma } ( \omega ) - g^{a}_{\sigma } ( \omega ) \right ] . \label{eq : gk0}\end{aligned}\ ] ] the function @xmath202 reduces to @xmath203 at zero temperature . taking the fourier transformation of eqs . , using eq . , and making further manipulations , we can rewrite @xmath204 and @xmath205 as @xmath206^ { * } , \label{eq : pir}\\ \pi^{k}(\varepsilon ) & = 2i\sum_{a , b } \frac{\gamma_{a } \gamma_{b}}{\gamma^{2 } } \coth \tfrac{\beta(\varepsilon-\mu_{ab})}{2 } { \mathop{\mathrm{im}}\nolimits}\left [ \frac{\gamma b_{ab}(\varepsilon)}{\pi \varepsilon ( \varepsilon+2i\gamma ) } \right ] , \label{eq : pik}\end{aligned}\ ] ] where @xmath207 , @xmath208 is the inverse temperature , and @xmath209 is given by @xmath210 \!\left [ g^{a}(\varepsilon ' ) - g^{r}(\varepsilon'+\varepsilon ) \right].\end{aligned}\ ] ] in the present work , we are interested in the zero temperature limit , for which @xmath211 becomes @xmath212 . the function @xmath213 in this limit is evaluated as ( with @xmath112 ) @xmath214 this corresponds to a multi - terminal extension of the result obtained for the two - terminal ph symmetric siam . to complete evaluating the remaining integral over @xmath215 of eqs . ( [ eq : i1],[eq : i2 ] ) , we take full advantage of various properties of dilogarithm function @xmath96 . a concrete integral formula we have utilized will be given in appendix d. for the sake of completeness , we here collect its definition and properties necessary to complete our evaluation . [ [ definition ] ] definition + + + + + + + + + + dilogarithm @xmath96 with a complex argument @xmath216 is not so commonly found in literature . as it is a multi - valued function , we need to specify its branch structure properly . one way to define dilogarithm @xmath96 all over the complex plane consistently is to use the integral representation @xmath217 the multivaluedness of dilogarithm @xmath218 originates from logarithm in the integrand . here we designate the principal value of logarithm as @xmath219 , which is defined by @xmath220 according to eq . , @xmath96 has a branch - cut just above the real axis of @xmath221 . accordingly , its values just above and below the real axis are different for @xmath221 : @xmath222 but @xmath223 . some special values are known analytically . we need @xmath224 , @xmath225 , @xmath226 and @xmath227 for evaluation later . [ [ sec : functional - relations ] ] functional relations + + + + + + + + + + + + + + + + + + + + dilogarithm @xmath96 has interesting symmetric properties regarding its argument @xmath228 ; values at @xmath228 , @xmath229 , @xmath230 , @xmath231 , @xmath232 and @xmath233 are all connected with one another . those points are ones generated by symmetric operations @xmath234 and @xmath235 defined by @xmath236 and @xmath237 forms a group . other operations correspond to @xmath238 applying a series of integral transformations in eq . , one can connect the values of dilogarithm at these values with one another @xcite . note , those functional relations are usually presented only for real arguments . extending them for complex variables needs examining its branch - cut structure carefully . by following and extending the derivations in ref for complex @xmath216 , we prove that the following functional relations are valid for any complex variable @xmath228 . @xmath239^{2 } , \label{eq : dilog - func - rel - start}\\ & { \mathop{\mathrm{li}_{2}}\nolimits}(tz ) = -{\mathop{\mathrm{li}_{2}}\nolimits}(z ) + \tfrac{\pi^{2}}{6 } - { \mathop{\mathrm{log}}\nolimits}(tz ) { \mathop{\mathrm{log}}\nolimits}z , \\ & { \mathop{\mathrm{li}_{2}}\nolimits}(tstz ) = - { \mathop{\mathrm{li}_{2}}\nolimits}(z ) -\tfrac{1}{2 } { \mathop{\mathrm{log}}\nolimits}^{2}(stz ) \nonumber \\ & \qquad \qquad \qquad - [ { \mathop{\mathrm{log}}\nolimits}(tz ) + { \mathop{\mathrm{log}}\nolimits}(stz ) ] { \mathop{\mathrm{log}}\nolimits}z , \\ & { \mathop{\mathrm{li}_{2}}\nolimits}\big ( tsz \big ) = { \mathop{\mathrm{li}_{2}}\nolimits}(z ) - \tfrac{\pi^{2}}{6 } \nonumber \\ & \qquad \qquad \qquad -\tfrac{1}{2 } { \mathop{\mathrm{log}}\nolimits}^{2}(sz ) - { \mathop{\mathrm{log}}\nolimits}(sz ) { \mathop{\mathrm{log}}\nolimits}(tz ) , \\ & { \mathop{\mathrm{li}_{2}}\nolimits}\big(stz \big ) = { \mathop{\mathrm{li}_{2}}\nolimits}(z ) + \tfrac{\pi^{2}}{6 } \nonumber \\ & \qquad \qquad \qquad + \tfrac{1}{2 } { \mathop{\mathrm{log}}\nolimits}^{2}(tz ) + { \mathop{\mathrm{log}}\nolimits}(tz ) { \mathop{\mathrm{log}}\nolimits}(tstz ) . \label{eq : dilog - func - rel - end}\end{aligned}\ ] ] to our knowledge , the above form of extension of functional relations of dilogarithm has not been found in literature . [ [ the - taylor - expansion ] ] the taylor expansion + + + + + + + + + + + + + + + + + + + + to examine various limiting behaviors , we need the taylor expansion of dilogarithm , which is derived straightforwardly from eq . , @xmath240 the presence of @xmath241 reflects the branch - cut structure of @xmath96 . in particular , we utilize the following expansion in our analysis . @xmath242 here we derive and present the central integral formula for evaluating eqs . ( [ eq : i1],[eq : i2 ] ) . by performing a simple integral transformation in eq . , we have the integration @xmath243 where all the parameters @xmath244 as well as @xmath228 may be taken as complex numbers . combined with fractional decomposition , we see the following integral be evaluated in terms of dilogarithm : @xmath245 the remaining task to complete calculating @xmath71 in the form of eqs . ( [ eq : sigmar][eq : xi123 ] ) is to collect all the relevant formulas and organize them in a form that conforms to eq . . to write concisely , we introduce the following notations @xmath247 where the hartree level @xmath248 is defined as before . we express the terms @xmath249 and @xmath250 defined in eqs . ( [ eq : i1],[eq : i2 ] ) as @xmath251 here the leads @xmath95 in @xmath249 as well as @xmath252 in @xmath250 carry spin @xmath253 , while @xmath13 in @xmath250 does spin @xmath16 . singularity on energy integration is prescribed by the principal values . equation enables us to perform and express the above integrals in terms of dilogarithm . the resulting expressions are still complicated , but we can simplify them further using functional relations of dilogarithm eqs ( [ eq : dilog - func - rel - start][eq : dilog - func - rel - end ] ) . these require straightforward but rather laborious manipulations . in this way , we reach the final expression of @xmath81 of eq . .

let @xmath0 be a connected nonorientable surface of genus @xmath1 with @xmath2 punctures and let @xmath5 denote the mapping class group of @xmath0 , the group of isotopy classes of all diffeomorphisms @xmath6 . the complex of curves @xmath7 on @xmath0 is defined to be the abstract simplicial complex whose vertices are the isotopy classes of nontrivial simple closed curves . a set of vertices @xmath8 forms a @xmath9 if and only if @xmath10 have pairwise disjoint representatives . each diffeomorphism @xmath11 acts on the set of nontrivial simple closed curves preserving the disjointness of simple closed curves . it follows that the mapping class group of @xmath0 acts on @xmath7 as simplicial automorphisms . in other words , there is a natural group homomorphism @xmath12 . the purpose of this paper is to prove the following theorem . [ theorem - main1 ] let @xmath0 be a connected nonorientable surface of genus @xmath1 with @xmath2 punctures . suppose that @xmath1 is odd and @xmath13 . then the natural map @xmath14 is an isomorphism . the complex of curves on an orientable surface @xmath15 was introduced by harvey @xcite . it was shown by ivanov @xcite and korkmaz @xcite that all automorphisms of the complex of curves on @xmath15 are induced by diffeomorphisms of the surface @xmath15 , with a few exception . another proof of this result was also obtained by luo @xcite . as a consequence of this result , it was proved that any isomorphism between two subgroups of finite index in the mapping class group of @xmath15 is given by the conjugation with a mapping class ( cf . @xcite , @xcite ) . ivanov also gave another proof of the fact that the isometries of the teichmller space are induced by diffeomorphisms of @xmath15 . schaller @xcite showed that the extended mapping class group of a hyperbolic surface is isomorphic to the automorphism group of the graph . the set of vertices of the graph is the set of nonseparating simple closed geodesic and the edges consisting of pairs of those nonseparating curves satisfying the property that the two curves in each pair intersect exactly once . margalit @xcite proved that the automorphism group of the pants complex is isomorphic to the extended mapping class group . irmak @xcite defined a superinjective simplicial map and showed that a superinjective simplicial map of the complex of curves and the complex of nonseparating curves is induced by a homeomorphism of an orientable surface . irmak and korkmaz @xcite proved that the automorphism group of the hatcher - thurston complex is isomorphic to the extended mapping class group modulo its center . brendle and margalit @xcite showed that any injection of a finite index subgroup of @xmath16 , generated by the dehn twists about separating curves , into the torelli group @xmath17 of a closed orientable surface is induced by a homeomorphism , conforming a conjecture of farb that @xmath18 . behrstock and margalit @xcite proved that for a torus with at least 3 punctures or a surface of genus 2 with at most 1 puncture , every injection of a finite index subgroup of the extended mapping class group into the extended mapping class group is the restriction of an inner automorphism . all the above results are about orientable surfaces . for nonorientable surfaces we prove that the automorphism group of the complex of curves of a nonorientable surface of odd genus is isomorphic to its mapping class group . first , we show that the natural group homomorphism is injective . second , we prove that the natural group homomorphism is surjective for the punctured projective plane using the results and the ideas of korkmaz s analogous work on the punctured sphere @xcite and a result by scharlemann @xcite . for higher genus , we use induction and some ideas contained in irmak s analogous work @xcite . [ remark - evengenus ] we believe that the same result holds for nonorientable surfaces of even genus . although we have some progress in proving the even genus case , the proof is not complete yet . * acknowledgement . * this work is the author s dissertation at middle east technical university . the author would like to mustafa korkmaz , the author s thesis advisor , for his continuous guidance , encouragement , suggestions and for reviewing this manuscript . let @xmath0 be a connected nonorientable surface of genus @xmath1 with @xmath2 marked points . we call these marked points punctures . recall that the genus of a nonorientable surface is the maximum number of projective planes in a connected sum decomposition . if @xmath19 is a circle on @xmath0 , by which we mean a simple closed curve , then according to whether a regular neighborhood of @xmath19 is an annulus or a mobius strip , we call @xmath19 two - sided or one - sided simple closed curve , respectively . we say that a circle is nontrivial if it bounds neither a disc nor annulus together with a boundary component , nor a disc with one puncture , nor a mobius band on @xmath0 . if @xmath19 is a circle , then we denote by @xmath20 the surface obtained by cutting @xmath0 along @xmath19 . a circle @xmath19 is called nonseparating if @xmath20 is connected and separating otherwise . if @xmath19 is separating , then @xmath20 has two connected components . if @xmath19 is separating and if one of the components of @xmath20 is a disc with @xmath21 punctures , then we say that @xmath19 is a @xmath21-separating circle . we denote circles by the lowercase letters @xmath22 and their isotopy classes by @xmath23 . an embedded arc connecting a puncture to itself or two different punctures will be denoted by @xmath24 and their isotopy classes by @xmath25 . let @xmath26 be the isotopy class of a circle @xmath19 . we say that @xmath26 is nonseparating ( respectively separating ) if @xmath19 is nonseparating ( respectively separating ) . similarly , we say that @xmath26 is one - sided , two - sided or @xmath21-separating vertex if @xmath19 is one - sided , two - sided or @xmath21-separating circle , respectively . the geometric intersection number @xmath27 of two isotopy classes @xmath26 and @xmath28 is defined to be the infimum of the cardinality of @xmath29 with @xmath30 , @xmath31 . the geometric intersection numbers @xmath32 and @xmath33 are defined similarly . the following lemma is proved in @xcite . [ flp ] let @xmath15 be a sphere with @xmath34 punctures . then * up to isotopy , there exists a unique nontrivial embedded arc joining a puncture @xmath35 to itself , or @xmath35 to another puncture @xmath36 , * all circles on @xmath15 are trivial . an abstract simplicial complex is defined as follows ( cf . @xcite ) : let @xmath37 be a nonempty set . an abstract simplicial complex @xmath38 with vertices @xmath37 is a collection of nonempty finite subsets of @xmath37 , called simplices , such that if @xmath39 , then @xmath40 , and if @xmath41 and @xmath42 is a nonempty subset of @xmath37 , then @xmath43 . the dimension dim@xmath44 of a simplex @xmath44 is card@xmath45 , where card@xmath44 is the cardinality of @xmath44 . a simplex @xmath44 is called a @xmath46-simplex if dim@xmath47 . the supremum of the dimension of the simplices of @xmath38 is called the dimension of @xmath38 . a subcomplex @xmath48 of an abstract simplicial complex @xmath38 is called a full subcomplex if whenever a set of vertices of @xmath48 is a simplex in @xmath38 , it is also a simplex in @xmath48 . the complex of curves @xmath49 on an orientable surface @xmath15 is the abstract simplicial complex whose vertices are the isotopy classes of nontrivial simple closed curves . similarly , the complex of curves @xmath7 on a nonorientable surface @xmath0 is the abstract simplicial complex whose vertices are the isotopy classes of nontrivial simple closed curves . in this complex of curves , we take one - sided vertices as well as two - sided vertices . clearly , the complex of curves of a surface of genus @xmath1 with @xmath2 punctures and with @xmath50 boundary components , and that of a surface of genus @xmath1 with @xmath51 punctures are isomorphic . therefore , sometimes we regard boundary components and the punctures the same . two distinct vertices @xmath52 are joined by an edge if and only if their geometric intersection number is zero . more generally , a set of vertices @xmath53 forms a @xmath46-simplex if and only if @xmath54 for all @xmath55 . clearly , the dimension of @xmath7 is @xmath56 if @xmath0 is a projective plane with @xmath2 punctures . if @xmath15 is a sphere with @xmath2 punctures , then the dimension of @xmath49 is @xmath57 . for higher genus , if @xmath0 is a connected nonorientable surface of genus @xmath58 with @xmath2 punctures such that the euler characteristic of @xmath0 is negative and if @xmath59 , then the dimension of @xmath7 is @xmath60 and if @xmath61 , then the dimension of @xmath7 is @xmath62 ( see section2.5 ) . if @xmath15 is a connected orientable surface of genus @xmath1 with @xmath2 punctures such that @xmath63 , then the dimension of @xmath49 is @xmath64 . let @xmath26 be a vertex in the complex of curves . we define the link @xmath65 of @xmath26 to be the full subcomplex of the complex of curves whose vertices are those of the complex of curves which are joined to @xmath26 by an edge in the complex of curves . the dual link @xmath66 of @xmath26 is the graph whose vertices are those of @xmath65 such that two vertices of @xmath66 are joined by edge if and only if they are not joined by an edge in the complex of curves ( or in @xmath65 ) . a pentagon is an ordered five - tuple @xmath67 , defined up to cyclic permutations and inversion , of vertices of the complex of curves such that @xmath68 for @xmath69 and @xmath70 otherwise . obviously , if @xmath15 is a sphere with @xmath71 , then there are no nontrivial circles on @xmath15 . therefore , @xmath49 is empty . if @xmath15 is a sphere with four punctures , then @xmath49 is infinite discrete . if @xmath0 is a projective plane or a projective plane with one puncture then @xmath7 consists of a unique vertex . if @xmath0 is a projective plane with two punctures , then @xmath7 is finite ( cf . it consists of two vertices , the isotopy classes of the circles @xmath72 and @xmath73 of the figure[2puncture ] . , width=377 ] we now define another abstract simplicial complex @xmath74 as follows . the vertices of @xmath74 are the isotopy classes of nontrivial embedded arcs on @xmath0 connecting two punctures . a set of vertices of @xmath74 forms a simplex if and only if these vertices have pairwise disjoint representatives . if @xmath19 is @xmath75-separating circle on @xmath0 , there exists up to isotopy a unique nontrivial embedded arc @xmath76 on the twice - punctured disc component of @xmath20 joining two punctures by lemma1.1 in @xcite and in @xcite . on the other hand , an arc @xmath76 connecting two different punctures of @xmath0 determines uniquely a @xmath75-separating circle up to isotopy , that is , the boundary of a regular neighborhood of the arc . so , we have a one - to - one correspondence between the set of @xmath75-separating isotopy classes and the set of isotopy classes of embedded arcs connecting two different punctures . if @xmath0 is a projective plane with one puncture , then @xmath74 consists of a unique vertex . scharlemann studied the arc complex of a twice - punctured projective plane . he showed that the vertices of the arc complex consist of the isotopy classes of arcs shown in the figure[2puncarc ] . , width=453 ] if @xmath19 and @xmath50 are two @xmath75-separating circles , and @xmath26 , @xmath28 are their isotopy classes , such that the corresponding arcs @xmath76 and @xmath77 can be chosen disjoint with exactly one common endpoint @xmath35 , then we say that @xmath19 and @xmath50 constitute a simple pair of circles and denote it by @xmath78 ( see figure[chain](a ) ) . similarly , we say that @xmath79 is a simple pair of arcs . we also say @xmath80 and @xmath81 simple pairs . let @xmath82 be embedded pairwise disjoint arcs , @xmath83 and @xmath84 the endpoints of @xmath85 , with @xmath86 for @xmath87 , @xmath88 . therefore , @xmath89 is a simple pair of arcs for each @xmath90 . let @xmath91 be the corresponding circles . we say that @xmath92 is a chain of arcs ( see figure[chain](b ) ) . similarly , @xmath93 is a chain of circles . we recall that the maximum number of disjoint pairwise nonisotopic nontrivial circles on a connected orientable surface @xmath15 of genus @xmath94 with @xmath2 boundary components is @xmath95 , whenever the euler characteristic @xmath96 of @xmath15 is negative . [ prop - odim1 ] let @xmath15 be a connected orientable surface of genus @xmath1 with @xmath2 punctures . suppose that @xmath63 . then all maximal simplices in @xmath49 have the same dimension @xmath64 . [ prop - odim2 ] let @xmath0 be a real projective plane with @xmath97 punctures . all maximal simplices in @xmath7 have the same dimension @xmath56 . let @xmath98 . the complex @xmath7 consists of only two vertices , hence all simplices are of dimension @xmath99 . let @xmath100 . we consider a maximal simplex @xmath44 of dimension @xmath101 . hence , @xmath44 contains @xmath102 elements only one of which is a one - sided vertex . by cutting @xmath0 along this one - sided simple closed curve , we get sphere with @xmath103 punctures . by lemma [ prop - odim1 ] , all maximal simplices in the complex of curves on the sphere with @xmath103 punctures have the same dimension @xmath104 . it follows that @xmath105 . [ prop - ndim1 ] let @xmath0 be a connected nonorientable surface of genus @xmath106 with @xmath2 punctures such that the euler characteristic of @xmath0 is negative . let @xmath107 and @xmath108 if @xmath109 , and @xmath110 and @xmath111 if @xmath112 . then there is a maximal simplex of dimension @xmath101 in @xmath7 if and only if @xmath113 . for each integer @xmath101 satisfying @xmath114 , the maximal simplicies are shown in figure[maxsimplex0 ] and figure[maxsimplex1 ] for closed nonorientable surface of odd genus . one can draw similar figures for nonorientable surface of odd genus with punctures and nonorientable surface of even genus . moreover , figure[maxsimplex ] helps to see the maximal simplex of dimension @xmath101 between @xmath115 and @xmath116 . we now prove the converse . let us consider a maximal simplex @xmath44 of dimension @xmath101 . hence , @xmath44 contains @xmath117 elements . choose pairwise disjoint simple closed curves representing elements of @xmath44 , and let @xmath118 denote the surface obtained by cutting @xmath0 along these simple closed curves . suppose that the number of one - sided simple closed curves is @xmath119 , so that we have @xmath120 two - sided elements in @xmath44 . the surface @xmath118 is a disjoint union of @xmath21 pair of pants for some positive integer @xmath21 . by the euler characteristic argument , it can be seen that @xmath121 . the number of boundary components and punctures on @xmath118 is @xmath122 . by counting the contribution of one - sided curves and two - sided curves to the boundary of @xmath118 , one can easily see that @xmath123 suppose first that @xmath109 . in this case @xmath124 . from the equality ( [ eqn3k ] ) , it is easy to see that @xmath119 must be odd and @xmath125 . suppose now that @xmath112 . in this case @xmath126 . from the equality ( [ eqn3k ] ) , it is easy to see that @xmath119 must be even and @xmath127 . the proposition follows from these . let @xmath128 be the subgroup of mapping class group of @xmath0 such that @xmath128 is generated by the dehn twist of two - sided nonseperating circles as below shown in figure[surf0-circles ] . [ prop - centralizer ] let @xmath0 be a connected nonorientable surface of odd genus @xmath1 and @xmath129 . suppose that @xmath130 is a collection of two - sided nonseparating circles in figure[surf0-circles ] such that @xmath128 is generated by the dehn twist @xmath131 along the circles @xmath132 of @xmath130 . then the centralizer @xmath133 is trivial . let @xmath134 \in \mathcal{c}_{\mathcal m_{n}}(\mathcal{t'})$ ] . unoriented circles will be denoted by @xmath135 . then , @xmath136 for @xmath137 . let @xmath19 be a one - sided circle such that @xmath20 is an orientable surface . we observe that @xmath138 . because there is one isotopy class of one - sided circle which does not intersect circles @xmath132 in @xmath130 . therefore , by cutting @xmath0 along @xmath19 , we get diffeomorphism @xmath139 . moreover , @xmath140 is orientation preserving . to see this , assume that @xmath141 is dual to @xmath132 , we know that @xmath140 preserves the orientation of a regular neighborhood of @xmath132 if and only if it preserves the orientation of @xmath141 . recall that tubular neighborhood of @xmath142 is a torus with one boundary component . since the product of the orientations of @xmath132 and @xmath141 gives the orientation of the torus , @xmath140 preserves the orientation of tubular neighborhood . using this argument for these circles in figure[surf0-circles ] , we deduce that @xmath140 preserves the orientation of the surface @xmath20 . since @xmath143 in @xcite , @xmath140 is isotopic to identity . since @xmath144 , we obtain that @xmath145 is isotopic to identity on @xmath0 . hence , @xmath146 . we first show that the natural map @xmath12 is injective . [ lemma - injec0 ] let @xmath0 be a connected nonorientable surface of genus @xmath1 with @xmath2 punctures . suppose that @xmath148 . then the natural map @xmath14 is injective . suppose that @xmath134\in \mathcal m_n$ ] acts trivially on @xmath7 . then @xmath149 , @xmath150 , @xmath151 ; i.e. , @xmath152 for all @xmath153 . first , suppose that @xmath109 is odd . let @xmath19 be a one - sided circle such that @xmath20 is orientable surface of genus @xmath94 with @xmath103 boundary components . we denote by @xmath26 the isotopy class of @xmath19 . then @xmath154 . this implies @xmath155 is isotopic to @xmath19 . then there exists a diffeomorphism @xmath156 , @xmath157 . let @xmath158 . we observe that @xmath159 and @xmath160 . therefore , we have a diffeomorphism @xmath161 such that @xmath162 . using ivanov s theorem in @xcite ( for @xmath163 and @xmath164 , using theorem1 in @xcite ) , we get @xmath165 . since @xmath159 , we see that @xmath166 on @xmath0 . in other words , it descends to a diffeomorphism of @xmath0 . so , we have @xmath167 . since @xmath156 , we get @xmath168 . hence , the natural map @xmath12 is injective . now , suppose that @xmath169 . let @xmath50 be a nonseparating two - sided circle such that @xmath170 is orientable surface of genus @xmath94 with @xmath171 boundary components . let @xmath28 be the isotopy class of @xmath50 . then @xmath172 . this implies @xmath173 is isotopic to @xmath50 . then there exists a diffeomorphism @xmath156 , @xmath174 . let @xmath158 . we see that @xmath175 and @xmath160 . therefore , we have a diffeomorphism @xmath176 such that @xmath162 . using ivanov s theorem in @xcite ( for @xmath163 and @xmath164 , using theorem1 in @xcite ) , we get @xmath165 . since @xmath175 , @xmath166 on @xmath0 . in other words , it descends to a diffeomorphism of @xmath0 . therefore , we obtain that @xmath167 . since @xmath156 , we get @xmath168 . hence , the natural map @xmath12 is injective . throughout this section unless otherwise stated , @xmath0 will denote a real projective plane with @xmath178 punctures . we need at least @xmath179 punctures for the proof of the lemma [ lemma - indep ] . in this section , we will prove that the natural homomorphism @xmath180 is surjective . hence it will be an isomorphism . for this , we first prove that automorphisms of @xmath7 preserve the topological type of the vertices of @xmath7 and that certain pairs of vertices of @xmath7 can be realized in the complex of curves . we conclude that automorphisms of @xmath7 preserve these pairs of vertices . next , we show that every automorphism of @xmath7 induces an automorphism of the complex @xmath74 in a natural way . the automorphisms of @xmath74 are determined by their action on a maximal simplex . then , we use the relation between maximal simplices of @xmath74 and isotopy classes of ideal triangulations of @xmath0 and in conclusion , we show that an automorphism of @xmath74 induced by some automorphism of @xmath7 agrees with a mapping class . we remind that if @xmath0 is a projective plane with @xmath181 punctures , then up to diffeomorphism there is only one nonseparating one - sided circle and also there is no nonseparating two - sided circle . the other circles are @xmath21-separating for some @xmath21 . [ lemma - iso ] let @xmath181 and @xmath182 . if @xmath0 is a projective plane with @xmath2 punctures and @xmath15 is a sphere with @xmath21 punctures , then @xmath7 and @xmath49 are not isomorphic . the complexes @xmath7 and @xmath49 have dimensions @xmath56 and @xmath183 , respectively . if @xmath184 , since these complexes of curves have different dimensions , @xmath7 and @xmath49 are not isomorphic . if @xmath185 , we proceed as follows . let @xmath186 . then @xmath7 is finite ( see @xcite ) , however , @xmath49 is infinite discrete since @xmath15 is a sphere with @xmath187 punctures . therefore , they are not isomorphic . now , assume that @xmath188 and @xmath7 and @xmath49 are not isomorphic when @xmath0 has @xmath189 punctures . we need to show that these complexes are not isomorphic if @xmath0 has @xmath2 punctures . assume that there is an isomorphism @xmath190 . note that for a vertex @xmath191 of @xmath7 , the dual link @xmath192 of @xmath191 is connected if and only if @xmath191 is either one - sided or @xmath75-separating . for a vertex @xmath193 of @xmath49 , the dual link @xmath194 of @xmath193 is connected if and only if @xmath193 is @xmath75-separating . from this , it follows that the image of the union of the set of one - sided vertices and the set of @xmath75-separating vertices of @xmath7 is precisely the set of @xmath75-separating vertices of @xmath49 . let @xmath191 be a @xmath75-separating vertex of @xmath7 . then @xmath195 takes @xmath191 to a @xmath75-separating vertex @xmath193 of @xmath49 and induces an isomorphism @xmath196 . clearly , @xmath197 is isomorphic to the complex of curves of a real projective plane with @xmath189 punctures and @xmath198 is isomorphic to the complex of curves of a sphere with @xmath103 punctures . by assumption , these complexes are not isomorphic . we get a contradiction . hence , @xmath7 and @xmath49 are not isomorphic . [ theorem - prese ] the group @xmath199 preserves the topological type of the vertices of @xmath7 . let @xmath195 be an automorphism of @xmath7 . note that for a vertex @xmath191 of @xmath7 , the dual link @xmath192 of @xmath191 is connected if and only if @xmath191 is either one - sided vertex or @xmath75-separating . therefore , @xmath195 can not take a one - sided vertex or a @xmath75-separating to a @xmath21-separating vertex with @xmath200 . assume that @xmath201 is a @xmath75-separating vertex for some one - sided vertex @xmath26 . let @xmath30 be a circle . then @xmath20 is disc with @xmath2 punctures . let @xmath31 . then @xmath170 is homeomorphic to @xmath177 with @xmath56 punctures and with one boundary component . clearly , @xmath65 is isomorphic to @xmath202 . similarly , @xmath203 is isomorphic to @xmath204 . since the complexes @xmath202 and @xmath204 are not isomorphic by lemma [ lemma - iso ] , @xmath28 can not be @xmath75-separating . it also follows from this that @xmath195 maps a @xmath75-separating vertex to a @xmath75-separating vertex . let @xmath191 be a @xmath21-separating vertex for some @xmath205 . then @xmath206 is an @xmath207-separating vertex for some @xmath208 . we must show that @xmath209 . let @xmath210 and @xmath211 be circles . @xmath212 and @xmath213 have two connected components . let @xmath214 and @xmath215 such that @xmath216 and @xmath217 have genera @xmath218 . the dual link @xmath192 has exactly two connected components . let us denote these components by @xmath219 and @xmath220 ; @xmath221 . we name @xmath222 so that the vertices of @xmath223 are the isotopy classes of circles on @xmath216 . let @xmath222 be the full subcomplex of @xmath7 with vertices @xmath223 . it follows that @xmath224 . then @xmath222 is isomorphic to @xmath225 . clearly , the dimension of @xmath226 is @xmath227 since @xmath228 is a sphere with punctures and the dimension of @xmath229 is @xmath230 . similarly , we define @xmath231 and @xmath232 so that @xmath232 is isomorphic to @xmath233 . if @xmath234 , then @xmath195 restricts to an isomorphism @xmath235 , which induces an isomorphism @xmath236 . then either @xmath237 or @xmath238 . hence , @xmath239 or @xmath240 . however , @xmath240 is not possible by lemma [ lemma - iso ] . therefore , @xmath241 . it follows that their dimensions are equal : @xmath242 . hence , @xmath209 . the proof of the theorem is now complete . [ korkmaz - penta ] let @xmath26 and @xmath28 be two @xmath75-separating vertices of the complex of curves @xmath49 . then @xmath243 is a simple pair if and only if there exist vertices @xmath244 of @xmath49 satisfying the following conditions . * @xmath245 is a pentagon in @xmath49 , * @xmath246 and @xmath247 are @xmath75-separating , @xmath248 is @xmath34-separating , and @xmath249 and @xmath250 are @xmath21-separating for @xmath251 , * @xmath252 , @xmath253 , @xmath254 and @xmath255 are codimension - zero simplices , where @xmath256 . [ theorem - penta ] let @xmath26 and @xmath28 be two @xmath75-separating vertices of @xmath7 . then @xmath243 is a simple pair if and only if there exist vertices @xmath257 of @xmath7 satisfying the following conditions . * @xmath245 is a pentagon in @xmath7 , * @xmath246 is @xmath75-separating , @xmath248 is @xmath34-separating , and @xmath249 is @xmath21-separating for @xmath258 , @xmath193 is one - sided , * @xmath259 , @xmath260 , @xmath261 and @xmath262 are codimension - zero simplices , where @xmath263 . suppose that @xmath264 is a simple pair . let @xmath30 and @xmath31 so that @xmath265 is a simple pair . it is clear that any two simple pairs of circles are topologically equivalent , that is ; if @xmath266 is any other simple pair , then there exists a diffeomorphism @xmath267 such that @xmath268 . so , we can assume that @xmath19 and @xmath50 are the circles illustrated in figure[penta ] . in the figure , we think of the sphere as the one point compactification of the plane and the cross inside the circle means that we delete open disc and identify the antipodal boundary points so that we get a real projective plane with punctures . the isotopy classes @xmath269 of the circles @xmath132 and the isotopy class @xmath193 of the one - sided circle @xmath270 satisfy ( i)-(iii ) . now , we prove the converse . assume that conditions ( i)-(iii ) above hold . let @xmath211 be a one - sided circle . deleting @xmath271 from conditions ( ii ) and ( iii ) , we have codimension - one simplices @xmath252 , @xmath253 , @xmath254 and @xmath255 . however , these simplicies are codimension - zero simplices in the complex @xmath272 . by theorem[korkmaz - penta ] , we see that @xmath243 is a simple pair on the sphere and @xmath213 is a sphere with @xmath273 punctures . let us say that a puncture is inside @xmath19 if it is one of the two punctures on the disc bounded by @xmath19 . similarly for @xmath50 . there are three possibilities the boundary component @xmath274 ( we see it as a puncture from point of view of the curve complex ) resulting from cutting along @xmath270 , as illustrated in figure[pair ] . in the figure , this boundary component @xmath274 is drawn as an oval . the case are : @xmath275 it may be outside of both @xmath19 and @xmath50 , @xmath276 it may be inside , say , @xmath19 and outside of @xmath50 , or @xmath277 it may be the unique puncture inside both @xmath19 and @xmath50 . after the identifying the antipodal boundary points of the oval , in order to get both @xmath19 and @xmath50 as @xmath75-separating circles , we must have the case in figure[pair](i ) . hence , we see that @xmath243 is a simple pair on @xmath0 as well . [ corollary - chain ] let @xmath195 be an automorphism of @xmath7 . if @xmath243 is a simple pair , so is @xmath278 . similarly , the image of a chain in @xmath7 under @xmath195 is also a chain . the conditions ( i ) and ( iii ) of theorem[theorem - penta ] are invariant under the automorphisms of @xmath7 . by theorem[theorem - prese ] , the condition ( ii ) is also invariant under the automorphisms of @xmath7 . corollary follows from these . we can define an action of the group @xmath279 on punctures of @xmath0 as follows . for @xmath280 and for a puncture @xmath35 of @xmath0 , choose two isotopy classes @xmath281 , @xmath282 of embedded arcs such that @xmath283 is a simple pair with the common endpoint @xmath35 . define @xmath284 to be the common endpoint of the simple pair @xmath285 . we note that by the one - to - one correspondence between the set of @xmath75-separating vertices of @xmath7 and the set of those vertices of @xmath74 which join different punctures , @xmath279 has a well - defined action on the latter set . next two lemmas below can be proved in the same way as lemma3.5 and lemma3.6 in @xcite . [ lemma - indep ] the definition of the action of @xmath199 on the punctures of @xmath0 is independent of the choice of the simple pair . [ lemma - punc ] let @xmath286 , @xmath26 be a @xmath21-separating vertex of @xmath7 and @xmath30 . if @xmath287 denotes the @xmath21-punctured disc component of @xmath20 and @xmath288 denotes @xmath289-punctured @xmath177 with one boundary component , then @xmath290 and @xmath291 . we define an action of @xmath279 on the vertices of @xmath74 as follows . let @xmath280 , @xmath281 a vertex of @xmath74 and let @xmath292 . if @xmath76 is joining two different punctures , then @xmath293 is already defined by the correspondence between the @xmath75-separating vertices of @xmath7 and the action of @xmath279 on @xmath7 . the other words , @xmath293 is the isotopy class of the arc , which is unique up to isotopy , joining two punctures on the twice - punctured disc component of @xmath294 for @xmath295 . suppose now that the arc @xmath76 joins a puncture @xmath35 to itself and it is a two - sided loop . then the action is same as in @xcite . to be more precise , let @xmath296 and @xmath297 be the boundary components of a regular neighborhood of @xmath298 and @xmath299 and @xmath300 be their isotopy classes , respectively . since @xmath76 can not be deformed to @xmath35 , at most one of @xmath296 and @xmath297 is trivial . if @xmath296 is trivial , then @xmath296 bounds either a disc with one puncture or a mbius band . if it bounds a disc with a puncture @xmath36 , then @xmath297 bounds a disc with two punctures @xmath35 and @xmath36 . by theorem [ theorem - prese ] , @xmath301 is @xmath75-separating . so for a representative @xmath302 of @xmath301 , one of the components , say @xmath303 of @xmath304 is a twice - punctured disc , with punctures @xmath284 and @xmath305 by lemma [ lemma - punc ] . define @xmath293 to be the isotopy class of a nontrivial simple arc on @xmath303 joining @xmath284 to itself . such an arc is unique up to isotopy by lemma [ flp ] . if @xmath296 bounds a mbius band , then @xmath297 bounds a projective plane with one puncture @xmath35 . by theorem [ theorem - prese ] , @xmath301 bounds a projective plane with one puncture @xmath284 . therefore , for a representative @xmath302 of @xmath301 , one of components , say @xmath306 , of @xmath304 is a projective plane with one puncture and one boundary component . the puncture on @xmath306 is @xmath284 by lemma [ lemma - punc ] . there is only one nontrivial two - sided loop ( arc ) joining a puncture @xmath284 to itself by example [ example - arcs ] . we define @xmath293 to be the isotopy class of this two - sided loop ( arc ) joining @xmath284 to itself . if neither of @xmath307 and @xmath308 is trivial , then @xmath296 and @xmath297 bound an annulus with a puncture @xmath35 . we claim that @xmath309 and @xmath302 also bound a once - punctured annulus with only one puncture @xmath284 . here , @xmath310 is a representative of @xmath311 for @xmath312 . to see this , let @xmath313 be the subsurface of @xmath0 bounded by @xmath314 not containing the puncture @xmath35 . similarly we define @xmath315 to be the component of @xmath316 not containing @xmath284 . now , assume that the set of punctures on @xmath317 and @xmath318 are @xmath319 and @xmath320 , respectively . then @xmath321 for all @xmath322 . by lemma [ lemma - punc ] , @xmath323 and @xmath324 . we deduce that since @xmath309 and @xmath302 are disjoint and nonisotopic , they must bound an annulus with only one puncture @xmath284 . the class @xmath293 is defined to be the isotopy class of the unique arc up to isotopy on this annulus joining @xmath284 to itself . suppose finally that @xmath76 is a one - sided loop ( arc ) joining a puncture @xmath35 to itself . let @xmath19 be the boundary component of a regular neighborhood of @xmath325 and let @xmath26 be the isotopy class of @xmath19 . the circle @xmath19 bounds a mbius band with a puncture @xmath35 . by theorem [ theorem - prese ] , @xmath326 bounds a mbius band @xmath327 with a puncture , say @xmath36 . by lemma [ lemma - punc ] , @xmath328 . by example [ example - arcs ] , there is up to isotopy a unique one - sided loop @xmath77 on @xmath327 joining @xmath284 to itself . we define @xmath293 to be the isotopy class of @xmath77 . [ lemma - disjoint ] let @xmath195 be an automorphism of @xmath7 and @xmath281 and @xmath282 be two distinct vertices of @xmath74 such that @xmath329 . then @xmath330 . therefore , every automorphism of @xmath7 yields an automorphism of @xmath74 . let @xmath76 and @xmath77 be two disjoint representatives of @xmath281 and @xmath282 , respectively . there are thirteen possible cases as illustrated in figure[puncture ] . in each figure , we assume that the arc on the left is @xmath76 and the other is @xmath77 . if @xmath76 ( respectively @xmath77 ) is joining two different punctures , we denote by @xmath26 ( respectively @xmath28 ) the @xmath75-separating vertex of @xmath7 corresponding to @xmath281 ( respectively @xmath282 ) , and by @xmath19 ( respectively @xmath50 ) a representative of @xmath26(respectively @xmath28 ) . if @xmath76 ( respectively @xmath77 ) is a two - sided arc joining a puncture @xmath35 to itself , we denote by @xmath296 and @xmath297 ( respectively @xmath331 and @xmath332 ) the boundary components of a regular neighborhood of @xmath298 ( respectively @xmath333 ) . if a regular neighborhood of @xmath298 ( respectively @xmath334 ) is a once - punctured mbius band @xmath327 , we denote the boundary component of @xmath327 by @xmath335 ( respectively @xmath336 ) . we also denote representatives of @xmath337 , @xmath293 , @xmath338 , @xmath301 , @xmath339 by @xmath340 , @xmath341 , @xmath309 , @xmath302 , @xmath342 , respectively . if , say , @xmath307 is trivial , then @xmath307 bounds either a disc with a puncture or a mbius band . in the first case , we think of @xmath307 as the puncture its bounds . in the second case , we think of @xmath307 as the core of the mbius band it bounds . @xmath343 has a well defined action on the isotopy classes of these trivial simple closed curves . we now examine each of the thirteen cases . the proof is similar to that of lemma3.7 in @xcite for ( i ) , ( ii ) , ( iv ) , ( vii ) , ( ix ) , ( xi ) and we will not repeat them here . \(iii ) in this case , @xmath342 bounds a mbius band @xmath327 with a puncture and @xmath344 bounds a disc @xmath345 with two punctures . since @xmath195 is an automorphism , and since @xmath335 and @xmath50 are disjoint and nonisotopic , @xmath342 and @xmath344 are disjoint and nonisotopic . then @xmath327 does not intersect @xmath345 . since @xmath341 is on @xmath327 and @xmath346 is on @xmath345 , it follows that @xmath341 is disjoint from @xmath346 . \(v ) the once - punctured annulus bounded by @xmath331 and @xmath332 and the once - punctured mbius band bounded by @xmath335 are disjoint . since @xmath331 , @xmath332 and @xmath335 are pairwise disjoint , so are @xmath347 , @xmath348 and @xmath342 . so the annulus @xmath349 bounded by @xmath347 , @xmath348 , and the once - punctured mbius band @xmath327 bounded by @xmath342 are disjoint . since @xmath341 is on @xmath327 and @xmath346 is on @xmath349 , they are disjoint . \(vi ) this case follows from corollary[corollary - chain ] . \(viii ) suppose that @xmath76 and @xmath77 are joining the punctures @xmath35 and @xmath36 . let @xmath350 be any other puncture and let @xmath351 be an arc from @xmath35 to @xmath350 disjoint from @xmath352 . let @xmath345 denote a regular neighborhood of @xmath353 , so that @xmath345 is a disc with three punctures . let @xmath274 denote the unique arc on @xmath345 joining @xmath36 to @xmath350 such that @xmath274 does not intersect @xmath354 . let @xmath355 be the boundary of @xmath345 . so any two arcs in the set @xmath356 is a simple pair . thus any two arcs in the set @xmath357 is a simple pair , and @xmath346 , @xmath358 and @xmath359 are contained in the three punctured disk component of @xmath360 . it follows that any arc disjoint from @xmath358 and @xmath359 is also disjoint from @xmath346 . since @xmath341 is disjoint from @xmath358 and @xmath359 by ( vi ) , it is also disjoint from @xmath346 . \(x ) let @xmath35 be the common endpoint of two - sided loop @xmath76 and one - sided loop @xmath77 , so that both arcs connect @xmath35 to itself . by theorem [ theorem - prese ] and lemma [ lemma - punc ] , @xmath309 and @xmath302 are boundaries of an annulus with one puncture @xmath284 . since @xmath346 is disjoint from @xmath302 , @xmath346 is also disjoint from @xmath341 . \(xii ) assume that @xmath76 is joining @xmath35 to itself such that @xmath76 is a one - sided loop , and @xmath77 is connecting @xmath35 to @xmath36 . let @xmath361 be the punctures other than @xmath35 and @xmath36 . choose a chain @xmath362 such that @xmath363 joins @xmath364 to @xmath83 for @xmath365 , where @xmath366 . we consider a two - sided loop @xmath274 joining @xmath35 to @xmath35 disjoint from @xmath76 , @xmath77 and @xmath367 such that one of the components of the complement of @xmath274 is a mbius band and the other is a disc with punctures @xmath368 . by ( vi ) , @xmath369 is also chain and disjoint from @xmath346 . by ( xi ) , @xmath359 is disjoint from @xmath346 and the chain @xmath369 . note that one of the components of the complement of @xmath359 contains @xmath346 and the chain @xmath369 . since @xmath341 is disjoint from @xmath359 by ( x ) , we obtain that @xmath341 is also disjoint from @xmath346 . \(xiii ) let @xmath35 be the common endpoint one - sided loops @xmath76 and @xmath77 . the complement of a regular neighborhood of @xmath370 is the union of two discs @xmath371 and @xmath372 with @xmath373 and @xmath21 punctures for some @xmath21 with @xmath374 . let @xmath375 be a chain on @xmath371 so that each @xmath376 is disjoint from @xmath352 . let @xmath377 be the punctures on @xmath372 . we can choose pairwise disjoint arcs @xmath363 connecting @xmath364 and @xmath83 such that each @xmath363 is also disjoint from @xmath378 , where @xmath379 and @xmath380 . it follows that a regular neighborhood of @xmath381 is a mbius band with @xmath382 punctures . then a regular neighborhood of @xmath383 is also a mbius band with @xmath382 punctures . now the surface obtained from @xmath0 by cutting along @xmath384 and @xmath385 is an annulus @xmath349 . the puncture @xmath284 gives rise to two punctures @xmath386 and @xmath387 on the same component of @xmath388 . the arcs @xmath341 and @xmath346 live on @xmath349 . in order to get a one - sided arc , each must connect @xmath386 to @xmath387 . up to isotopy there are two arcs from @xmath386 and @xmath387 which are disjoint . these two arcs must be @xmath341 and @xmath346 . [ lemma - injective ] the natural map @xmath389 is injective . from the lemma above , every element of @xmath279 yields an element of @xmath390 . this gives a homomorphism . now , we need to show that the kernel of this homomorphism is trivial . in other words , if an automorphism of @xmath7 induces the identity automorphism of @xmath74 , then this automorphism must be the identity . let @xmath195 be an automorphism of @xmath7 . suppose that @xmath195 induces the identity automorphism of @xmath74 . we recall that there is a one - to - one correspondence between @xmath75-separating vertices of @xmath7 and the vertices of @xmath74 joining different punctures . it follows that @xmath195 is the identity on @xmath75-separating vertices of @xmath7 . let @xmath26 be a one - sided vertex of @xmath7 and let @xmath391 . let us denote by @xmath392 the punctures on the connected component of @xmath20 . let us take any chain @xmath393 disjoint from @xmath19 such that @xmath394 connects @xmath83 to @xmath84 . let @xmath395 be the isotopy class of @xmath394 . since @xmath396 , we have @xmath397 and hence @xmath398 . let @xmath399 . the surface @xmath400 obtained from @xmath0 by cutting along @xmath401 is a projective plane with one boundary component . up to isotopy , there is only one one - sided simple closed curve on @xmath400 . both @xmath26 and @xmath337 are on @xmath400 . so , we must have @xmath402 . let @xmath26 be a @xmath21-separating vertex of @xmath7 with @xmath258 and let @xmath30 . let us denote by @xmath403 and @xmath404 the punctures on the two connected components of @xmath20 . let us take any two chains @xmath405 and @xmath406 disjoint from @xmath19 such that @xmath407 connects @xmath83 to @xmath84 and @xmath408 connects @xmath409 to @xmath410 . let @xmath411 and @xmath412 be the isotopy classes of @xmath407 and @xmath408 , respectively . let @xmath274 be a one - sided loop joining to @xmath413 to itself disjoint from the other arcs . let @xmath414 be the isotopy class of @xmath274 . because @xmath415 , @xmath416 and @xmath417 , we have @xmath418 and @xmath419 . then @xmath420 and @xmath421 . let @xmath422 . the surface @xmath423 obtained from @xmath0 by cutting along @xmath424 is an annulus . since , up to isotopy , there is only one two - sided simple closed curve on @xmath423 , we must have @xmath402 . all maximal simplices in @xmath74 have the same dimension , and there is a well - defined action of the group @xmath390 on maximal simplices . any realization of a maximal simplex is an ideal triangulation of @xmath0 . an ideal triangulation is a triangulation of @xmath0 whose vertex set is the set of punctures on @xmath0 in the sense that vertices of a triangle can coincide as can a pair of edges . note that isotopy class of any ideal triangulation forms a maximal simplex in @xmath74 . the converse of that is also true . so , @xmath390 acts on the isotopy classes of ideal triangulations . we quote the following definition from @xcite . a good ideal triangle is a set @xmath425 of nontrivial embedded disjoint arcs such that @xmath76 , @xmath77 and @xmath351 connect @xmath426 to @xmath427 , @xmath427 to @xmath428 and @xmath428 to @xmath426 , respectively , for three different punctures @xmath426 , @xmath427 and @xmath428 , and such that @xmath429 bounds a disc in @xmath0 . the following lemma can be proven similar to the proof corollary in @xcite . [ lemma - hatcher ] let @xmath0 be a projective plane with at least five punctures . then given any two maximal simplices @xmath44 and @xmath430 of @xmath74 , there exists a sequence of maximal simplices @xmath431 such that @xmath432 is a codimension - one simplex for each @xmath218 . the lemmas below are analogous to lemma3.9 and lemma3.10 in @xcite . [ lemma - ideal ] let @xmath433 , @xmath434 be a good ideal triangle and let @xmath281 , @xmath282 , @xmath435 be the isotopy classes of @xmath76 , @xmath77 , @xmath351 , respectively . then @xmath436 and , hence @xmath437 is a @xmath75-simplex in @xmath74 . if @xmath438 is a realization of the latter simplex , then it is a good ideal triangle on @xmath0 . [ lemma - agree ] let @xmath439 and @xmath440 be two automorphisms of @xmath74 . if they agree on a maximal simplex , then they agree on all of @xmath74 . in section3 , we showed that the natural map @xmath441 is injective . we show that this natural homomorphism is onto . let @xmath442 and let @xmath44 be the isotopy class of a good ideal triangulation of @xmath0 . so , @xmath44 is a maximal simplex of @xmath74 . by lemma[lemma - ideal ] , @xmath443 , the automorphism induced by @xmath195 , takes a good ideal triangle to a good ideal triangle and @xmath439 can be realized by a homeomorphism . also , because each edge of a good ideal triangulation is an edge of two good ideal triangles , the homeomorphism of these triangles gives a homeomorphism @xmath444 of @xmath0 . by replacing @xmath444 by a diffeomorphism isotopic to @xmath444 if necessary , we may assume that @xmath445 is a diffeomorphism . if @xmath446 $ ] is the isotopy class of @xmath444 , then @xmath439 agrees with @xmath447 the automorphism induced by @xmath444 , on the maximal simplex @xmath44 of @xmath74 . from lemma[lemma - agree ] , they agree on all of @xmath74 . thus , @xmath448 . since the map @xmath449 is injective , we get @xmath450 . the proof of the theorem for punctured @xmath177 is now complete . throughout this section unless otherwise stated , @xmath0 will denote a connected nonorientable surface of genus @xmath1 with @xmath2 punctures where @xmath1 is odd and @xmath148 . in this section , we first show that automorphisms of @xmath7 preserve the topological type of the vertices of @xmath7 . we then prove that every automorphism of @xmath7 is induced by a diffeomorphism of @xmath0 . we prove this by induction on @xmath94 , where @xmath109 . let @xmath246 , @xmath248 , ... , @xmath453 be pairwise disjoint one - sided vertices such that each @xmath269 is disjoint from @xmath26 and @xmath28 . consider the link @xmath454 of these vertices . then @xmath195 restricts to an automorphism @xmath455 . @xmath454 is isomorphic to the complex of curves of a torus with @xmath456 punctures . observe that @xmath26 and @xmath28 vertices are contained in @xmath454 and by assumption @xmath451 . since @xmath457 , by theorem1 in @xcite . @xmath458 is induced by a diffeomorphism . in particular , @xmath452 . [ lemma - sequence ] let @xmath19 be a two - sided nonseparating circle . if @xmath459 and @xmath270 are two two - sided nonseparating circles , both intersecting @xmath19 transversely once , then there is a sequence @xmath460 of two - sided nonseparating circles such that each @xmath132 intersects the circle @xmath19 transversely once and @xmath132 is disjoint from @xmath461 for each @xmath462 . the complexes @xmath7 and @xmath465 have dimensions @xmath56 and @xmath466 , respectively . if @xmath467 , since these complexes of curves have different dimensions , @xmath7 and @xmath465 are not isomorphic . if @xmath468 , we proceed as follows . let @xmath186 . thus @xmath0 is a projective plane with two punctures and @xmath464 is a torus with one puncture . in this case , @xmath7 is finite ( @xcite ) , however , @xmath465 is infinite discrete since @xmath464 is a torus with one puncture . therefore , they are not isomorphic . now , assume that @xmath7 and @xmath465 are not isomorphic when @xmath0 has @xmath189 punctures . since there are @xmath189 punctures on @xmath0 , there are @xmath56 punctures on @xmath464 . we need to show that the complexes @xmath7 and @xmath465 are not isomorphic if @xmath0 has @xmath2 punctures . assume the contrary that there is an isomorphism @xmath469 . for any vertex @xmath191 in @xmath7 , @xmath195 induces an isomorphisms @xmath470 and @xmath471 . note that for a vertex @xmath191 of @xmath7 , the dual link @xmath192 of @xmath191 is connected if and only if @xmath191 is either one - sided or @xmath75-separating . for a vertex @xmath193 of @xmath465 , the dual link @xmath194 of @xmath193 is connected if and only if @xmath193 is either nonseparating or @xmath75-separating vertex . from this , it follows that the image of the union of the set of one - sided vertices and the set of @xmath75-separating vertices of @xmath7 is precisely the union of the set of nonseparating vertices and the set of @xmath75-separating vertices of @xmath465 . let @xmath191 be a @xmath75-separating vertex of @xmath7 . then @xmath472 is either a nonseparating vertex or a @xmath75-separating vertex of @xmath465 . first , we assume that @xmath472 is a nonseparating vertex @xmath193 of @xmath465 . clearly , @xmath197 is isomorphic to the complex of curves of a projective plane with @xmath189 punctures and @xmath198 is isomorphic to the complex of curves of a sphere with @xmath103 punctures . by lemma [ lemma - iso ] , these complexes are not isomorphic . therefore , @xmath472 can not be a nonseparating vertex @xmath193 of @xmath465 . now , we assume that @xmath472 is a @xmath75-separating vertex @xmath473 of @xmath465 . then @xmath474 is isomorphic to the complex of curves of a torus with @xmath56 punctures . by assumption , these complexes are not isomorphic . therefore , we get a contradiction . hence , @xmath7 and @xmath465 are not isomorphic . [ lemma - notisogenel ] let @xmath475 and @xmath476 . if @xmath0 is a connected nonorientable surface of genus @xmath477 with @xmath2 punctures and @xmath327 is a connected nonorientable surface of genus @xmath478 with @xmath21 punctures , then @xmath7 and @xmath479 are not isomorphic . the complexes @xmath7 and @xmath479 have dimensions @xmath60 and @xmath480 , respectively . if @xmath184 , since these complexes of curves have different dimensions , @xmath7 and @xmath479 are not isomorphic . if @xmath185 , we proceed as follows . let @xmath481 . we will prove that @xmath7 and @xmath479 are not isomorphic by induction on @xmath482 . let @xmath483 . in this case , @xmath484 and @xmath485 . in other words , @xmath0 is a connected closed nonorientable surface of genus @xmath34 and @xmath327 is a klein bottle with @xmath75 punctures . we need to show that the complexes @xmath7 and @xmath479 are not isomorphic . assume the contrary that there is an isomorphism @xmath486 . for any vertex @xmath191 in @xmath7 , @xmath195 induces an isomorphisms @xmath487 and @xmath488 . note that for a vertex @xmath191 of @xmath7 , the dual link @xmath192 of @xmath191 is connected if and only if @xmath191 is either one - sided vertex and let @xmath489 such that @xmath212 is an orientable surface or one - sided vertex such that @xmath212 is a nonorientable surface or nonseparating two - sided vertex . for a vertex @xmath193 of @xmath479 , the dual link @xmath194 of @xmath193 is connected if and only if @xmath193 is either one - sided or nonseparating two - sided or @xmath75-separating vertex . let @xmath191 be a nonseparating two - sided vertex of @xmath7 . then @xmath472 is either a one - sided vertex or a nonseparating two - sided vertex or a @xmath75-separating vertex of @xmath479 . first , we assume that @xmath472 is a one - sided vertex @xmath193 of @xmath479 . obviously , @xmath197 is isomorphic to the complex of curves of a projective plane with @xmath75 punctures and @xmath198 is isomorphic to the complex of curves of a projective plane with @xmath34 punctures . since these complexes have different dimensions zero and one , respectively , these complexes are not isomorphic . second , we suppose that @xmath472 is a nonseparating two - sided vertex @xmath473 of @xmath479 . then @xmath474 is isomorphic to the complex of curves of a sphere with @xmath187 punctures . by lemma [ lemma - iso ] , these complexes are not isomorphic . now , we assume that @xmath472 is a @xmath75-separating vertex @xmath490 of @xmath479 . clearly , @xmath491 is isomorphic to the complex of curves of a klein bottle with one puncture . since these complexes have different dimensions , these complexes are not isomorphic . therefore , we get a contradiction . hence , the complex of curves of a connected closed nonorientable surface of genus @xmath34 and the complex of curves of a klein bottle with @xmath75 punctures are not isomorphic . assume that the complexes @xmath7 and @xmath479 are not isomorphic for all integers @xmath164 to @xmath492 . now , let us prove it for @xmath482 . in this case , @xmath0 is a connected nonorientable surface of genus @xmath477 with @xmath2 punctures and @xmath327 is a connected nonorientable surface of genus @xmath478 with @xmath171 punctures . suppose the contrary that there is an isomorphism @xmath486 . note that for any vertex @xmath26 of @xmath7 , the dual link @xmath66 of @xmath26 is connected if and only if @xmath26 is either one - sided vertex and let @xmath30 such that @xmath20 is an orientable surface or one - sided vertex such that @xmath20 is a nonorientable surface or nonseparating two - sided vertex or @xmath75-separating vertex . for a vertex @xmath28 of @xmath479 , the dual link @xmath493 of @xmath28 is connected if and only if @xmath28 is either one - sided vertex or nonseparating two - sided vertex and @xmath494 such that @xmath495 is an orientable surface or nonseparating two - sided vertex such that @xmath495 is a nonorientable surface or @xmath75-separating vertex . let @xmath26 be a @xmath75-separating vertex of @xmath7 . first , we assume that @xmath337 is a one - sided vertex @xmath496 of @xmath479 . clearly , @xmath65 is isomorphic to the complex of curves of a connected nonorientable surface of genus @xmath477 with @xmath189 punctures and @xmath497 is isomorphic to the complex of curves of a connected nonorientable surface of genus @xmath498 with @xmath499 punctures . although these complexes have the same dimensions , these complexes are not isomorphic . because , by proposition [ prop - ndim1 ] , there is a maximal simplex of dimension @xmath500 in the complex of curves @xmath501 , whereas there is no any maximal simplex of dimension @xmath500 in the complex of curves @xmath502 . second , we suppose that @xmath337 is a nonseparating two - sided vertex @xmath503 of @xmath479 and let @xmath504 such that @xmath505 is an orientable surface . then @xmath506 is isomorphic to the complex of curves of a connected orientable surface of genus @xmath507 with @xmath508 punctures . by lemma [ prop - odim1 ] and proposition [ prop - ndim1 ] , these complexes @xmath509 and @xmath501 are not isomorphic . third , we assume that @xmath337 is a nonseparating two - sided vertex @xmath510 of @xmath479 and let @xmath511 such that @xmath512 is a nonorientable surface . then @xmath513 is isomorphic to the complex of curves of a connected nonorientable surface of genus @xmath514 with @xmath508 punctures . since these complexes @xmath515 and @xmath501 have different dimensions @xmath516 and @xmath517 , respectively , these complexes are not isomorphic now , we suppose that @xmath337 is a @xmath75-separating vertex @xmath518 of @xmath479 . then @xmath519 is isomorphic to the complex of curves of a connected nonorientable surface of genus @xmath478 with @xmath103 punctures . by assumption of the induction , these complexes @xmath520 and @xmath501 are not isomorphic therefore , we get a contradiction . hence , @xmath7 and @xmath479 are not isomorphic . [ lemma - take ] let @xmath0 be a connected nonorientable surface of genus @xmath521 @xmath522 with @xmath476 boundary components . the group @xmath199 preserves the topological type of the vertices of @xmath7 . let @xmath195 be an automorphism of @xmath7 . note that for a vertex @xmath191 of @xmath7 , the dual link @xmath192 of @xmath191 is connected if and only if @xmath191 is either one - sided vertex or nonseparating two - sided vertex or @xmath75-separating . therefore , @xmath195 can not take a one - sided vertex or a nonseparating two - sided vertex or a @xmath75-separating to a @xmath21-separating vertex with @xmath523 or separating vertex . let @xmath26 be a one - sided vertex of @xmath7 and let @xmath391 such that @xmath20 is an orientable surface . clearly , @xmath65 is isomorphic to the complex of curves of an orientable surface of genus @xmath94 with @xmath103 boundary components @xmath524 . let @xmath28 be a one - sided vertex and let @xmath525 such that @xmath170 is a nonorientable surface of genus @xmath478 with @xmath103 boundary components . obviously , @xmath203 is isomorphic to the complex of curves of a nonorientable surface of genus @xmath478 with @xmath103 boundary components @xmath526 . let @xmath191 be a nonseparating two - sided vertex of @xmath7 and @xmath197 is isomorphic to the complex of curves of a nonorientable surface of genus @xmath498 with @xmath171 boundary components @xmath527 . let @xmath193 be a @xmath75-separating vertex of @xmath7 and @xmath198 is isomorphic to the complex of curves of a nonorientable surface of genus @xmath477 with @xmath189 boundary components @xmath528 . since there are maximal simplices of different dimensions in the complexes of curves @xmath529 , @xmath530 and @xmath501 , these complexes are not isomorphic to @xmath531 . moreover , the dimensions of @xmath529 , @xmath530 and @xmath501 are @xmath517 , @xmath516 and @xmath517 , respectively . therefore , the complex of curves @xmath530 is not isomorphic to @xmath529 and @xmath501 . furthermore , although these complexes of curves @xmath529 and @xmath501 have the same dimensions , by lemma [ lemma - notisogenel ] , these complexes are not isomorphic . let @xmath490 be a separating vertex and let @xmath532 be circle . @xmath533 has two connected components . let @xmath534 such that @xmath228 and @xmath535 are nonorientable surfaces . more precisely , @xmath228 is a nonorientable surface of genus @xmath207 for some @xmath536 with @xmath382 punctures for some @xmath537 denoted by @xmath538 , then @xmath535 is a nonorientable surface of genus @xmath539 with @xmath540 punctures denoted by @xmath541 . the dual link @xmath542 has exactly two connected components . let us denote these components by @xmath543 and @xmath544 ; @xmath545 . we name @xmath546 so that the vertices of @xmath547 are the isotopy classes of circles on @xmath216 . let @xmath546 be the full subcomplex of @xmath7 with vertices @xmath547 . it follows that @xmath548 . then @xmath546 is isomorphic to @xmath225 . let @xmath473 be a separating vertex and @xmath549 be circle . similarly , we define @xmath550 and @xmath551 so that @xmath551 is isomorphic to @xmath233 . in other words , @xmath552 is isomorphic to the complex of curves of an orientable surface of genus @xmath207 for some @xmath553 with @xmath382 punctures for some @xmath554 denoted by @xmath555 , then @xmath556 is isomorphic to the complex of curves of a nonorientable surface of genus @xmath557 with @xmath540 denoted by @xmath558 . since @xmath552 is isomorphic to @xmath559 and all maximal simplices in @xmath559 have the same dimension , @xmath560 is not isomorphic to @xmath542 . hence , @xmath490 can not be mapped @xmath473 under @xmath195 . [ remark - take ] in the above proof , in case of @xmath484 and @xmath476 , one can see that the complex of curves of a torus with @xmath103 boundary components and the complex of curves of a projective plane with @xmath171 boundary components are not isomorphic by lemma [ lemma - notiso ] . [ thm - onto ] let @xmath0 be a connected nonorientable surface of odd genus @xmath109 with @xmath2 punctures . suppose that @xmath148 . then @xmath195 agrees with a map @xmath561 which is induced by a diffeomorphism @xmath562 . we assume that every automorphism @xmath563 is induced by a diffeomorphism of @xmath0 if @xmath0 is of odd genus @xmath564 . now , we show that @xmath195 is induced by a diffeomorphism of @xmath0 for genus @xmath109 . let @xmath459 be any two - sided nonseparating circle and @xmath191 denote its isotopy class . by lemma [ lemma - take ] , @xmath195 takes @xmath191 to a two - sided nonseparating vertex , say @xmath435 , and let @xmath565 . there is a diffeomorphism @xmath145 such that @xmath566 . then @xmath567 where @xmath568 the automorphism induced by @xmath145 . by replacing @xmath195 by @xmath569 , we can assume that @xmath570 . @xmath195 restricts to an automorphism @xmath571 . since @xmath197 is isomorphic to the complex of curves @xmath572 of a nonorientable surface @xmath212 of genus @xmath498 with @xmath171 boundary components , we get an automorphism @xmath573 of the complex of curves of a nonorientable surface of genus @xmath498 with @xmath171 boundary components . by induction , we can assume that @xmath573 is equal to a map @xmath574 which is induced by a diffeomorphism @xmath575 . by gluing two boundary components of @xmath212 obtained @xmath459 in a convenient way @xmath576 induces a diffeomorphism @xmath577 . it follows that @xmath578 . therefore , @xmath195 agrees with @xmath579 on every element of @xmath197 . the composition @xmath580 with @xmath195 , @xmath581 fixes @xmath191 and every element of @xmath197 . we may replace @xmath582 by @xmath195 . now , @xmath195 is an automorphism of @xmath7 such that it is identity on @xmath583 . let @xmath270 be a two - sided nonseparating circle dual to @xmath459 . in other words , @xmath270 intersects @xmath459 transversely once and there is no other intersection . let @xmath464 be a regular neighborhood of @xmath584 . the surface @xmath464 is a torus with one boundary component . let @xmath355 be the boundary component of @xmath464 . let @xmath193 and @xmath503 denote the isotopy class of @xmath270 and @xmath355 , respectively . obviously , since @xmath585 , @xmath586 . also , @xmath570 . since @xmath587 , @xmath588 . since @xmath589 by lemma [ lemma - intersect ] , there exists an integer @xmath2 such that @xmath590 . let @xmath591 be the set of isotopy classes of two - sided nonseparating circles which are dual to @xmath459 on @xmath0 . let @xmath592 be a two - sided nonseparating circle which is disjoint from @xmath270 and dual to @xmath459 . similarly , there exists an integer @xmath593 such that @xmath594 , where @xmath595 is the isotopy class of @xmath592 . since @xmath596 , @xmath597 . if @xmath598 , then @xmath599 since both @xmath270 and @xmath592 are dual to @xmath459 . thus , we have @xmath600 . this is a contradiction . therefore , we must have @xmath601 . now for any two - sided nonseparating circle @xmath482 which is dual to @xmath459 , by lemma [ lemma - sequence ] we can find a sequence of two - sided nonseparating circles dual to @xmath459 , connecting @xmath270 to @xmath482 such that each consecutive pair is disjoint . it follows that @xmath195 agrees with @xmath602 on every element of @xmath591 . therefore , @xmath195 agrees with @xmath602 on @xmath191 and on @xmath603 . let @xmath604 . in the following , we will denote @xmath605 simply by @xmath606 . if @xmath607 and @xmath608 are any other two - sided nonseparating circles dual to each other , then there exists a diffeomorphism @xmath609 mapping @xmath610 to @xmath611 , where @xmath72 and @xmath73 are the circles in figure[sur - circles ] . the set @xmath612 can be completed to a set @xmath130 of two - sided circles except @xmath19 as shown in figure[sur - circles ] . let @xmath269 be the isotopy class of @xmath132 . we orient tubular neighborhoods of elements of @xmath130 in such way that these orientation agree with an orientation of a regular neighborhood @xmath327 of @xmath613 . we note that @xmath327 is an orientable surface . the isotopy classes of dehn twists about elements of @xmath130 generate a subgroup @xmath128 . as shown above , the restriction of @xmath195 to @xmath614 agrees with the induced map @xmath615 of a diffeomorphism @xmath616 . any circle in @xmath130 is either disjoint from @xmath72 or is dual to @xmath72 , @xmath617 for all @xmath218 . similarly , @xmath618 for all @xmath218 . hence , @xmath619 . in particular , @xmath620 is isotopic to @xmath132 and @xmath621 where @xmath622 . let us denote @xmath623 by @xmath624 . for a circle @xmath137 dual to @xmath72 , we have the braid relation @xmath625 by conjugating with @xmath624 , we get @xmath626 but this relation holds if and only if @xmath627 and @xmath628 has the same sign . now if @xmath132 is dual to @xmath72 and if @xmath141 is dual to @xmath132 , by similar reasoning we get @xmath629 and @xmath627 have the same sign . it follows that all @xmath628 have the same sign , as we can pass from @xmath72 to any circle in @xmath130 through dual circles . suppose that @xmath630 , so that @xmath631 for all @xmath137 . thus @xmath624 reverses the orientation of tubular neighborhoods of @xmath132 . let @xmath632 be the reflection in @xmath633plane as shown in figure[sur - circles ] . the reflection @xmath632 leaves each @xmath137 invariant , but reverses the orientations of tubular neighborhoods of @xmath132 . here @xmath634 for all @xmath218 . thus @xmath635 in other words , @xmath636 . since @xmath637 is trivial by proposition [ prop - centralizer ] , we get that @xmath638 on the other hand , there exists a two - sided nonseparating circle @xmath19 as shown in figure[sur - circles ] such that the class @xmath26 of @xmath19 is contained in @xmath639 and @xmath640 . however , this is a contradiction since @xmath641 . let @xmath270 be any two - sided nonseparating circle and @xmath193 denote the isotopy class of @xmath270 on @xmath0 . since @xmath0 is a nonorientable surface of genus @xmath646 , @xmath72 and @xmath270 are dually equivalent by theorem3.1 of @xcite . in other words , there exists a sequence of two - sided nonseparating circles @xmath647 on @xmath0 such that @xmath648 , @xmath649 and the circles @xmath650 and @xmath651 are dual . using this sequence , we obtain that @xmath652 . indeed , for any two - sided nonseparating circle @xmath270 , @xmath653 agrees with @xmath195 . furthermore , since the isotopy classes of every separating circle and of any one - sided circle are in @xmath497 where @xmath496 is the isotopy class of some two - sided nonseparating circle @xmath482 , we obtain that @xmath653 agrees with @xmath195 on @xmath7 . [ [ proof - of - theoremtheorem - main1-for - nonorientable - surface - of - odd - genus ] ] proof of theorem[theorem - main1 ] for nonorientable surface of odd genus ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ it is shown that @xmath12 is injective in section3 . theorem[thm - onto ] implies that automorphisms of @xmath7 are induced by diffeomorphisms of @xmath0 . hence , this completes the proof theorem[theorem - main1 ] . harvey , w. j. , boundary structure of the modular group , in : i. kra and b. maskit , eds . , riemann surfaces and related topics , proc . 1978 stony brook conference , ann . of math . 97 , 245 - 251 , princeton university press , princeton , nj , 1981 .

the emerging field of optomechanics concerns with the study of the mechanical effects of light on mesoscopic and macroscopic mechanical oscillators . this phenomenon has been realized in the optomechanical systems consisting of an optical cavity with a movable end - mirror or with a membrane - in - the middle . the radiation pressure exerted by the light inside the optical cavity couples the moving mirror or the membrane which acts as a mechanical oscillator to the optical field . this optomechanical coupling has been employed for a wide range of applications such as the cavity cooling of microlevers and nanomechanical resonators to their quantum mechanical ground state @xcite , producing high precision detectors for measuring weak forces and small displacements and also providing a good approach for fundamental studies of the transition between the quantum and the classical world @xcite . on the other hand an alternative path to the studies of cavity optomechanics has been provided experimentally by systems consisting of ultracold atomic ensembles trapped in optical cavities @xcite where the excitation of a collective mode of the cold gas plays the role of the vibrational mode of the mirror . the standing electromagnetic wave inside the cavity forms a periodic potential , the so - called optical lattice , in which the cold atoms exhibit phenomena typical of solid state physics like the formation of energy bands and bloch oscillations @xcite . in such systems with high finesse cavities the atom - light interaction is enhanced because the atoms are collectively coupled to the same optical mode . besides , in the dispersive regime where the laser pump is far detuned from the atomic resonance the excited electronic state of the atoms can be adiabatically eliminated and consequently the only degrees of freedom of atoms will be their mechanical motions @xcite . for low photon numbers when the optical grating produced by the intracavity optical field is very shallow , one can approximately restrict the dynamics to the first two motional modes @xcite . in more recent theoretical investigations @xcite it has been shown that the simple two - mode model of a bose - einstein condensate ( bec ) can be improved by considering higher motional modes . in spite of similarities between the two kinds of optomechanical systems ( with a moving mirror and with a bec ) , there are some essential differences between them . firstly , in contrast to the moving mirror of the optomechanical sytems which is placed in the harmonic potential of its spring , the excitation modes of the bec are not based on the presence of such an external harmonic potential @xcite . secondly , their parameters are realized in different regimes ; the frequency of the excitation mode of the bec , i.e. , recoil frequency , is well below the dipole coupling strength while the oscillation frequency of a moving mirror is of the same order of magnitude as the coupling strength @xcite . on the other hand in addition to pumping the cavity from one of the end mirrors , in the cavities equipped with atomic gas it is also possible to pump the atoms from the side of the cavity@xcite . one of the most important characteristics of the many - body systems is the two - body collision which can affect the properties of the system . so in order to study the dynamics of a bec gas in a realistic experimental situation , it is necessary to take it into account . in the optomechanical systems containing the atomic gas there are two kinds of atom - atom interactions which are the origins of the system nonlinearities . firstly , due to the atom - photon interaction the potential acting on the condensate depends in a highly nonlocal and nonlinear way on the condensate itself which leads to the long range atomic interaction mediated by the cavity field @xcite . on the other hand there is an intrinsic nonlinearity due to the s - wave scattering which can take place at arbitrary momentum values and causes a broadening of the atomic momentum distribution @xcite due to the intraband transitions and also can scatter atoms to the other bands ( interband transitions)@xcite . furthermore , there is another kind of nonlinearitiy due to the finiteness of the particle number of the system which can be neglected in the thermodynamic limit where the total number of atoms is very large @xcite . in this work , we are going to extend the two - mode model considering the effects of atom - atom interaction for a one - dimensional bec in an optical cavity . we consider the system in the low photon regime but will not restrict our treatment to the weak atom - atom interaction . in fact the atomic collisions can populate several nonzero quasimomenta of the energy bands ( intraband transition ) and can also cause the atoms to be scattered to the other bands ( interband transitions ) . in order to have a simplified optomechanical model we take into account just the lowest nozero quasimomenta in the first brillouin zone . we show that in the bogoliubov approximation the atom - atom interaction not only shifts the energies of the excited modes but also play the role of an oprtical parametric amplifier ( opa ) for the bogoliubov side mode which can affect its normal - mode splitting behaviour @xcite . on the other hand due to the atomic collisions the resonance frequency of the cavity is shifted which leads to the decrease of the number of cavity photons and the depletion of the bogoliubov mode . these results are in good agreement with those obtained by the numerical solutions of the full description of gross - pitaevskii equation ( gpe ) @xcite . the paper is structured as follows . in sec . ii we will first give a thorough theoretical description of the many - body system under consideration and then derive a simplified optomechanical model . in sec . iii the quantum langevin equations ( qles ) are derived and linearized around the semiclassical steady state . in sec . iv we study the mean - field solutions and fluctuations of the system and in sec . v we derive the spectrum of the bogoliubov mode and investigate the normal mode splitting ( nms ) . finally , our conclusions are summarised in sec . and the decay rate is @xmath0.,width=3 ] we are going to study a gas of @xmath1 ultracold bosonic two - level atoms with mass @xmath2 and transition frequency @xmath3 inside the optical lattice of a single - mode , high - finesse fabry - perot cavity with length @xmath4 . the cavity is driven at rate @xmath5 through one of its mirrors by a laser with frequency @xmath6 , and wavenumber @xmath7 . we assume the bec is confined in a cylindrically symmetric trap with a transverse trapping frequency @xmath8 and negligible longitudinal confinement along the @xmath9 direction ( fig.[fig : fig1 ] ) . in this way we can describe the dynamics within an effective one - dimensional model by quantizing the atomic motional degree of freedom along the @xmath9 axis only . in the dispersive regime where the laser pump is far detuned from the atomic resonance ( @xmath10 exceeds the atomic linewidth @xmath11 by orders of magnitude ) , the excited electronic state of the atoms can be adiabatically eliminated and spontaneous emission can be neglected @xcite . in the frame rotating at the pump frequency , the many - body hamiltonian reads @xmath12 where @xmath13 is the annihilation operator for a cavity photon and @xmath14 is the cavity - pump detuning . @xmath15 is the single - particle hamiltonian of an atom inside the optical lattice of the cavity and @xmath16 is the atom - atom interaction that are respectively given by [ hamil1 ] @xmath17 here @xmath18 is the optical lattice barrier height per photon and represents the atomic backaction on the field , @xmath19 is the vacuum rabi frequency , @xmath20 and @xmath21 is the two - body s - wave scattering length @xcite . the second term in the hamiltonian of eq.([hamil1]a ) is a periodic potential of period @xmath22 . it is well - known that the eigenfunctions of a particle inside a periodic potential are the bloch functions @xmath23 with eigenvalues @xmath24 where @xmath25 is the bloch band index and @xmath26 $ ] is the quasimomentum of the particle @xcite . if there are @xmath27 periods inside the cavity , then @xmath28 . using the born - von karman periodic boundary condition @xmath29 and based on the bloch theorem , @xmath30 it is deduced that @xmath31 , where @xmath32 . in this way the quasimomentum @xmath33 in the first brillouin zone can be written in terms of optical wave number , i.e. , @xmath34 where @xmath35 . now using the bloch theorem , the eigenfunctions @xmath36 can be written as @xmath37 here @xmath38 is a periodic function with the optical lattice period @xmath39 and reads @xmath40 by substituting eq.([u ] ) into eq.([bloch2 ] ) , the eigenfunctions of the single - particle hamiltonian @xmath15 is obtained as @xmath41 the second quantized atomic wave field @xmath42 can be expanded as a fourier series in terms of the bloch functions , i.e. , @xmath43 here @xmath44 ( @xmath45 ) is the annihilation ( creation ) operator for the atomic field that annihilates ( creates ) a particle in a state determined with the bloch band index @xmath25 and quasimomentum @xmath33 . by using this expansion the hamitlonian of the system is diagonalized and one can investigate the bloch band structure of the condensate @xcite . however , the fourier coefficients @xmath46 can only be determined numerically and there is no closed form for the bloch functions . instead , by substituting eq.([blochf ] ) into eq.([psi ] ) one can obtain a fourier expansion of the atomic field in therms of plane waves @xmath47 in the derivation of eq.([psi2 ] ) we have done a bogoliubov transformation @xmath48 where @xmath49 ( @xmath50 ) is the atomic field operator that annihilates ( creates ) a particle in a state determined with the band index @xmath51 and quasimomentum @xmath33 . by substituting eq.([psi2 ] ) into eq.([h1 ] ) the hamiltonian of the systems can be written as @xmath52 in these sums @xmath53 , @xmath35 , and @xmath54 are the energy eigenvalues in which @xmath55 is the recoil energy . as seen , the field expansion of eq.([psi2 ] ) can no longer diagonalize the hamiltonian of the system . in fact that part of hamiltonian which corresponds to the atom - photon interaction remains nondiagonal . due to the interaction with the optical filed , the atoms are scattered to the nearest bands ( @xmath56 ) while their quasimomentum ( @xmath57 ) remains unchanged . since the optical potential is symmetric with respect to the origin , the hamiltonian of the system has the parity symmetry . so if the sytem starts from a homogeneous bec , only the cosine parts of the exponential functions in eq.([psi2 ] ) are excited because of the parity conservation . in the limit of weak photon - atom coupling , when either the photon number inside the cavity or @xmath58 is small , the lowest bands @xmath59 can be excited by fluctuations resulting from the atom - light interaction @xcite . on the other hand , the s - wave scattering populates fluctuations with arbitrary @xmath51 and @xmath57 @xcite . in a very simplified optomechanical model one can consider the scattering from @xmath60 to @xmath61 in the lowest band @xmath62 . in this way the atomic field operator ( eq.([psi2 ] ) ) can be truncated as @xmath63 where we have done the following bogoliubov transformations @xmath64 in the case that the system does not have parity symmetry , for example when the bec is inside a ring cavity , one should also consider sine modes with annihilation operators @xmath65 which in our model have been set aside @xcite . by substituting the atomic field operator , eq.([opaf ] ) , into the hamiltonian of eq.([h1 ] ) , one can arrive at the following hamiltonian @xmath66 where [ ac aa ] @xmath67 the first two terms in eq.([opa ] ) denote the cavity and the pump hamiltonians . the third term is the energy of the side - modes @xmath68 and @xmath69 which is just the third term in eq.([h2 ] ) with energy eigenvalues @xmath70 given by eq.([epsilon ] ) . the hamiltonian @xmath71 denotes the atom - photon interaction and @xmath16 is the atom - atom interaction . furthermore , @xmath72 is the s - wave scattering frequency and @xmath73 is the waist of the optical potential . this hamiltonian can be further simplified since for weak optical fields and large @xmath74 the depletion of the initial condensate remains weak . so we can use the bogoliubov approximation and treat the zero - momentum mode classically , i.e. , @xmath75 @xcite . in this way eqs.([opa ] ) reduce to the following forms [ hb ] @xmath76 here @xmath77 is the effective stark - shifted detuning . as seen from the hamiltonian of eq.([hb]b ) there are two kinds of optomechanical coupling . the first term of eq.([hb]b ) is the linear radiation pressure which couples the side mode @xmath68 to the optical field with the optomechanical coupling constant @xmath78 while the second term is the nonlinear optomechanical coupling of the two side modes @xmath68 and @xmath69 with the optical field . in the atom - atom interaction hamiltonian @xmath16 we have disregarded all terms proportional to @xmath79 . the influence of s - wave scattering has partly appeared as a shift in the side modes energies ( the third and fourth tems in eq.([subhb ] ) ) . since @xmath80 , where @xmath81 is the density of atoms , it is apparent that the energy shits obtained here is the same as that of bogoliubov theory @xcite . on the other hand the role of @xmath16 in eq.([hb ] ) for the bogoliubov side modes is very similar to that of an optical parametric amplifier ( opa ) in an optomechanical system with the nonlinear gain parameter @xmath82 @xcite . in sec . v we will show how this parameter can affect the nms behaviour of the coupled bogoliubov mode and the cavity field . the dynamics of the optomechanical system described by eqs.([hb ] ) is fully characterized by the following set of nonlinear qles , written in the frame rotating at the input laser frequency , [ nql ] @xmath83-\eta+\xi,\\ \dot{c}_{10}&=&-(i\omega_{10}+\gamma)c_{10}-i\frac{\sqrt{2n}}{4}u_{0}a^{\dagger}a-\frac{iu_{0}}{2}a^{\dagger}a c_{10}\nonumber\\ & & -\frac{i}{2}\omega_{sw}c^{\dagger}_{10}+f_{10},\\ \dot{c}_{01}&=&-(i\omega_{01}+\gamma)c_{01}-\frac{iu_{0}}{2}a^{\dagger}a c_{01}-\frac{i}{2}\omega_{sw}c^{\dagger}_{01}+f_{01},\nonumber\\\end{aligned}\ ] ] where @xmath84 and @xmath85 . here @xmath0 and @xmath11 characterize the dissipation of the cavity field and collective density excitations of the bec , respectively . the cavity - field quantum vacuum fluctuation @xmath86 satisfies the markovian correlation functions , i.e. , @xmath87 , @xmath88 with the average thermal photon number @xmath89 which is nearly zero at optical frequencies @xcite . besides , @xmath90 and @xmath91 are the thermal noise inputs for the two side modes of bec which also satisfy the same markovian correlation functions as those of the optical noise . the noise sources are assumed uncorrelated for the different modes of both the matter and light fields . now we are going to study the weak excitations of the condensate from its ground state . such excitations include small deviations of both the atomic wave function and the optical field from their respective stationary states . so we decompose each operator in eqs.([nql ] ) as the sum of its steady - state value and a small fluctuation . by substituting @xmath92 , @xmath93 and @xmath94 into eqs.([nql ] ) one can obtain a set of nonlinear algebraic equations for the steady - state values , [ ss ] @xmath95 where we have assumed @xmath96 and @xmath97 are real numbers @xcite and @xmath98 is the effective detuning , @xmath99 and @xmath100 . eq.([ss]c ) shows that the mean value of the mode @xmath69 is zero . it is the consequence of the fact that the ground state of the translationally invariant system is also invariant under discrete translation @xcite . on the other hand , the linearized qles for the fluctuating operators take the followinf forms [ flc ] @xmath101 where @xmath102 is the effective atom - photon coupling parameter . by defining the quadrature fluctuations [ flc ] @xmath103 with @xmath104 , the qles can be written in the compact matrix form @xmath105 where @xmath106^{t}$ ] is the vector of continuous variable fluctuation operators and @xmath107^{t}.\ ] ] is the corresponding vector of noises . the @xmath108 matrix @xmath109 is the drift matrix given by @xmath110 as is seen from the drift matrix , the side mode @xmath111 has been decoupled from the optical mode . it is due to the fact that in the qles , eqs.([nql]c ) , there is no standard radiation pressure coupling between the side mode @xmath69 and the optical field . instead , it is coupled to the radiation field via the nonlinear term , @xmath112 which is as small as @xmath113 of the radiation pressure term , and is deleted during the linearization process ( because the mean field value of this mode , @xmath114 , is zero ) . in this way the side mode @xmath69 acts as a medium level in the dynamics . in order to study the stationary properties of the system it is enough to focus our attention on the subspace spanned by the optical mode and the side mode @xmath68 . it means that we can consider only the upper block of the drift matrix of eq.([a ] ) . the system is stable only if the real part of all the eigenvalues of the matrix @xmath109 are negative . these stability conditions can be obtained by using the routh - hurwitz criterion @xcite . due to the linearized dynamics of the fluctuations and since all noises are gaussian the steady state is a zero - mean gaussian state which is fully characterized by the @xmath115 stationary correlation matrix ( cm ) @xmath116 , with components @xmath117 . when the system is stable such a cm is given by@xcite @xmath118 versus the normalized cavity detuning @xmath119 for two values of @xmath120 ( thin line ) and @xmath121 ( thick line ) . the parameters are @xmath122 , @xmath123 , @xmath124 , @xmath125 , @xmath126 and @xmath127.,title="fig:",width=288 ] versus the normalized cavity detuning @xmath119 for two values of @xmath120 ( thin line ) and @xmath121 ( thick line ) . the parameters are @xmath122 , @xmath123 , @xmath124 , @xmath125 , @xmath126 and @xmath127.,title="fig:",width=288 ] and ( b ) the normalized effective atom - photon coupling @xmath128 ( thick line ) and the mean photon number magnified by @xmath129 ( thin line ) versus the normalized s - wave scattering frequency @xmath130 . the parameters are the same as those of fig.[fig : fig2 ] . the cavity detuning has been set at @xmath131.,title="fig:",width=288 ] and ( b ) the normalized effective atom - photon coupling @xmath128 ( thick line ) and the mean photon number magnified by @xmath129 ( thin line ) versus the normalized s - wave scattering frequency @xmath130 . the parameters are the same as those of fig.[fig : fig2 ] . the cavity detuning has been set at @xmath131.,title="fig:",width=288 ] where @xmath132 and @xmath133 is the matrix of stationary noise correlation functions . for the noise diffusion matrix we have @xmath134 , where @xmath135 $ ] is the diffusion matrix , with @xmath136^{-1}$ ] as the mean number of thermal excitations of the bogoliubov side mode @xmath137 whose frequency of oscillations is given by @xmath138 therefore , eq.([vij ] ) becomes @xmath139 when the stability conditions are satisfied @xmath140 and eq . ( v2 ) will be equivalent to the following lyapunov equation for the steady - state cm @xmath141 equation([lyap ] ) is linear in @xmath116 and can be straightforwardly solved . however , the explicit form of @xmath142 is complicate and is not reported here . in this section we first discuss our results based on the numerical solutions of eqs.([ss ] ) for the mean fields and then solve eq.([lyap ] ) to obtain fluctuations in the number of atoms and photons and their entanglement . we will show how the atom - atom interaction affects the mean fields , cavity resonance and atom - photon entanglement . we analyse our results based on the experimentally feasible parameters given in ref.@xcite . here we are going to study the effect of s - wave scattering on the behaviour of the mean - field values of the optical field and the bogoliubov side mode @xmath68 , given by eqs.([ss ] ) , quantitatively . we consider @xmath143 atoms distributed in the optical cavity of length @xmath4=178 @xmath144 m with bare frequency @xmath145 corresponding to a wavelength of @xmath146=780 nm . the optical mode is coherently driven at rate @xmath5=80.06@xmath147 with the recoil frequency for rubidium atoms @xmath148 3.57 khz and the one - atom light shift @xmath58=0.96 @xmath147 @xcite . in fig.[fig : fig2 ] we have plotted the mean number of photons and the fraction of condensate atoms occupying the bogoliubov mode @xmath68 versus the normalized atomic detuning @xmath119 at a fixed temperature of @xmath149 k , and for two values of s - wave scattering frequencies @xmath82 . as is seen the increasing of the s - wave scattering interaction shifts the resonance frequency of the cavity to the lower values . besides , it causes the number of atoms in the bogoliubov mode to decrease i.e. , leads to the depletion of this mode . this is in good accordance with the results obtained from numerical solution of the gpe ( fig.1 of ref.@xcite ) . from eq.([ss ] a ) one can obtain the resonance condition as @xmath150 such a resonance shift has been obtained in ref.@xcite without considering atomic collisions . as was mentioned in that reference the bec acts as a kerr medium that shifts the empty - cavity resonance . however , due to the atomic collisions the bogoliubov side mode @xmath68 is depleted to other modes i.e. , the mean value @xmath97 decreases . so a bec with atom - atom interaction has a different resonance in comparison to a non - interacting one . for clarity , we have shown the effect of atom - atom interaction on the resonance of the cavity in fig.[fig : fig3](a ) . for a non interacting bec , i.e. , @xmath151 , the resonance occurs at @xmath152=28966 @xmath147 while for very large values of @xmath82 it goes to @xmath153 . in fig.[fig : fig3](b ) the effective atom - photon coupling parameter @xmath154 ( thick line ) , given by eq.([g ] ) , and the mean photon number of the cavity ( thin line ) have been plotted versus the normalized s - wave scattering frequency . this figure shows that both the effective atom - photon coupling parameter and the mean number of photons decrease with @xmath82 . all of these reductions are the direct consequence of the resonance shift exerted by atomic collisions . in the next subsection it will be shown that increasing the strength of atomic interaction causes the number of fluctuating photons to decrease . so the decrease of the total cavity photon number is a sign of atomic collisions . these results coincide with those obtained by numerical solution of gpe @xcite . ( b ) and the the entanglement between the bogoliubov mode @xmath68 and the optical field ( c ) versus normalized cavity detuning @xmath119 for two different values of @xmath120 ( thin line ) and @xmath121 ( thick line ) . all parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] ( b ) and the the entanglement between the bogoliubov mode @xmath68 and the optical field ( c ) versus normalized cavity detuning @xmath119 for two different values of @xmath120 ( thin line ) and @xmath121 ( thick line ) . all parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] ( b ) and the the entanglement between the bogoliubov mode @xmath68 and the optical field ( c ) versus normalized cavity detuning @xmath119 for two different values of @xmath120 ( thin line ) and @xmath121 ( thick line ) . all parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] after calculating the mean - field values we can obtain the elements of the drift matrix @xmath109 and solve for the steady solutions of eq.([na ] ) . as explained before , by solving the lyapunov equation ( eq.([lyap ] ) ) we can obtain the correlation matrix @xmath116 which gives us the second - order correlations of the fluctuations . the correlation matrix corresponding to the upper block of @xmath109 in eq.([a ] ) can be written as @xmath155 where @xmath156 and @xmath157 represent the correlations of the photonic and atomic degrees of freedom respectively , and @xmath158 describes the cross correlations . in this way we can calculate the incoherent excitation number of photons @xmath159 and the incoherent excitation number of atoms in the bogoliubov side mode @xmath160 on the other hand the bipartite entanglement between the atomic and photonic degrees of freedom can also be calculated by using the logarithmic negativity @xcite @xmath161,\ ] ] where @xmath162^{1/2}$ ] is the lowest symplectic eigenvalue of the partial transpose of the @xmath163 cm , @xmath116,with @xmath164 . and the s - wave scattering frequency has been considered to be @xmath120 . the other parameters are the same as those of fig.[fig : fig2 ] . for clarity , the values of @xmath165 and @xmath166 have been multiplied by 100 and 1000 respectively.,width=288 ] in figs.[fig : fig4 ] ( a ) and ( b ) we have plotted , respectively , the incoherent excitation number of photons and atoms versus the normalized cavity detuning @xmath119 for two values of @xmath82 . as is seen , fluctuations are small far from resonance but they increase rapidly near the resonance . on the other hand increasing atom - atom interaction decreases the fluctuations in the number of photons and atoms . since the total number of photons is the mean value @xmath167 that obtained in previous subsection plus the incoherent excitation number of photons , we can conclude that the stronger atom - atom interaction , the weaker optical field of the cavity . in fact the atom - atom interaction makes larger shifts in the cavity resonance frequency and consequently reduces the cavity field intensity . hence the decrease of the cavity output provides a direct measure of the atom - atom interaction within the condensate . in fig.[fig : fig4 ] ( c ) the entanglement between the bogoliubov mode @xmath68 and the optical field has been plotted versus the normalized cavity detuning @xmath119 for two values of @xmath82 . again , the maximum of entanglement occurs at resonance . besides , an increasing in the s - wave scattering frequency causes the entanglement of the atoms and photons to decrease . as it was mentioned above , an increase in atom - atom interaction shifts the resonance of the cavity which reduces the number of photons of the cavity and also causes the depletion of the bogoliubov mode.this reduction in the mean values of photons and atoms leads to a reduction in the effective atom - photon coupling ( fig.[fig : fig3](b ) ) which causes the entanglement of atoms and photons to decrease . the thermal effects on the incoherent excitation numbers of photons and atoms in the bogoliubov side mode as well as the entanglement between photons and atoms have been illustrated in fig.[fig : fig5 ] . as is seen from the figure , increasing the temperature causes the fluctuations of the number of photons and atoms to increase and on the other hand reduces the atom - photon entanglement . furthermore , the thermal effects for the range of temperatures below @xmath168 are negligible . finally , we are going to obtain the power spectrum of the bogoliubov side mode @xmath68 and also drive its effective frequency and damping rate . we will show that the coupling between the cavity field and the bogoliubov mode which behaves like a mechanical mirror , leads to the splitting of the normal mode into two modes ( nms ) . a similar theoretical approach has been done in ref.@xcite for opto - mechanical bose - hubbard hamiltonian ( ombh ) by expanding the atomic wave field in terms of the wannier functions which is valid only for weak atom - field nonlinearity @xcite . instead , here we use the atomic field expansion , eq.([opaf ] ) , in the momentum space . for this purpose , we solve the linearized qles for the fluctuations in the displacement operator @xmath169 as @xmath170 where @xmath171 is the fourier transformation of the fluctuations in the total force acting on the bogoliubov mode and @xmath172 is its susceptibility which are given by , respectively , @xmath173 where @xmath174 is the effective frequency of the bogoliubov mode , @xmath175 and @xmath176 is its effective damping rate , @xmath177 , ( b ) normalized effective bogoliubov mechanical damping @xmath178 and ( c ) normalized power spectrum of the displacement operator of the bogoliubov mode versus the normalized frequency @xmath179 for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) and for @xmath182 and @xmath183 . the other parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] , ( b ) normalized effective bogoliubov mechanical damping @xmath178 and ( c ) normalized power spectrum of the displacement operator of the bogoliubov mode versus the normalized frequency @xmath179 for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) and for @xmath182 and @xmath183 . the other parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] , ( b ) normalized effective bogoliubov mechanical damping @xmath178 and ( c ) normalized power spectrum of the displacement operator of the bogoliubov mode versus the normalized frequency @xmath179 for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) and for @xmath182 and @xmath183 . the other parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] , ( b ) normalized effective bogoliubov mechanical damping @xmath178 and ( c ) normalized power spectrum of the displacement operator of the bogoliubov mode versus the normalized frequency @xmath179 for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) and for @xmath182 and @xmath184 . the other parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] , ( b ) normalized effective bogoliubov mechanical damping @xmath178 and ( c ) normalized power spectrum of the displacement operator of the bogoliubov mode versus the normalized frequency @xmath179 for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) and for @xmath182 and @xmath184 . the other parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] , ( b ) normalized effective bogoliubov mechanical damping @xmath178 and ( c ) normalized power spectrum of the displacement operator of the bogoliubov mode versus the normalized frequency @xmath179 for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) and for @xmath182 and @xmath184 . the other parameters are the same as those of fig.[fig : fig2].,title="fig:",width=288 ] to calculate the spectrum , we need the correlation functions of the optical noise sources in the frequency domain @xmath185 where we have assumed @xmath186 . for the atomic noise sources in the frequency domain we have @xmath187 where we have omitted the indices of @xmath188 and @xmath189 for simplicity . here , @xmath190^{-1}$ ] is the number of thermal excitations for the bogoliubov mode and @xmath191 is the mechanical bogoliubov frequency of oscillation given by eq.([wm ] ) . by using the relation @xmath192 we can calculate the power spectrum of the displacement operator @xmath193 as follows @xmath194.\end{aligned}\ ] ] the modification of the frequency of the bogoliubov excitations of the condensate due to the radiation pressure given by eq.([omegaeff ] ) is equivalent to the optical spring effect in cavity optomechanical systems with movable mirrors @xcite . in figs . [ fig : fig6](a ) and [ fig : fig7](a ) we have plotted the normalized effective bogoliubov mechanical frequency @xmath195 versus the normalized frequency @xmath179 for two values of cavity damping rates @xmath183 ( fig.[fig : fig6 ] ( a ) ) and @xmath184 ( fig.[fig : fig7 ] ( a ) ) , and for two different values of @xmath180 ( thin line ) and @xmath181 ( thick line ) . as is seen from both figures , a higher two - body interaction makes the condensate more robust and the bogoliubov frequency of the condensate does not significantly deviate from @xmath191 . besides , a decrease in the damping rate of cavity causes the appearance of the peaks in the curves . in figs.[fig : fig6](b ) and [ fig : fig7](b ) the normalized effective bogoliubov mechanical damping @xmath178 has been plotted versus the normalized frequency @xmath179 for two values of cavity damping rates @xmath183 ( fig.[fig : fig6 ] ( b ) ) and @xmath184 ( fig.[fig : fig7 ] ( b ) ) . in the case of fig.[fig : fig6](b ) where the damping rate of the cavity is nearly equal to the effective atom - photon coupling constant @xmath196 , the effective damping rate @xmath176 is much lower compared to the case of fig.[fig : fig7](b ) where @xmath197 . it means that the stronger atom - photon coupling , the higher atom loss and hence the higher value of the effective damping . an experimental observation of this phenomenon of light induced back - action heating and consequent loss of atoms has been reported in ref.@xcite . it was also found that the atom loss rate is increased near the resonance . on the other hand , the effect of atomic collisions on the effective damping rate has been demonstrated in figs.[fig : fig6](b ) and [ fig : fig7](b ) which show that a stronger atom - atom interaction causes the effective damping rate to be increased near the resonance points . therefore increasing the rate of atomic collisions can help us in cooling of the bogoliubov mode of the bec by the radiation pressure . finally , we are going to investigate the influence of atomic collisions on the phenomenon of nms in the power spectrum of bogoliubov displacement operator . the nms is associated with the mixing between the fluctuation of the cavity field around the steady state and the fluctuations of the condensate ( bogoliubov mode ) around the mean field . the origin of the fluctuations of the cavity field is the beat of the pump photons with the photons scattered from the condensate atoms . the phenomenon of nms is observable whenever the energy exchange between the two interacting modes takes place on a time scale faster than the decoherence of each mode . in figs.[fig : fig6](c ) and [ fig : fig7](c ) we have shown the normalized power spectrum of the displacement operator of the bogoliubov mode @xmath198 versus the normalized frequency @xmath179 for two values of cavity damping rate @xmath183 ( fig.[fig : fig6 ] ( c ) ) and @xmath184 ( fig.[fig : fig7 ] ( c ) ) , and for two values of @xmath180 ( thin line ) and @xmath181 ( thick line ) . in the case of fig.[fig : fig6](c ) where the damping rate of the cavity is nearly equal to the effective coupling constant of atoms and photons @xmath196 , there is no splitting in the power spectrum ( neither for @xmath199 nor for @xmath200 ) while in fig.[fig : fig7](c ) where @xmath197 the nms is appeared for @xmath201 so in order to observe nms firstly the system should be in the strong coupling regime where the atom - photon coupling is larger than the decay rates of photons an atoms . when this condition is fulfilled , the nms can be observable with increasing atom - atom interaction . normal mode splitting of a system of large number of atoms coupled to a ring cavity has been observed experimentally in the strong cooperative coupling regime @xcite . the splitting of the normal mode has been observed by increasing the number of atoms which leads to the increase in the s - wave scattering frequency . in conclusion , we have done a theoretical investigation on the optomechanical properties of a one - dimensional bose - einstein condensate inside a driven optical cavity considering the effects of atomic collisions . due to the dispersive atom - photon interaction the atoms develop a band structure in the optical lattice of the cavity . in the limit of weak photon - atom coupling the lowest bands @xmath59 can be excited by fluctuations due to the atom - light interaction . on the other hand , the atomic collisions scatter the atoms to the higher band and also populate states with non zero quasimomentum . it has been shown that there is a nonlinear optomechanical coupling between the nonzero quasimomentum states with the optical field proportional to @xmath113 which can be disregarded in the bogoliubov approximation where the number of atoms in the bec mode is very large . therefore , there is no radiation pressure coupling between the atomic modes with nonzero quasimomentom and the cavity field in the bogoliubov approximation . in this way we have obtained a simplified optomechanical model where the bogoliubov side mode @xmath68 is coupled to the optical field through the radiation pressure term and the atom - atom interaction hamiltonian behaves like an opa for the bogoliubov mode with the non - linear gain parameter @xmath82 . it has been shown that in the strong coupling regime where the effective atom - photon coupling is greater than the cavity damping rate , an increase in the s - wave scattering frequency would lead to the nms in the power spectrum of the bogoliubov mode . furthermore , the atom - atom interaction causes the depletion of the bogoliubov mode and also shifts the cavity resonance frequency which leads to a decrease in the mean number of cavity photons . besides , it decreases fluctuations in the number of photons and atoms . hence the decrease in the cavity output provides a direct measure of the atom - atom interaction within the condensate . on the other hand , a stronger atomic collision rate causes a decrease in the effective atom - photon coupling which leads to a decrease in the entanglement between the bogoliubv mode and the optical field . the authors wish to thank the office of graduate studies of the university of isfahan for their support . c. genes , d. vitali , p. tombesi , s. gigan , and m. aspelmeyer , phys . a * 77 * , 033804 ( 2008 ) . j. d. teufel , j. w. harlow , c. a. regal , and k. w. lehnert , phys . * 101 * , 197203 ( 2008 ) ; j. d. teufel _ et al _ , nature ( london ) * 475 * , 359 ( 2011 ) ; j. d. teufel _ et al _ , nature ( london ) , * 471 * , 204 ( 2011 ) . s. de liberato , n. lambert , and f. nori , phys . rev . a * 83 * , 033809 ( 2011 ) . m. ludwig , c. neuenhahn , c. metzger , a. ortlieb , i. favero , k. karrai , and f. marquardt , phys . lett . * 101 * , 133903 ( 2008 ) . barzanjeh , m. h. naderi , and m. soltanolkotabi phys . a * 84 * , 023803 ( 2011 ) . c. bradaschia _ et al _ , nucl . methods phys . res . a * 289 * , 518 ( 1990 ) ; a. abramovici _ et al _ , science * 256 * , 325 ( 1992 ) ; p. fritschel , proc . spie * 4856 * , 282 ( 2003 ) . m. d. lahaye , o. buu , b. camarota , and k. c. schwab , science * 304 * , 74 ( 2004 ) . t. j. kippenberg and k. j. vahala , opt . express * 15 * , 17172 ( 2007 ) . f. brennecke , t. donner , s. ritter , t. bourdel , m. kohl , t. esslinger , nature ( london ) * 450*,268 ( 2008 ) . s. gupta , k.l . moore , k.w . murch , and d.m . stamper - kurn , phys . * 99 * , 213601 ( 2007 ) . f. brennecke , s. ritter , t. donner , and t. esslinger , science * 322 * , 235 ( 2008 ) . s. ritter , f. brennecke , k. baumann , t. donner , c. guerlin , and t. esslinger , appl . b * 95 * , 213 ( 2009 ) . o. morsch and m. oberthaler , rev . phys . * 78 * , 179 ( 2006 ) . c. maschler and h. ritsch , opt . commun . * 243 * , 145 ( 2004 ) p. domokos and h. ritsch , j. opt . b * 20 * , 1098 ( 2003 ) . c. maschler and h. ritsch , phys . rev . lett.*95 * 260401 ( 2005 ) . c. kittel , _ quantum theory of solids _ ( wiley , new york , 1963 ) . a. bhattacherjee , opt . commun . * 281 * , 3004 ( 2008 ) . g. szirmai , d. nagy , and p. domokos , phys . a * 81 * , 043639 ( 2010 ) . s. k. steinke and p. meystre , phys . a * 84 * , 023834 ( 2011 ) . w. chen , d. s. goldbaum , m. bhattacharya , and p. meystre , phys . a * 81 * , 053833 ( 2010 ) . g. konya , g. szirmai , and p. domokos , eur . phys . j. d * 65 * , 33 ( 2011 ) . r. kanamoto and p. meystre , phys . * 82 * 038111 ( 2010 ) . c. j. pethick and h. smith , _ bose einstein condensation in dilute gases _ ( cambridge university press , 2002 ) . d. nagy , p. domokos , a. vukics , and h. ritsch , eur . j. d * 55 * , 659 ( 2009 ) . c. maschler , mekhov , and h. ritsch , eur . j. d * 46 * , 545 ( 2008 ) . asboth , p. domokos , and h. ritsch , phys . a * 70 * , 013414 ( 2004 ) . d. s. goldbaum , k. zhang , and p. meystre , arxiv:0911.3234v1 . i. s. gradshteyn and i. m. ryzhik , _ table of integrals , series and products _ ( academic press , orlando , 1980 ) ; a. hurwitz , _ selected papers on mathematical trends in control theory _ , edited by r. bellman and r. kabala ( dover , new york , 1964 ) . c. w. gardiner and p. zoller , _ quantum noise _ ( springer , berlin , 2000 ) . c. genes , d. vitali , and p. tombesi , new j. phys . * 10 * , 095009(2008 ) . j. m. zhang , f. c. cui , d. l. zhou , and w. m. liu , phys . rev . a * 79 * , 033401 ( 2009 ) . p. horak , s. m. barnett , and h. ritsch , phys . a * 61 * , 033609 ( 2000 ) . j. eisert , ph.d . thesis , university of potsdam , 2001 ; g. vidal and r. f. werner , phys . a * 65 * , 032314 ( 2002 ) ; g. adesso , a. serafini , and f. illuminati , phys . a * 70 * , 022318 ( 2004 ) . a. bhattacherjee , j. phys . b : at . mol . . phys . * 43 * 205301 ( 2010 ) . j. larson _ et al _ , phys . lett . * 100 * 050401 ( 2008 ) . k. w. murch , k. l. moore , s. gupta , and d. m. stamper - kurn , nature phys . * 4 * , 561 ( 2008 ) . j. klinner , m . lindholdt , b . nagorny , and a. hemmerich , phys . * 96 * , 023002 ( 2006 ) . c. w. gardiner , phys . a * 56 * , 1414 ( 1997 ) .

until a few years ago there was a common feeling in the condensed matter physics community that we had a very good understanding of the universal properties of one - dimensional systems of fermions through our knowledge of luttinger liquid theory and its instabilities @xcite . one of the key results of luttinger liquid theory @xcite is that interacting electrons in one - dimension have separate dynamics in the spin and charge sectors at low energies . these low energy collective spin and charge modes are bosonic and generally propagate with different velocities , with the spin mode typically being the slower of the two . luttinger liquid theory provides a particularly convenient theoretical framework for discussing the response of the electron gas to a time - dependent local potential , even in higher dimensions : the mathematical trick of bosonization " makes the evaluation of correlation functions relatively simple , and a potential local on the scale of the fermi wavelength results in scattering of primarily @xmath3-wave type which effectively converts the problem to just one dimension ( the radial coordinate ) . a few years back it was realized that at finite temperatures a luttinger liquid system " can take on qualitatively new behaviors in transport @xcite and the single - particle green s function @xcite . some of the results for the single - particle green s function were anticipated in an earlier work @xcite , but the many novel features that can arise due to finite temperatures were not fully appreciated until the more recent works mentioned above which really started proper the study of the spin - incoherent luttinger liquid @xcite . the spin - incoherent regime of a one - dimensional interacting electron gas is defined in the following way . let us assume that at zero temperature our system is spin - charge separated with spin velocity @xmath4 and charge velocity @xmath5 such that @xmath6 . if the interactions are strong ( regardless of the range of interactions ) it can be shown that @xmath4 can be _ exponentially small _ compared to @xmath5 @xcite . defining a characteristic spin energy @xmath7 and a characteristic charge energy @xmath8 for mean particle spacing @xmath9 , it is evidently clear that the spin energy can be exponentially suppressed relative to the charge energy : @xmath10 . an exponentially separated spin and charge energy makes it possible for the thermal energy to lie in the range @xmath11 where @xmath12 is boltzman s constant and @xmath13 is the temperature . a one dimensional electron system in the energy window @xmath11 is said to be in the spin - incoherent regime . if the charge sector is gapless it will be described by a zero - temperature luttinger liquid theory and the resulting system is called a spin - incoherent luttinger liquid ( sill ) . note that if @xmath14 refers to a spin _ gap _ then a sill is still realized when @xmath11 provided the charge sector remains gapless . for a general discussion of the sill see @xcite . for sill physics in ferromagnetic and superconducting hybrid structures see @xcite , and in cold atomic gases see @xcite . in this article we will only discuss a particularly simple form of a time - dependent spatially local potential whereby its value is changed instantaneously from zero to non - zero . the prototypical physical realization of this situation is when a photon in the x - ray energy range ejects a deep core " electron from a metal instantly changing the local charge by @xmath15 . the electrons of the metal then rush in to screen the positive potential and thereby change their states . if the electrons interact strongly with each other the processes of the electrons rushing in " to screen the local potential will obviously be impacted . according to luttinger liquid theory , interactions qualitatively change the nature of the low energy physics in such a way that electron - like quasi - particles are absent from the theory @xcite leaving only the collective bosonic excitations . on this basis alone , we expect the response of one - dimensional systems to differ in some ways from their higher dimensional counter parts . in fact , this expectation is correct but to fully appreciate the differences , it is important to review some of basic ideas and qualitative expectations for the singular responses of non - interacting electron systems . the seminal theoretical ideas of the singular response of a non - interacting electron gas due to a localized core hole were worked out by @xcite and @xcite . a nice textbook - level discussion can be found in @xcite and a more recent review in @xcite . the most important results to emerge are : ( i ) there is a threshold frequency @xmath1 above which there is photon absorption and below which there is not . ( ii ) the photon absorption at frequencies just above @xmath1 follows a power law with exponent @xmath2 . the photon absorption intensity is computed from a fermi s golden rule - type treatment in the electron - photon coupling and from the results just cited above takes the form @xmath16 near threshold where @xmath17 is a constant , and @xmath18 for @xmath19 while @xmath20 for @xmath21 . a schematic of the possible behaviors is shown in fig . [ fig : schematic_threshold ] . schematic behavior of photon absorption intensity , eq.([eq : i_fl ] ) , for an infinite mass core hole in three dimensions . the exponent @xmath2 given by eq.([eq : a ] ) contains positive contributions from excitonic effects and negative contributions from anderson s orthogonality catastrophe . it is possible for @xmath2 to be either positive ( a ) or negative ( b ) . in the spin - incoherent regime of one dimensional systems the absorption is not generally given by eq.([eq : i_fl ] ) but instead exhibits logarithmic corrections and additional physics enters the exponent @xmath2 . ] the important microscopic details of the electron dynamics and the electron - core hole interaction are encoded in the exponent @xmath2 and the power - law form itself . the power - law arises from summing an infinite series of diagrams containing logarithmic divergences @xcite . as we will see , in the sill the power law form is not sacred . it will be modified by logarithmic corrections , and the physics encoded in @xmath2 will depend on whether or not the one - dimensional system is in the spin - incoherent regime . for a non - interacting electron gas the exponent takes the form @xcite @xmath22 ^ 2,\ ] ] where @xmath23 is the phase shift induced by the core hole in the @xmath24 channel of the conduction electrons . often @xmath3-wave scattering ( @xmath25 ) is the dominant channel and we will assume this is the case throughout this review . as the total amount of screening charge around the core hole must be one unit of charge , there is a friedel sum rule : @xmath26 if the scattering is not too strong , the phase shifts @xmath23 can be computed in the born approximation and one finds @xcite @xmath27 where @xmath28 is the density of states at the fermi energy and @xmath29 is the strength of the electron - hole interaction assumed momentum and energy independent . ( the actual core hole - electron interaction is @xmath30 and therefore attractive . ) within the born approximation and assuming only @xmath3-wave scattering , the exponent becomes @xmath31 ^ 2.\ ] ] evidently , eq.([eq : a ] ) and eq.([eq : a_approx ] ) imply there are competing effects in @xmath2 . the first term proportional to @xmath32 is positive while the second term proportional to @xmath33 ^ 2 $ ] is negative . going back to eq.([eq : i_fl ] ) one sees that a positive contribution to @xmath2 tends to make the threshold response more singular , while a negative contribution to @xmath2 tends to make the threshold response less singular . in fact , the origin of these two contributions are very well understood . the first contribution to @xmath2 in eq.([eq : a ] ) comes from excitonic effects arising from the attractive interaction between the electrons and the core hole . the second contribution to @xmath2 in eq.([eq : a ] ) arises from anderson s orthogonality catastrophe @xcite and it tends to suppress the absorption by the small ( vanishing as @xmath34 ) overlap between initial and final states . note that for a non - interacting fermi sea both excitonic effects and orthogonality effects disappear if @xmath35 and the absorption threshold becomes a simple step . looking ahead for a moment to luttinger liquid physics , it is instructive to point out that if a collection of fermions is interacting , there will be orthogonality effects if an additional fermion of like charge is added to the system because the other fermions must adjust their state to accommodate the newcomer , but no excitonic effects . this rearrangement of electronic states is what is responsible for the famous power law suppression of the tunneling density of states in a luttinger liquid @xcite . finally , it is worth emphasizing that depending on the microscopic details involved , it is possible for @xmath2 to be positive or negative as shown in fig . [ fig : schematic_threshold ] . if the mass of the core hole is finite and one considers only direct transitions like those in fig . [ fig : direct_transition ] one finds that the threshold is smeared out and disappears @xcite . we will find this is not the case when we go to one dimension . schematic of the possible direct transitions for exciting a hole with ( a ) finite mass and ( b ) infinite mass . here @xmath36 denotes the fermi energy . in ( a ) an electron is excited from a valence band up to a conduction band leaving behind a single mobile hole in the valence band . in ( b ) an electron is excited from an infinite mass core level . the photo - absorption of both process ( a ) and ( b ) differ in one dimension compared to their counterparts in three dimensions . furthermore , the response in one dimension depends on whether the system is in the luttinger liquid or spin - incoherent luttinger liquid regime . ] before we turn our attention to the spin - incoherent luttinger liquid , it is useful for orientation to discuss the singular response of its closest cousin , the luttinger liquid . the behavior exhibited by the one - dimensional luttinger liquid ( ll ) will already take us part of the way to understanding the response of the sill . in all of the results quoted below finite system size effects will not become important until frequencies smaller than @xmath37 are probed , where @xmath38 is the length of the system and @xmath39 is the fermi velocity . also , finite temperature will cut - off the singular behaviors when relevant frequencies are of order @xmath40 and smaller . in this sense , finite system size and temperature can act as a non - zero threshold for tunneling into a finite length system or an infinite system at finite temperature , for example . let us first consider the simplest kind of singular response : abruptly adding a particle to a ll as occurs during tunneling . a nice pedagogical discussion of the essential physics of tunneling into a ll is given by @xcite . here we adopt the central elements of their discussion . the tunneling process is conveniently visualized within a wigner crystal - type picture for spinless electrons , as shown in fig . [ fig : wigner_schematic ] . schematic of the tunneling process of spinless electrons into a wigner crystal " type electron arrangement . electrons are schematically represented by black dots evenly spaced on either side of the tunneling barrier of strength @xmath41 at @xmath42 . the tunneling occurs in two steps : a fast process where the electron enters the region @xmath21 from the left followed by a slow process of charge relaxation where each electron shifts over to the right by a distance of one lattice constant . at long times @xmath43 as discussed in the text . the resulting behavior of the tunneling density of states is @xmath44 the qualitative behavior of the tunneling density of states looks very much like that in fig . [ fig : schematic_threshold](b ) with @xmath45 . ] such a picture implies very strong electron interactions , but the physical result and the final mathematical expression that describes it are actually valid at any interaction strength . we imagine a process where an electron tunnels from the left of the barrier at @xmath42 to the right of the barrier at time @xmath46 . the rate at which the electron tunnels is determined by the debye frequency of the wigner crystal , @xmath47 , where @xmath48 is the fermi wave vector and @xmath49 is the phonon ( charge ) velocity . when an electron tunnels the potential near the barrier suddenly ( on a time scale @xmath50 ) changes . the energy cost of this new configuration is large , @xmath51 , where @xmath52 is the interaction parameter for the spinless luttinger liquid . ( here @xmath53 if the electrons are non - interacting . ) this large energy cost is reduced by the relatively slow process of the electrons relaxing to new equilibrium positions in order to minimize the energy . once the relaxation process is complete , all the electrons will be shifted to the right by one crystalline period . in the thermodynamic limit this new wavefunction is orthogonal to the original one which implies at zero energy ( @xmath54 ) there is no tunneling . this is a concrete example of anderson s orthogonality catastrophe that we earlier mentioned plays a role in the exponent @xmath2 of the threshold behavior of photon absorption . for finite energy the final state may be in an excited state so the relaxation is not complete and the initial and final wavefunction can have finite overlap . we will now see how this directly leads to a power law tunneling density of states , one of the singular " responses of interest in this article . we can compute the tunneling density of states from the single particle green s function @xcite . in the semi - classical approximation the electron tunneling takes the path of minimal action which is given by the classical trajectory @xmath55,\ ] ] where @xmath56 is the electron annihilation operator at position @xmath57 and time @xmath58 , and @xmath59 is the creation operator . the initial density deformation @xmath60 will eventually relax to @xmath61 . as the processes is classically forbidden , the trajectory runs in imaginary time @xmath62 . the classical equation of motions for a harmonic lattice give @xmath63 the potential energy of the deformation is @xmath64 ^ 2},\ ] ] which gives @xmath65 . by the virial theorem , the kinetic energy is equal to the potential energy . therefore the action describing the slow part of the tunneling is @xmath66 whence @xmath67 . taking the fourier transform one finds the tunneling density of states @xmath68 . since we have assumed @xmath69 , the tunneling density of states @xmath70 as @xmath34 , as we noted above in the context of the orthogonality catastrophe . the discussion above illustrates some of the most central aspects of singular responses of interacting one - dimensional systems . however , there is much more to the story . let us now suppose that instead of an electron tunneling , we consider an incident photon that abruptly creates a particle - hole pair as shown in fig . [ fig : direct_transition ] . note that due to the presence of the gap between the hole states and conduction band , there will be a minimum frequency @xmath1 required to create the particle - hole pair . if we first focus on the electrons in the conduction band , we see that the process is similar to tunneling : conduction band electrons see the sudden appearance of a new electron and they must adjust their positions by relaxing to a new state just as if the electron tunneled from outside the system . as we have seen , this relaxation rate is related to the strength of the electron interactions @xmath71 in the luttinger liquid , and based on the power law density of states we obtained , we expect the threshold behavior will also contain at least a power law contribution with an exponent that contains information about the interactions in the system . it is also important to consider the interaction between the hole created and the electrons in the conduction band . as holes and electrons generally have different masses and they attract rather than repel each other , the creation of a hole adds new physics to the electron tunneling physics above . in fact , to compute the photon absorption spectrum near threshold we must _ simultaneously _ account for the interaction of the electrons with each other and with the hole . in general , this is a very difficult problem . fortunately , though , the power of luttinger liquid theory and the commonly used technical methods such as bosonization allow us to make some precise and quite general statements about @xmath72 . pioneering theoretical work in the study of the fermi - edge singularity in luttinger liquids was carried out more than 15 years ago @xcite , and later works soon followed @xcite . a recent review of the optical response of low dimensional systems ( including some aspects of one - dimension ) exists @xcite , so here we will only highlight the main issues and results to set the stage for our discussion of the spin - incoherent one - dimensional systems . in one - dimensional electronic systems the effect of impurities is especially dramatic compared to three dimensions . even for an unrealistic non - interacting one - dimensional system the electrons localize leading to insulating behavior . however , for a realistic system which contains repulsive interactions even a single local potential of arbitrary strength will localize the electrons to the half - line @xcite . the central physics of this result is the relevance of electron backscattering as a perturbation : at low energies the strength of the impurity grows to arbitrarily large values and eventually divides the system into two semi - infinite segments . this physics is important for the fermi - edge singularity involving an infinitely massive core hole , _ i.e. _ one that is localized in space . on the other hand , when the mass of the core hole is finite the physics is very different . the properties of a heavy particle ( holes are typically heavier than electrons ) moving in a luttinger liquid have been investigated @xcite . it was shown that the mobility of the heavy particle diverges at low energies , implying that electron backscattering from it becomes irrelevant . we therefore see that the fermi - edge singularity in a luttinger liquid falls into two categories : ( i ) infinite hole mass and ( ii ) finite hole mass . note that in three dimensions it was also important to divide the physics into the same two categories , but the reason for that division was different . in three dimensions finite hole mass led to excitonic effects that smeared the edge singularity of the direct transition and eliminated the threshold . this has to do with a large ( compared to one - dimension ) phase space for recoil effects . a sharp threshold only exists for the infinitely massive hole . on the other hand , in one dimension the issue is about the relevance / irrelevance of backscattering and that is decided by the interactions ( repulsive or attractive ) in the ll . another way to see that it is natural to divide the classes into infinite and finite masses cases is that the symmetry is different . an infinitely massive hole breaks the translational symmetry of the problem , while a hole with finite mass does not . one of the first issues addressed regarding the fermi - edge singularity in the ll is whether the threshold still exists for the case of finite hole mass . this was answered in the affirmative @xcite . the next logical question is how does the threshold exponent @xmath2 depend on the electron - hole and electron - electron interactions ? initially it was argued that @xmath2 was independent of the hole mass @xcite , but this was later challenged and corrected @xcite . knowing that any finite mass hole is an irrelevant perturbation @xcite we can immediately see that only forward scattering interactions will contribute to @xmath2 . indeed this is found as well as a schematic phase diagram for when @xmath2 is positive or negative @xcite . the result is @xcite @xmath73,\ ] ] where @xmath74 is an asymmetric phase shift and @xmath75 a symmetric phase shift defined below in the discussion of the spin - incoherent case . the presence of @xmath74 implies that the exponent depends on the mass of the hole . note that both @xmath74 and @xmath75 correspond to forward scattering processes , and @xmath69 is the luttinger parameter for repulsively interacting spinless electrons . we can directly compare the result eq.([eq : a_ll_mobile ] ) with the non - interacting case eq.([eq : a ] ) by taking @xmath53 and @xmath76 ( appropriate for an infinitely massive hole ) . doing so , we note the form is identical to an s - wave only scattering model . from this comparison we also see that @xmath75 has the meaning of a phase shift up to factors of 2 and @xmath77 . note that we are allowed to compare the result for a mobile impurity in a ll with an infinitely massive impurity in a free electron gas in three dimensions because in the latter backscattering is not a relevant perturbation . they are therefore in a similar class " . the situation is different when we turn to immobile infinite mass holes . here , as before , a threshold exists , but now @xmath2 is modified by the relevant backscattering and in fact obtains a universal contribution of 1/8 @xcite , @xmath78 in the next section we will discuss how these two results are modified in the spin - incoherent regime . but before we do that , it is important to comment on the experimental situation . unfortunately , there are few published results @xcite and the overall level of agreement between theory and experiment is poor . clearly there is great need for further experiments on fermi - edge singularities on clean , single mode one - dimensional systems . we hope this article will help to stimulate more . we finally come to the main topic of this review well prepared from the earlier discussions of singular responses of non - interacting electron gases and luttinger liquids . the question to ask is what are the similarities and differences between the singular responses of a sill and a ll or fermi liquid ? as we will see , the behavior is qualitatively different from both the ll and fermi liquid making it possible to distinguish them in experiment . perhaps the simplest and most striking difference is the behavior of the tunneling density of states . we begin there . the first theoretical study of tunneling into a sill was reported in the infinite @xmath79 limit of the hubbard model in a lengthy bethe ansatz calculation @xcite . immediately afterwards a simpler and more general method based on bosonization and the imaginary time path integral representation was given @xcite . extensions of the method to finite length systems including boundary effects and externally applied magnetic fields have also been reported @xcite . while the imaginary time path integral formulation is well suited to the regime of the sill , it does not provide a conceptually straightforward generalization to the regime @xmath80 . in order to address that regime , we need a representation of the spin degrees of freedom that are valid at arbitrary energy scales . such a formulation is now in hand @xcite and it provides a good starting point for numerical studies in the regime @xmath80 . as is typical of interacting one dimensional electron systems , we assume the hamiltonian in the strongly interacting case is spin - charge separated @xcite , @xmath81 . here , @xmath82 is identical to that of the ll , and is given by @xcite @xmath83 , \label{eq : h_holon}\ ] ] where @xmath84 and @xmath85 are bosonic fields satisfying @xmath86=-i\frac{\pi}{2 } { \rm sgn}(x - x')$ ] , and @xmath87 is the effective luttinger interaction parameter of the electron gas @xcite . on the other hand , the spin hamiltonian at arbitrary temperatures is to a very good approximation given by a nearest neighbor antiferromagnetic heisenberg spin chain @xcite , @xmath88 where evidently the spin energy is set by @xmath89 : @xmath90 . the basic idea is to represent the electron operator as a product of operators that describe the holons @xmath91 [ spinless fermions that naturally arise in the context of strongly interacting fermions and the spin - incoherent regime @xcite ] and the spin degrees of freedom @xmath92 . the holon operators ( denoted by @xmath93 ) by construction satisfy the equation @xmath94 where @xmath95 is the electron annihilation operator for electrons of spin projection @xmath96 , and @xmath97 the corresponding electron creation operator . the issue of how to bosonize the electron operator for a strongly interacting system earlier arose in the context of the large @xmath79 limit of the one - dimensional hubbard model where @xcite wrote the electron creation operator as @xmath98 where @xmath99 creates a site on the spin chain eq.([eq : s_chain ] ) with spin projection @xmath96 . the expression in eq.([eq : penc_bos ] ) can be physically motivated as follows . from eq.([eq : psi ] ) it is clear that the creation of an electron is also accompanied by the creation of a holon . however , electrons also carry spin so there must be a component of the electron operator that also creates spin . this is accomplished by @xmath99 . in general , one has @xmath100 as the object that adds a new site to the spin chain between @xmath101 and @xmath102 . while this appears physically intuitive , the expression suffers from the drawback that it does not naturally account for the variation of electron density with position in a real electron gas @xcite . the remedy for this issue @xcite is to define the position at which the spin site is added to the chain eq.([eq : s_chain ] ) in terms of the number of holons to the left of the site , @xmath103 in terms of eq.([eq : l_x ] ) the electron creation and annihilation operators are defined as @xmath104 the operators given above explicitly account for the fact that the spins are attached to electrons , and the formulas are valid at all energy scales . it is perhaps worth noting that even though the hamiltonian is spin - charge separated , the electron operators are not written as a product of a spin piece and a charge piece because the spin " pieces @xmath105 also depend on the electron density via eq.([eq : l_x ] ) . in writing eq.([eq : psi+_x ] ) and eq.([eq : psi_x ] ) no assumptions have been made about the energy scale relative to the spin and charge energies . we now restrict our considerations to energies small compared to @xmath106 , but arbitrary with respect to @xmath14 . in this case , we are free to bosonize the holon sector . we have already given the form of the charge hamiltonian in eq.([eq : h_holon ] ) . the spinless fields @xmath84 and @xmath85 can be related to the holon density as @xcite @xmath107,\ ] ] where the holon fermi wave vector is twice the electron fermi wavevector @xcite @xmath108 and the bosonic fields satisfy the commutation relations @xmath109=i\pi\delta(x - y)$ ] . since we are interested in low energies with respect to the charge energy , the electron operator may be expanded about the two holon fermi points at @xmath110 , @xmath111 where @xmath112 destroys an holon near the right fermi point and @xmath113 destroys an electron near the left fermi point . the left and right holon operaters are bosonized as @xmath114},\ ] ] where @xmath115 is a short distance cut - off of order the interparticle spacing @xmath9 . combining the results of eq.([eq : l_x ] ) , eq.([eq : psi_x ] ) , eq.([eq : den_rel ] ) , eq.([eq : psi_rl ] ) and eq.([eq : psi_rl_fields ] ) one obtains the bosonized form of the electron annihilation operator for spin @xmath96 @xmath116}+e^{-i[k_f^hx+\theta(x)]}\right ) z_{l,\sigma}\bigg |_{l=\frac{1}{\pi}[k_f^hx+\theta(x)]},\end{aligned}\ ] ] and an analogous expression for the electron creation operator @xmath117 . expression eq.([eq : psi_bos ] ) , however , it not quite complete as it does not account for the discreteness of the charge of the electron . this can be accomplished by interpreting @xmath118 } \to \sum_l z_{l,\sigma}\delta\left({1 \over \pi}[k_f^hx+\theta(x)]-l\right),\ ] ] after which the full electron annihilation operator ( include both left and right moving parts ) becomes @xcite @xmath119},\ ] ] where @xmath120 the expression for the electron annihilation operator eq.([eq : psi_final_bos ] ) is the key result of @xcite who also showed that in the limit of small energies compared to @xmath14 the expression correctly recovers the standard ll formulas for the electron annihilation operator . with eq.([eq : psi_final_bos ] ) correlation functions involving electron operators can be expressed in terms of the correlation functions of the holon and spin sectors at arbitrary temperatures with respect to @xmath14 , but small energies compared to @xmath106 . the tunneling density of states in a sill is qualitatively different from the ll . while the tunneling density of states in a ll qualitatively resembles the behavior shown in fig.[fig : schematic_threshold](b ) with @xmath45 , the tunneling density of a sill qualitatively resembles the behavior in fig.[fig : schematic_threshold](a ) with @xmath45 , at least for a range of frequencies , @xmath121 and for not too strong interactions . for this reason , the tunneling density of states may provide the simplest and most transparent test for sill physics in experiment . tunneling into an infinitely long sill has been considered by @xcite . deep in the spin - incoherent regime , the first - quantized path integral representation of the bosonized electron s green s function is a convenient method @xcite . here we instead follow @xcite as it provides a better launching point for numerical studies in the regime @xmath80 , and also readily captures some other important features of the tunneling density of states . the tunneling density of states , @xmath122 $ ] , where @xmath123 is the fourier transform of the retarded green s function . the retarded green s function generally contains particle and hole contributions , but for illustrative purpose , we will only focus on the particle contributions . hole contributions are conceptually and technical similar , and in the spin - incoherent regime it can be shown that @xmath124 for @xmath125 @xcite . here we focus on the particle contribution : @xmath126 . using the general formulas eq.([eq : psi_final_bos ] ) and eq.([eq : little_z ] ) we have @xmath127-\phi(x , t ) ] } e^{-i[(1+\frac{q_2}{\pi})\theta(0,0)-\phi(0,0)]}\rangle,\;\;\;\end{aligned}\ ] ] where we have assumed the time evolution of the spin degrees of freedom are slow compared to the charge degrees of freedom so that we can neglect their time dependence @xcite . in a translationally invariant system , one has @xmath128 . making a change of variables @xmath129 , the summation over the sites of the spin chain results in a delta function , @xmath130 , which immediately kills one of the momentum integrals and sets @xmath131 . the resulting green s function is @xmath132 where @xmath133 and @xmath134 } e^{-i[(1+\frac{q}{\pi})\theta(0,0)-\phi(0,0)]}\rangle.\ ] ] since the holon sector is described the the gaussian theory eq.([eq : h_holon ] ) we can make use of the identity @xmath135 for operator @xmath136 and evaluate the correlation function eq.([eq : g ] ) at zero temperature with respect to the charge energy . [ generalization to small but finite temperatures in the charge sector is trivial @xcite . ] the expression for the hole - like contributions to the green s function has the same structure as eq.([eq : g_final_trans ] ) and is reported in @xcite . since for low temperatures compared to the charge energy the holon correlations in eq.([eq : g ] ) can be analytically evaluated in a straightforward way , the emphasis for arbitrary temperatures with the respect to the spin energy is on the spin correlations described by eq.([eq : c ] ) . for arbitrary temperatures the evaluation of @xmath137 is non - trivial @xcite and it is precisely at this point that numerical input for both zero and non - zero magnetic fields would be most helpful . keeping for the moment temperature arbitrary with respect to @xmath14 , but effectively zero compared to @xmath106 one finds for the tunneling density of states @xcite , @xmath138 where @xmath139 - 1 $ ] and @xmath87 is the luttinger interaction parameter appearing in eq.([eq : h_holon ] ) . from the exponent @xmath140 one sees that the frequency dependence @xmath141 which is maximum for @xmath142 . expanding about this point and doing the guassian integration , one finds @xmath143 which agrees with the result earlier obtained for the infinite @xmath79 limit of the hubbard model @xcite and for general interaction parameter @xmath71 in the spin - incoherent regime @xcite . note , however , that the result eq.([eq : nu_structure ] ) is valid at @xmath144 . the only assumptions were that the dominant weight comes from the point @xmath142 and that @xmath145 . if one is deep in the spin - incoherent regime @xmath146 _ independent of the details and form _ of @xmath147 [ a simple argument for this is given in @xcite ) ] , and one has @xmath148 . hence , the assumptions leading to eq.([eq : nu_structure ] ) are valid and the result indeed applies in the spin - incoherent regime as well . on the other hand , if the spins are fully polarized ( along the z - axis , say ) then @xmath149 and the tunneling density of states will necessarily have a different form ( in fact , the familiar ll form ) from eq.([eq : nu_structure ] ) @xcite . also at @xmath144 and zero magnetic field one has @xmath150 so the logarithmic factor appearing in eq.([eq : nu_structure ] ) is absent and the ll form is obtained . finally , note that in the spin - incoherent regime where eq.([eq : nu_structure ] ) applies the tunneling response of a sill looks qualitatively like the behavior in fig . [ fig : schematic_threshold](a ) with @xmath45 for @xmath151 , but it _ does not _ follow the analytical form of eq.([eq : i_fl ] ) because of the logarithmic corrections . if the interactions are both strong and long - ranged , it is possible to obtain @xmath152 in which case the tunneling density of states would look qualitatively like the behavior in fig . [ fig : schematic_threshold](b ) with @xmath45 although it would still not fit the analytic form in eq.([eq : i_fl ] ) . [ recall that for zero - range interactions the smallest value @xmath153 can take is @xmath154 , obtained in the infinite-@xmath79 limit of the hubbard model @xcite . ] in this sense , the tunneling response of the sill is in a different universality class from both the fermi liquid and luttinger liquid . we will now see that the same logarithmic corrections can also appear in the frequency dependence of the photon absorption . as we discussed earlier , the physics of the fermi - edge singularity falls into two classes : ( 1 ) electron backscattering from the hole is an irrelevant perturbation . ( 2 ) electron backscattering from the hole is a relevant perturbation . for a luttinger liquid these two regimes are distinguishes by whether the hole has a finite mass , case ( 1 ) , or an infinite mass , case ( 2 ) . in the sill there is a subtlety because the relevance or irrelevance of backscattering from a local potential ( for example , an infinite mass core hole ) depends on the interactions in the system through the luttinger parameter @xmath71 . in particular , a local impurity is only a relevant perturbation if @xmath155 [ which is the same as @xmath69 ] @xcite . if @xmath157 the interactions of ll are repulsive , but a local impurity is an _ irrelevant _ perturbation . therefore , in the sill we effectively have 3 possible regimes for the fermi - edge singularity : ( i ) finite mass mobile hole for which backscattering is irrelevant . ( ii ) infinite mass core hole for which backscattering is irrelevant . ( iii ) infinite mass core hole for which backscattering is relevant . let us first consider the case of a finite mass hole . in this case the backscattering is always an irrelevant perturbation . for addressing the threshold behavior of the photon absorption we can therefore neglect backscattering terms and focus only on forward scattering interactions between the electron and the hole . in the initial study of the fermi - edge singularity in a ll @xcite it was incorrectly concluded that the exponent @xmath2 is independent of the dynamics of the hole , _ i.e. _ its mass . a more careful treatment showed that indeed the mass of the hole enters @xmath2 @xcite . the work of tsukamoto _ et al . _ employed a combined bethe ansatz - conformal field theory study of a zero range interaction between particles . the asymptotic structure of the energy and critical exponents in this model could be obtained exactly . from the formulas obtained it was shown explicitly that the exponent @xmath2 depends on the mass of the hole . perhaps the most important result is the structure of the electron - hole interaction : there are actually two forward scattering terms that emerge . one is the familiar density - density interaction between hole density and electron density . the other is less familiar and it describes the interaction of the hole with a current of electrons that is flowing by it . in other words , it is natural to transform to frame co - moving with the hole @xcite . in this frame the hole sees a current of electrons . one might think that such a picture would break time - reversal symmetry , but in an actual experiment electrons would be excited near both the right and left fermi points shown in fig.[fig : direct_transition](a ) . thus , as a whole , time - reversal symmetry is preserved . as the contribution to the absorption edge is the same in both cases , one need only focus on one fermi point . our discussion below will cover both fermi points in one fell swoop , as the electron operator eq.([eq : psi_final_bos ] ) contains both right and left moving pieces . the fermi - edge singularity in a spin - incoherent luttinger liquid was discussed by @xcite . the full hamiltonian for our problem in the frame co - moving with the hole is @xmath158 , where @xmath159 is given by eq.([eq : h_holon ] ) and eq.([eq : s_chain ] ) , and @xmath160 with @xmath161 and @xmath162 . here @xmath163 is the symmetric part of the forward scattering from the hole and @xmath164 is the antisymmetric part of the forward scattering @xcite . ( in our convention @xmath165 represents the density fluctuations and @xmath166 the particle current . ) again , the antisymmetric part appears since in the frame of the hole , it sees a net current of particles scattering from it . the `` + '' sign is for a right - moving hole and the `` - '' sign is for a left - moving hole . the parameter @xmath164 depends on the momentum and mass of the hole , and when it is at rest , @xmath167 @xcite . the operator @xmath168 creates a hole with spin @xmath96 and @xmath169 annihilates a hole with spin @xmath96 . the energy of such a hole in its rest frame is @xmath170 , which in the presence of an externally applied magnetic field can depend on @xmath96 . to compute the photon absorption we treat the electron - photon interaction in second order perturbation theory . standard manipulations show @xcite @xmath171 to evaluate the correlations appearing in eq.([eq : threshold ] ) we must diagonalize the hamiltonian . the hamiltonian @xmath172 can be diagonalized with the unitary transformation @xmath173h^\dagger h\}\ ] ] where @xmath174 and @xmath175 applying this transformation we find @xmath176 , where the only change to @xmath177 is a shift in the hole energy @xmath178 , which is unimportant to us here . our main concern is with computing the threshold exponent @xmath2 and the functional form of the response . we begin with the evaluation of the correlation function @xmath179 , where @xmath180 is the imaginary time . at finite temperature , @xmath181 \\ & = & \frac{1}{z}{\rm tr}\left[e^{-\beta \bar h } \bar \psi_\sigma(\tau)\bar h_\sigma(\tau)\bar h_\sigma^\dagger(0)\bar \psi_\sigma^\dagger(0)\right],\end{aligned}\ ] ] where in the second line we have introduced the unitary transformation @xmath182 and used the cyclic property of the trace . we have already determined @xmath183 , so it remains to determine @xmath184 and @xmath185 . direct evaluation gives @xmath186 ( up to unimportant multiplicative factors ) and @xmath187 } $ ] . therefore , the correlation function @xmath188 separates as @xmath189 . one readily finds @xmath190 at zero temperature . this factor will enter the threshold frequency @xmath1 in eq.([eq : threshold ] ) and will in general also depend on the external magnetic field . our main interest here is in the evaluation of the part of the correlation function that will give us the frequency dependence just above threshold , @xmath191 } e^{i[\delta_a\theta(0)+\delta_s\phi(0 ) ] } \psi_\sigma^\dagger(0)\biggr],\end{aligned}\ ] ] where @xmath192 $ ] . formally , eq.([eq : c_psi ] ) bears a striking resemblance to the single particle green s function evaluated in the spin - incoherent regime above . it is worth pausing a moment to understand the physics of eq.([eq : c_psi ] ) . recalling that the operator @xmath193 creates a particle at space - time point @xmath194 in the many - body system and the electron number is related to the @xmath84 field via @xmath195 , with @xmath196 the average particle density , we see immediately that @xmath197 involves adding an electron _ plus _ an additional background excitation " ( from @xmath198 ) and then removing the same particle and its additional background at a time @xmath180 later . the factors @xmath199 contribute additional density fluctuations coming from the motion of the finite mass valence hole . this factor is absent in the infinite mass limit . it remains to evaluate the trace in eq.([eq : c_psi ] ) . we again follow the more general bosonization scheme for strongly interacting systems described above which reduces the to previously obtained results deep in the spin - incoherent regime @xcite . we have @xmath200 where @xmath201 } e^{-i[(1+\frac{q}{\pi}-\delta_a)\theta(0)-(1-\delta_s)\phi(0)]}\rangle.\ ] ] in the zero temperature limit with respect to the charge degrees of freedom , we can evaluate eq.([eq : f ] ) using eq.([eq : h_holon ] ) as we did for the single - particle green s function . the result is @xmath202}.\ ] ] the frequency dependence of @xmath72 can now be computed from eq.([eq : threshold ] ) using @xmath189 . the computation is essentially identical to that of the tunneling density of states , eq.([eq : tdos ] ) , only @xmath203 and @xmath204 - 1 $ ] . therefore , we have @xmath205 where the @xmath206 factor tells us there is no photon absorption below below a threshold frequency @xmath1 . from fig.([fig : direct_transition])(a ) it is clear that @xmath207 is of order the band gap plus the fermi energy of the electrons plus the absolute value of the energy of the hole near the fermi points of the electron gas . as with the expression eq.([eq : tdos ] ) for the tunneling density of states , eq.([eq : i_gen ] ) is the general expression for the photon absorption in an interacting one - dimensionsal system . for temperatures low compared to the charge energy , but arbitrary with respect to the spin energy , the heavy lifting is in the evaluation of @xmath137 as we emphasized in the discussion of the single - particle green s function . following the same arguments we used there , we see that the frequency dependence @xmath208(1+q/\pi-\delta_a)^2}$ ] which is a maximum for @xmath209 . expanding about this point and doing the guassian integration , one finds @xmath210 note that the final form of the photon absorption eq.([eq : i_mobile_final ] ) is almost identical to that of the tunneling density of states , eq.([eq : nu_structure ] ) , aside from the factor @xmath75 that appears in the threshold exponent . it is instructive to compare the similarities and differences of the fermi - edge singularity with a mobile hole in the ll and sill . comparing with the general formula eq.([eq : i_fl ] ) we see that the threshold behavior is different in the sill compared to the ll and even the fermi liquid . the threshold behavior of the sill contains the same logarithmic corrections that we found earlier in the tunneling density of states . focusing on the power - law part of the threshold behavior we see that @xmath211 so that it is independent of @xmath74 , unlike the ll case in eq.([eq : a_ll_mobile ] ) . since the asymmetric phase shift does not appear , this implies the threshold exponent is _ independent of the hole dynamics and therefore the mass of the hole _ in the spin - incoherent case . of course , from eq.([eq : i_mobile_final ] ) it is clear that the effects of @xmath74 are still felt , but only as a prefactor to the frequency dependence . the effects of an externally applied magnetic field can also be easily incorporated @xcite . for a quantum wire the magnetic field does not affect the orbital part of the wavefunction until the magnetic length is of the order of the wire width . we therefore only discuss the coupling of the magnetic field to the spin of the electron , _ i.e. _ a zeeman coupling . the magnetic field will influence the spin correlations described by @xmath137 . in the spin - incoherent regime with zero field one has @xmath212 , independent of @xmath96 . in a finite external field one instead has @xmath213 where @xmath214 and @xmath215 is the zeeman energy @xcite . while the presence of the magnetic field may shift the threshold energy , it will not affect the exponent eq.([eq : a_sill_mobile ] ) @xcite . the magnetic field also gives rise to a frequency scale @xmath216 below which sill behavior is observed and above which behavior characteristic of a spinless ll is observed @xcite , namely the form eq.([eq : i_fl ] ) with the exponent eq.([eq : a_ll_mobile ] ) . we next turn to the case of an infinitely massive impurity for which backscattering is irrelevant . in this case the effective electron - hole interaction is given by eq.([eq : e - h ] ) only with @xmath76 as this corresponds to the case of a stationary hole with no backscattering . the photon absorption in this case is thus given by eq.([eq : i_mobile_final ] ) with @xmath76 . while an infinite mass hole breaks translational symmetry , the irrelevance of backscattering effectively restores translational symmetry at low energies and the threshold behavior maps onto the finite mass case . on the other hand , when backscattering is relevant the situation is very different . at low energies the hole acts as a boundary to a semi - infinite one dimensional system @xcite . in this case , the sill maps onto a _ spinless _ electron system @xcite which allows us to immediately exploit @xcite the results obtained for that system . in particular , one obtains a universal contribution of 1/8 to the threshold exponent @xcite , @xmath217 and the frequency dependence follows eq.([eq : i_fl ] ) _ without _ the logarithmic corrections . in fact , for the infinitely massive hole with relevant backscattering the only signature of spin - incoherent physics is hidden in the mapping @xmath87 @xcite . while our theoretical understanding of the spin - incoherent regime has advanced rapidly , the difficulty of obtaining the regime @xmath10 has hampered experimental tests of the theory . to date our best experimental evidence comes from momentum resolved tunneling on cleaved - edge overgrowth quantum wires @xcite , transport in split - gate devices @xcite , and somewhat more speculatively in transport measurements on gated single wall carbon nanotubes with low electron density @xcite . given that the existing theory now encompasses hybrid structures of sills with ferromagnets and superconductors as well @xcite , the number of experimental groups poised to make contributions to this exciting field is greatly enlarged . for a recent discussion of numerous possible experiments that would provide strong evidence for spin - incoherent luttinger liquid behavior see @xcite . on the theoretical side , a number of open issues remain . in particular , it is desirable to have a better understanding of the behavior on temperature scales @xmath218 which is most likely very relevant for many experiments that approach but do not quite reach the spin - incoherent regime . related to this is the need for a better understanding of the crossover between the luttinger liquid and the spin - incoherent luttinger liquid regimes . both of these issues will likely require a numerical attack as there are no obvious analytical methods available to address them . finally there is the issue of spin - orbit coupling that has so far received no attention . for very strong spin - orbit coupling is there novel behavior in the regime @xmath219 ? these issues await further study . i am grateful to my collaborators ophir auslaender , leon balents , adrian feiguin , thierry giamarchi , bert halperin , anibal iucci , markus kindermann , karyn le hur , jiang qian , hadar steinberg , dagim tilahun , yaroslav tserkovnyak , and amir yacoby for their important contributions to our work on spin - incoherent luttinger liquids . i am also indebted to matthew fisher , akira furusaki , leonid glazman , and kostya matveev for many enlightening discussions on this topic . this work was supported by nsf grant no . phy05 - 51164 and the lee a. dubridge foundation .

recently , important efforts in the fields of molecular , atomic and nuclear physics have been devoted to the analysis of collision processes involving composite quantum systems . despite the peculiarities of the different fields , the theoretical description of the collision has several common features . r - matrix theory @xcite is used for re - arrangement collisions , when several electrons or nucleons may be exchanged between the colliding systems . coupled channels calculations @xcite are used to describe excitation and dissociation in atomic , molecular and nuclear systems . the `` adiabatic '' or `` sudden '' approximation is often invoked in molecular and nuclear collisions , because it simplifies significantly the description of the scattering process , by considering that some of the relevant co - ordinates are effectively frozen during the scattering process . also , the description of the collision of atomic and molecular beams with surfaces requires coupled channels descriptions , but the combined difficulties of the atom - surface interactions and molecular vibrations makes the `` sudden '' approximation almost essential for its solution @xcite . in the case of atomic and molecular physics , one - dimensional atom - molecule collision reactions have been studied in detail by solving the schrdinger equation @xcite . in particular , collisions between an atom and a diatomic molecule represented by harmonic @xcite , an - harmonic @xcite and morse oscillators @xcite have been analysed with different degrees of approximation . diverse computational methods have been also implemented to study three - dimensional molecular and atomic collisions @xcite . moreover , in some recent papers algebraic approaches have been proposed for describing one- and three - dimensional atom - molecule collision processes @xcite . in the field of nuclear physics , much of the interest in recent years has been focused on the study of the properties of halo nuclei , weakly bound systems characterised by the existence of one or two particles ( generally neutrons ) with a high probability of being at distances larger than the typical nuclear radius . different approaches have been used in the analysis of reactions involving halo nuclei . the adiabatic approximation @xcite assumes that , for sufficiently high scattering energies , the internal hamiltonian is accurately represented by its corresponding eigenvalue for the ground state . the sudden approximation @xcite relies on two main assumptions : i ) the impulse approximation , i.e. , the multiple scattering expansion for the t - matrix is approximated by the sum on the individual t - matrices for the scattering of the separated constituents , and ii ) only one of the particles of the projectile interacts with the target . in the high energy regime , the glauber approach @xcite , which combine eikonal dynamics with the adiabatic approximation , provides a simple tool to analyse reactions involving halo nuclei . we have recently developed an alternative approach to the description of weakly bound systems . the approach , called `` uncorrelated scattering approximation '' ( usa ) @xcite , is based on the fact that , for a weakly bound projectile , the correlations between the constituents are weak and so , to some extent , they are expected to evolve independently in the strong field of a heavy target . thus , the three body s - matrix can be expressed in terms of the individual two body s - matrices for the scattering of the constituents . in this framework , the scattering observables of the process are given in terms of two body constituent - target observables . this model has been applied to describe elastic scattering and break - up of deuterons on heavy targets , with encouraging results . the uncorrelated scattering approximation has certain relation with the r - matrix approach . in a r - matrix calculation of deuteron scattering , the wave - function within the range of the target interaction is given in terms of products of single - particle ( protons and neutrons ) wave - functions with the proper boundary conditions , provided the interaction between the fragments of the projectile is neglected . these single - particle wave - functions should be matched with the proper asymptotic wave - functions . in the usa calculation , the incident wave is expanded in terms of products of fragment - target wave - functions , which then scatter independently from the target . the objective of this work is to investigate the scattering of a composite system from a target with interactions which have a very short range compared not only with the size of the system , but also with the associated wave - length of the projectiles . for that purpose , we consider the case of two particles , interacting through a harmonic oscillator potential , which collide with an infinite wall . we develop two different methods to solve the problem exactly , obtaining the wave - function as well as the s- matrix , or reflection coefficients , which give the probability amplitudes for the excitation of the different oscillator states . we compare the exact results with the ones obtained using the adiabatic approximation , and the uncorrelated scattering approximation . the model discussed here does not pretend to be a realistic representation of any specific molecular , atomic or nuclear system . however , it has the virtue that the only length scale is the oscillator length @xmath0 , while the only energy scale is the oscillator parameter @xmath1 . then , the results obtained , which are expressed in terms of @xmath2 and @xmath3 , can be applied in principle to arbitrary energy and length scales , which may be nuclear , atomic or molecular . this fact makes the model attractive as a bench - mark to test the validity of the approximations which are used for the description of composite systems . one aspect of the model which seems odd is the infinite nature of the harmonic oscillator interaction between the fragments . this would prevent dissociation in the case of molecules , ionization in the case of atoms or break - up for nuclei . on the other hand , the harmonic oscillator basis is complete , and thus the results of the model incorporate effects on the scattering due to coupling to all possible , open or closed , states . the paper is organised as follows : in section ii we describe the problem to be solved . two different approaches to extract the exact solutions are developed . in section iii we derive the s - matrix in the adiabatic model approach . in section iv we present the uncorrelated scattering approximation . in section v we apply it to evaluate the scattering coefficients and compare with the exact solution . section vi is devoted to summary and conclusions . let us consider the scattering of a two - particle bound system by the short range potential due to a heavy target , placed at the origin of coordinates . the hamiltonian for the system may be written in terms of the coordinates of the two particles ( @xmath4 , @xmath5 ) and their momenta ( @xmath6 , @xmath7 ) as @xmath8 where @xmath9 are the masses of the constituents , @xmath10 is the potential exerted by the target on each particle and @xmath11 represents the binding interaction . alternatively , it can be expressed in terms of the relative ( @xmath12 ) and centre of mass coordinate ( @xmath13 ) : @xmath14 where @xmath15 is the total mass , @xmath16 ( @xmath17 ) , and @xmath18 is the internal hamiltonian . here , @xmath19 represents the reduced mass of the two - particle system and @xmath20 is the relative coordinate . we refer to the specific case in which @xmath10 corresponds to an infinite wall potential , i.e. : @xmath21 for negative values of @xmath4 and @xmath5 the total wave function must vanish . for positive values , the hamiltonian comprises two terms : one associated with the internal motion of the projectile ( @xmath18 ) and the other describing the centre of mass motion ( @xmath22 . we take @xmath23 to be a harmonic interaction . therefore , the eigenfunctions for the internal hamiltonian are given by @xmath24 where @xmath25 is the oscillator length , @xmath26 a normalisation constant and @xmath27 the hermite polynomial of order @xmath28 . denoting by @xmath29 the number of open channels , the total wave function will be then expanded in terms of eigenstates of the hamiltonian as follows : @xmath30 in this expression @xmath31 represents the centre of mass momentum associated with the internal state @xmath32 and @xmath33 its velocity . energy conservation applied to the whole system leads to the constrain @xmath34 where @xmath35 is the eigenvalue corresponding to the internal state @xmath36 . the first term in ( [ eq : wf ] ) represents an incoming wave , normalised to unit flux , coming from @xmath37 . the second term contains the set of scattered waves travelling in the positive @xmath13 direction corresponding to open channels , i.e. , @xmath38 . the coefficients @xmath39 are the s - matrix elements or , strictly speaking , reflection coefficients . then , @xmath40 represents the amplitude probability of populating the state _ @xmath32 _ during the collision starting from a wave function in its ground state . the last term in ( [ eq : wf ] ) contains the contribution to the wave function due to the ( infinite ) set of closed channels , for which @xmath41 . for these states , the associated momentum @xmath31 is a pure imaginary quantity , and the centre of mass motion is described by an exponential decaying behaviour . therefore , they do not contribute to the asymptotic wave function and so they do not give direct contribution to the outgoing flux . although they are usually ignored in practical calculations , the peculiarity of the infinite potential requires however the inclusion of these states in order to describe correctly the wave function in all the space . in this problem there is no transmitted wave due to the presence of the wall and so the s - matrix should fulfil unitarity . therefore , the ingoing and outgoing flux must be equal : @xmath42 where the sum extends to the set of open channels . the coefficients @xmath39 and @xmath43 are determined by imposing the boundary conditions at @xmath44 . for the infinite wall the wave function must vanish at @xmath45 and @xmath46 which , in terms of @xmath12 and @xmath13 means @xmath47 in what follows we present two alternative methods to solve _ exactly _ this problem . as described above the exact solution of the scattering problem of a two - body projectile by a rigid wall is accomplished by applying the boundary conditions ( [ eq : boundary_cond_wall ] ) to the general solution ( [ eq : wf ] ) . for simplicity we assume that the system is initially in its ground state . the first condition in ( [ eq : boundary_cond_wall ] ) leads to the following equation for the scattering coefficients : @xmath48 the second condition in ( [ eq : boundary_cond_wall ] ) leads to a similar equation ( with @xmath49 instead of @xmath50 ) which holds for @xmath51 . in particular , we have performed calculations for the particular case of equal masses , i.e. , @xmath52 , for which both equations are identical , due to the symmetry of the problem under the exchange of @xmath4 and @xmath5 . in the remaining of this section we restrict to this particular case . in order to transform eq . ( [ eq : phi(r , r/2 ) ] ) into an ordinary algebraic equation , we require that @xmath53 which gives for the scattering coefficients the relation : @xmath54 where @xmath55 and @xmath56 the expressions for @xmath57 and @xmath58 for @xmath59 are obtained substituting @xmath60 for @xmath61 . differentiating eq . ( [ eq : bound_1 ] ) with respect to @xmath62 we get the following linear system in the variables @xmath63 : @xmath64 for practical calculations the resolution of this system requires the truncation of the sum at some finite value of @xmath32 . it should be noticed that in order to achieve convergence for the s - matrix elements one is forced to include in the calculation several closed channels . otherwise , the boundary conditions ( [ eq : boundary_cond_wall ] ) are not accurately fulfilled . in this section we present an alternative method to obtain the exact solution of the problem stated above . the method relies on the introduction of a new basis of states which are particular linear combinations of the internal wave functions . the new functions have the property of being highly localised in configuration space . as we shall see this peculiarity allows to apply more easily the boundary conditions . we start with a truncated basis of @xmath29 eigenstates for the internal hamiltonian that we denote by @xmath65 . thus , according to our previous notation , @xmath66 . in the appendix we show how these states can be decomposed in terms of configuration localised states ( cls ) as @xmath67 represents a configuration localised state . explicit expressions for the cls associated with the ho wave functions can be found in the appendix _ _ _ _ and in ref . the function @xmath68 has the property of being highly localised around @xmath69 , the @xmath70-th zero of the eigenfunction @xmath71 . the problem is significantly simplified in the case of fragments of equal masses . parity conservation guarantees that only states with the same parity as the incident wave function will be populated in the process . this allows to work in the subspace : \{@xmath72 ; @xmath73 } ( for simplicity of the notation , and without loss of generality , we take @xmath29 even ) . these functions are symmetric with respect to their natural variable , @xmath12 . from this set of @xmath74 states it is possible to construct , by means of a transformation similar to ( [ eq : cls ] ) , a set of _ symmetric configuration localised states _ ( scls ) , which are also even functions with respect to the variable @xmath12 . we denote this new set of states by \{@xmath75 } , where the index @xmath70 runs over the positive zeros of @xmath76 . the details of its derivation are presented in the appendix . the state @xmath77 has the property of being localised around the points @xmath78 , where @xmath79 is the @xmath80th positive zero of the hermite polynomial @xmath76 . in the treatment that follows we make extensive use of this remarkable signature . the boundary condition due to the wall requires that the total wave function vanishes for @xmath45 and @xmath46 or , in terms of coordinates @xmath12 and @xmath13 , along the lines @xmath81 ( @xmath82 ) and @xmath83 ( @xmath84 in the ( @xmath85 ) plane . at each value of @xmath13 a reflected wave is generated , interfering with the other outgoing waves to construct the total scattered wave . this picture is simplified working in terms of scls . the part of the wave function associated to the state @xmath77 is peaked around @xmath86 and , therefore , it will be mainly scattered around @xmath87 . in the limit case @xmath88 , @xmath77 becomes a delta function in @xmath12 and the associated wave is exactly scattered at @xmath89 . moreover , continuity of the wave function implies that a reflected wave , affected by a phase factor @xmath90 , is generated at this point . obviously , this is not exactly our situation as , in practise , @xmath29 is always finite and so our localised states have a finite dispersion around @xmath91 . however , we can make this dispersion as small as required by increasing the number of states . consequently , the process is considered as a distribution of localised states that are reflected at some definite _ barriers _ in the @xmath13 direction . the total scattered wave is given by the superposition of the reflected waves . to make the treatment clearer , we divide the @xmath13 axis in @xmath74 regions , delimited by the values of @xmath92 . we order the zeros of the hermite polynomial @xmath93 in such way that @xmath94 . let us introduce an index @xmath95 to label each region ( @xmath96 , such that @xmath97 corresponds to the asymptotic region , i.e. , @xmath98 , before any localised state has been reflected . in this region the basis space associated with the internal motion is described in terms of the first @xmath74 ho eigenfunctions with positive parity . alternatively , it can be described in terms of @xmath74 symmetric localised states . at @xmath99 the scls corresponding to @xmath100 is reflected and removed from the incident wave function , while the rest of scls remain unaltered . therefore , in the region @xmath101 the basis is limited to the subspace spanned by the remaining @xmath102 states ( @xmath103 ) . subsequently , our original set of states are no longer eigenstates of the hamiltonian in this region . instead , a new family of @xmath102 eigenstates must be calculated , by diagonalizing the hamiltonian in the basis constituted by the remaining @xmath102 localised states . the method is schematically illustrated in fig . [ fig : exact_cls ] . the picture represents the scattering problem in the @xmath104 plane . the total wave function travels in the @xmath13 direction ( vertical axis ) and must vanish along the lines @xmath45 and @xmath46 , which have been also plotted for reference . the case with @xmath105 has been considered , in which the incoming wave is decomposed in a set of four scls . each one of these scls is consider to scatter at a definite barrier in the @xmath13 axis , labelled from @xmath106 to @xmath107 . note that they partially extend to the forbidden region ( @xmath108 ) , which is a consequence of the truncation of the ho basis . let us specify explicitly the boundary condition at @xmath109 . this barrier separates regions @xmath95 and @xmath110 . the eigenstates corresponding to region @xmath95 will be denoted by \{@xmath111 ; @xmath112 } where @xmath113 is the number of states in region @xmath95 . the total wave function is then expanded in each region in terms of the corresponding eigenstates . for region @xmath95 , corresponding to the interval @xmath114 , we have : @xmath115 where @xmath116 ( @xmath117 ) are the coefficients of the incoming ( outgoing ) waves . energy conservation requires , in analogy with ( [ eq : e(kn , en ) ] ) , @xmath118 where @xmath119 is the @xmath32-th eigenvalue of the hamiltonian in region @xmath95 and @xmath120 its corresponding centre of mass momentum . similarly , for region @xmath110 the wave function will be written as @xmath121 at @xmath109 the part of the wave function associated with the localised state @xmath122 is reflected by the wall . therefore , continuity of the wave function requires that @xmath123 by contrast , the rest of localised states are unaffected by this barrier , so we may require continuity of the wave function and its derivative for the components @xmath124 : @xmath125 _ { r = r_{i } } & = & \left [ \frac{d}{dr}\langle scls;n\ , s|\psi ^{(i+1)}(e)\rangle \right]_{r = r_{i}}.\label{eq : condicion_ri+1_b } \end{aligned}\ ] ] equations ( [ eq : cont_ri+1 ] ) to ( [ eq : condicion_ri+1_b ] ) give rise to a set of _ @xmath126 _ conditions for the scattering coefficients @xmath127 and @xmath128 ( _ @xmath129 _ ) . when applied to all regions a total of @xmath130 equations with @xmath131 coefficients is obtained . from these , the @xmath74 coefficients @xmath132 are known , as they are determined by the initial conditions . therefore , there remain @xmath133 coefficients to be determined , that coincides with the number of equations . thus , the resolution of the problem reduces to the calculation of the inverse of a @xmath134 matrix . this can be a slow computational task when @xmath29 takes large values . this drawback has led us to adopt an alternative method that avoids this sort of calculations , speeding significantly the computational time . for its application , it is convenient to define a generalised `` s - matrix '' for region @xmath95 as : @xmath135 for region @xmath97 this definition gives just the usual s - matrix . the factor @xmath136 has been introduced to ensure unitarity . substituting this expression into equations ( [ eq : cont_ri+1 ] ) to ( [ eq : condicion_ri+1_b ] ) we get a set of @xmath126 equations relating the s - matrix elements of @xmath137 and @xmath138 . then , instead of solving the system of equations as a whole , we perform an iterative calculation in which the s - matrix for each region is determined in terms of the s - matrix for the neighbouring region . the starting point is the region @xmath139 , for which the s - matrix is known . in this region there is only one state left , which is completely reflected at @xmath140 . the s - matrix in this case is just the phase factor : @xmath141 . from this , we can determine the s - matrix for region @xmath102 , and so on . finally , we obtain the s - matrix , @xmath142 , in terms of @xmath143 . taking into account ( [ eq : bsa ] ) and the fact that @xmath144 are given by the initial conditions , it is possible to determine the coefficients @xmath145 . one can derive also an iterative procedure to determine the coefficients @xmath127 for all regions in terms of the corresponding s - matrices . in order to compare both treatments we have plotted in figure [ fig : sexact(e ) ] the s - matrix coefficients @xmath146 , @xmath147 and @xmath148 , as a function of the scattering energy . the energy scale is in units of @xmath149 . a basis space with 15 even parity ho states was used for the calculation . the figure shows a good agreement between both treatments , specially at low scattering energies . as the scattering energy increases , the effect of the truncation of the basis becomes more important and both calculations differ slightly . notice that this discrepancy appears to be more evident for the phase of the s - matrix elements . nevertheless , this difference between both calculations is reduced as the number of basis states is increased . we notice that , as expected , the elastic coefficient is identically one for scattering energies below @xmath150 . this corresponds to the energy of the first excited state that is suitable to be populated . accordingly , the inelastic coefficients , @xmath151 @xmath152 are identically zero below this threshold . for scattering energies higher than @xmath150 , the modulus of the elastic coefficient is less than one and the rest of the coefficients are non zero . note that the modulus of @xmath146 presents a minimum at @xmath153 , indicating a maximal loss of flux from the elastic channel to other channels at this energy . for higher scattering energies the modulus of elastic scattering coefficient tends gradually to unity . we have also compared the wave functions in both approaches . in our problem the wave function is a complex quantity depending on two variables , @xmath12 and @xmath13 . for the exact direct calculation the wave function is given by ( [ eq : wf ] ) , where the scattering coefficients are calculated by imposing the boundary conditions ( [ eq : boundary_cond_wall ] ) . it must be noticed , however , that this expression only holds for @xmath154 or , equivalently , for @xmath155 . outside this range , the total wave function must be identically zero . the projection of the wave function on a state @xmath156 is calculated as @xmath157 for large values of @xmath13 ( compared to the spatial extension of the internal wave function ) the integral can be extended to the interval @xmath158 and the projection above is directly related to the corresponding s - matrix element : @xmath159 in the analysis based on cls , the total wave function is given in terms of a piecewise function of @xmath13 , according to ( [ eq : phi_cls ] ) . for a certain region @xmath95 , this wave function is written as a superposition of the eigenstates of the hamiltonian in this region . these eigenstates travel freely between the consecutive barriers @xmath160 and @xmath161 . thus , the projection on a state @xmath162 for a value of @xmath13 belonging to a region @xmath95 reads @xmath163 we notice that , except for the incoming region ( @xmath97 ) , the states @xmath156 and @xmath111 are eigenstates belonging to different hamiltonians and so orthogonality can not be directly applied to them . however , they are both given in terms of the scls and so the calculation of their overlap is straightforward . the results are shown in figure [ fig : wfmod_e5 ] , where we present the modulus of the projection of the total wave function on the internal states . the solid line corresponds to the treatment of section [ sec : exact_direct ] and the dashed line to the cls method . these wave functions have been obtained for a scattering energy @xmath153 . the @xmath13 scale is in units of @xmath164 . as shown in this figure both treatments give almost identical results , the small differences being attributed to the truncation of the infinite basis . a similar agreement between both approaches is found at other energies , provided that a sufficient number of states is included in each case . from this picture it becomes apparent the role played by each internal state . the state @xmath28=0 corresponds to the initial state . then , the asymptotic wave function contains both incoming and outgoing contributions of this state . this is reflected in the characteristic interference pattern of the curve @xmath165 . at this scattering energy , the states @xmath28=2 and @xmath28=4 are open channels , and so they give a non vanishing contribution to the asymptotic scattered wave . in fact , the curves tend to a constant value for large distances which , according ( [ eq:<n|phi>_asym ] ) , is proportional to the corresponding coefficient @xmath166 . by contrast , the state @xmath28=6 is closed . its exponential decaying tail reflects the fact that this state does not contribute _ directly _ to the asymptotic wave function , but it does give a non negligible contribution to the total wave function in the vicinity of the wall . as mentioned before , the inclusion of these states is essential in order to reproduce accurately the boundary conditions and to achieve convergence for the s - matrix elements . then , they indirectly affect the wave function in all the space . the approach based on cls presents some advantages compared to the treatment described in section [ sec : exact_direct ] . it allows to evaluate all the matrix elements @xmath39 , @xmath167 in a single calculation , i.e. , the initial state does not need to be specified . by contrast , in the previous approach a new calculation is required for each initial state . moreover , the method based on the cls preserves the general properties of the s - matrix for any value of @xmath29 . in particular , conditions ( [ eq : cont_ri+1 ] ) and ( [ eq : condicion_ri+1_b ] ) ensure that the total flux is conserved at each barrier . as a relevant consequence , the resulting s - matrix fulfils unitarity , regardless of the number of initial states chosen for the basis . we derive in this section an expression for the s - matrix in the adiabatic approximation @xcite . the standpoint of the approach is that a fundamental distinction is made between the two relevant coordinates of our problem , namely , the centre of mass coordinate , @xmath13 , and the internal variable , @xmath12 . the former is identified as a high - energy ( _ fast _ ) variable and the latter as a low - energy ( _ slow ) _ variable . in this dynamical regime , it is expected that the excitation energies @xmath168 associated with those excited states which are significantly populated , are such that @xmath169 , where @xmath170 is the incident energy of the projectile . under this assumption it seems reasonable to replace the internal hamiltonian by a representative constant . by choosing this constant as @xmath171 , the ground state energy , it is also guaranteed that the solution of the resulting approximate three - body equation satisfies the correct incident wave boundary condition . then , applying this approximation to the hamiltonian ( [ eq : hrr ] ) the schrdinger equation reads : @xmath172 \psi ^{ad}(r , r)=0.\ ] ] note that this approximate schrdinger equation is independent of the relative momentum between the fragments . then , its conjugate coordinate , @xmath12 , is a constant of motion , remaining _ frozen _ during the collision . thus , eq.([eq : sch_ad ] ) has to be solved for all values of a fixed separation @xmath12 . in the case of a rigid wall potential , @xmath173 and @xmath174 are zero for positive values of @xmath4 and @xmath5 , respectively . therefore , in this case the solution of eq . ( [ eq : sch_ad ] ) is given by the plane wave , @xmath175 , multiplied by an arbitrary function of @xmath12 . the most general solution verifying the boundary incident condition at infinity can be written as @xmath176 where @xmath177 is the ground state wave function and @xmath178 is a function determined by imposing the boundary condition at the wall . this requires that the wave function vanishes at @xmath179 . then @xmath180 the scattering coefficients defined by eq . ( [ eq : wf ] ) can be obtained by projecting the wave function ( [ eq : ad_wf ] ) onto the basis states @xmath181 . this gives rise to the following simple expression : @xmath182 we notice that in this expression no distinction is made between open and closed channels . actually , the approximation treats the full excitation spectrum of the internal hamiltonian as being degenerate in energy with the ground state . as a consequence , unitarity of the adiabatic s - matrix is only achieved when summing over the infinite set of eigenstates . in what follows we show that the adiabatic approximation is not adequate for the problem treated in this work . in fig . [ fig : sad ] the quantity @xmath183 is plotted versus the collision energy for the exact ( solid line ) and adiabatic ( dashed line ) calculations . this quantity can be interpreted as an excitation probability . the adiabatic prediction completely disagrees with the exact calculation , indicating that the assumptions involved in the adiabatic approximation are not adequate for this problem . we attribute this failure to the fact that , as revealed by the exact calculation , many internal states participate in this process__. _ _ this is probably due to the peculiarities of an infinite zero - range interaction . in this case , the momentum transferred to each particle by the wall is twice the incident momentum and so , when increasing the scattering energy , the expected excitation energy increases . during the collision time , i.e. , while one particle has collided with the wall , but the other still has not , the internal motion of the projectile is strongly excited . then , the assumption of the adiabatic model , i.e. , to consider the whole spectrum to be degenerated with the ground state , does not work properly in this case . after the collision , that is , when both particles have collided with the wall , the centre of mass momentum is reversed and the final excitation energy is small . thus , this model represents a case in which , although the final excitation of the projectile is small , the adiabatic approximation is inadequate , because during the collision the internal motion is strongly excited . from this discussion we conclude that one should be very careful in applying the adiabatic approximation when dealing with strong , short - range interactions . in the next section we develop a new method to treat this type of situations . the main goal of this section is to derive an approximated expression for the s - matrix corresponding to the one dimensional scattering of a two particle system in terms of the constituent - target scattering amplitudes . the results presented here are not restricted to the case of an infinite potential . thus , we start with a general derivation of the model ant later we particularize the results to the problem of a wall potential in order to compare with the exact solution . there are two opposite effects acting on a projectile in the process of the collision . the first one is the binding potential @xmath23 that tends to keep the system bound . the second one is the interaction with the target which , apart from governing the motion of the projectile centre of mass , is the responsible for exciting or breaking the system . the relative importance of these two effects depends importantly on the separation between the projectile and the target . in particular , for sufficient large distances between them the dominant interaction is clearly the mutual interaction between the constituents . thus , it seems reasonable to approximate the projectile - target potential by an average ( folding ) potential at sufficiently large distances . by contrast , when the bound projectile is close enough to the target the dynamical evolution of the projectile is mainly governed by the target interaction . in this case it is reasonable to neglect the correlations between the fragments . let us introduce a characteristic centre of mass distance @xmath184 separating these two regions . for distances @xmath185 , referred as the `` asymptotic region '' and denoted by the index i , we adopt an approximate hamiltonian in which the interaction with the target is neglected @xmath186 where @xmath18 is the internal hamiltonian ( [ eq : hint ] ) and @xmath187 with @xmath188 representing the folding potential between the projectile and target . asymptotically , the eigenstates of @xmath189 are just the product of the eigenstates of the internal hamiltonian @xmath18 times a plane wave in @xmath13 , subject to the restriction ( [ eq : e(kn , en ) ] ) . as mentioned before , there is only contribution to the asymptotic wave function coming from the open channels . denoting by @xmath29 the number of these states , we restrict the basis space to the set @xmath190 then , the total wave function corresponding to an incoming wave in an internal state @xmath28 will be written in the asymptotic region as @xmath191 for distances @xmath192 , in what we call `` interaction region '' ( denoted by ii ) , the hamiltonian ( [ eq : hr1r2 ] ) is approximated by @xmath193 where @xmath194 is a constant that substitutes @xmath23 . the eigenfunctions of @xmath195 can be expanded in terms of the product of eigenfunctions of the hamiltonians @xmath196 and @xmath197 . an eigenstate of @xmath198 corresponds to the distorted wave for the scattering of a particle under the potential @xmath10 : @xmath199 where @xmath200 is the asymptotic incident momentum for the constituent @xmath95 , and @xmath201 . asymptotically this distorted wave behaves as @xmath202 where @xmath203 is a plane wave with momentum @xmath200 and @xmath204 is the constituent - target s - matrix for the scattering energy @xmath205 . note that , in the case of the infinite wall , this expression is valid for all values of @xmath206 . thus , the eigenstates of the hamiltonian @xmath195 , corresponding to the asymptotic momenta @xmath207 and @xmath208 , can be expressed as products of the form @xmath209 according to ( [ eq : distwaves ] ) , the asymptotic expansion of the eigenfunction ( [ eq : wf(k1,k2 ) ] ) contains an incoming wave , @xmath210 , an scattered wave , , and two cross terms mixing ingoing and outgoing contributions , namely @xmath211 and @xmath212 . in practise , these last two terms do not contribute to the incoming nor the scattered wave function at large distances . this can be verified by considering an incoming wave packet in @xmath207 and @xmath208 . the superposition of states of the form @xmath213 or @xmath214 contains an incoming part which vanishes for @xmath215 , and an outgoing part which cancels for @xmath216 . then , these cross terms can be omitted as far as the asymptotic behaviour concerns . as we will show later , the use of mls allows to demonstrate that these terms do not contribute to the wave function at large distances in a time - independent formalism . the form of the scattered wave , @xmath217 , indicates that the s - matrix for an incoming wave with definite values of @xmath207 and @xmath208 , denoted @xmath218 , appears to be the product of the individual s - matrices for the constituents , i.e. , @xmath219 . this s - matrix is unitary , provided the individual s - matrices are unitary , i.e. , @xmath220 . this condition is satisfied for the infinite potential , but it also holds for any other situation for which the transmission coefficient is zero . therefore , the scattering problem for the hamiltonian @xmath195 corresponding to a situation characterised by an incoming wave with definite values of the energies of the constituents can be easily solved . however , our physical initial state is not characterised by the individual energies of the two particles , but by a certain internal state of the projectile and the energy of the collision . the general solution in the interaction region for a total energy @xmath170 will be a certain superposition of eigenstates ( [ eq : wf(k1,k2 ) ] ) , verifying ( [ eq : e(e1,e2,v ) ] ) and the adequate asymptotic boundary conditions . these boundary conditions require that the wave function in region ii matches smoothly with the asymptotic wave function of eq . ( [ eq : wf_asym_usm ] ) . one possible way to proceed might be to expand the total wave function in each region in terms of the eigenstates of the approximated hamiltonian for that region . the coefficients of the expansion are determined by imposing the continuity of the wave function and its derivative at the matching radius @xmath184 . apart from the complexity of the calculation , this method has the problem that incoming waves coming from the asymptotic region do not match exactly with incoming waves of the interaction region , due to the discontinuity of the hamiltonian at @xmath184 . as a result , part of the incoming flux is reflected at @xmath184 and spurious outgoing waves are generated . to overcome this difficulty we proceed on a different way . in order to avoid the unphysical reflections we relax the meaning of the matching radius @xmath184 . we assume that the hamiltonian @xmath189 is smoothly transformed into @xmath195 in a finite transition region around @xmath184 . although we do not make an explicit description of this transition region in our model , we include its effect by requiring that the wave function passes from one region to the other without loss of flux . let us consider the incoming part of the general solution ( [ eq : wf_asym_usm ] ) : @xmath221 but , provided the number of states @xmath29 is large , their momentum distribution is highly localised around a certain value . the two basis of states are connected by an orthogonal transformation ( see appendix ) and so , the incoming state can be rewritten as @xmath222 where @xmath223 denotes a mls and @xmath224 are the transformation coefficients . the function @xmath225 has the property of being highly localised around @xmath226 , the @xmath70-th zero of the eigenfunction @xmath227 . as noted before , the eigenstates of the approximated hamiltonian in the interaction region are characterized by the energies of the two particles . these energies are directly related to their incident momenta . equivalently , they can be characterised by the asymptotic values of the internal momentum @xmath228 and the centre of mass momentum @xmath229 . in our approach , the internal eigenstates will be approximated in region ii by the discrete basis of mls . the incoming wave function for the interaction region is then expressed at large distances as @xmath230 with the centre of mass momentum @xmath231 defined by the relation @xmath232 the coefficients @xmath233 are determined in order the waves ( [ eq : wf_asym_usm ] ) and ( [ eq : wfin_ii ] ) match smoothly . in particular , we impose the incoming flux to be conserved in the transition . this can be achieved by taking @xmath234 with @xmath235 and @xmath236 . the phases @xmath237 and @xmath238 must be real numbers to preserve the incoming flux . in order to evaluate the phases @xmath237 and @xmath238 , we impose the two incoming solutions to have the same phase at @xmath184 . this is achieved by taking @xmath239 and @xmath240 , which leads to @xmath241 once the coefficients @xmath233 are known , the total wave function in the interaction region is completely determined . it can be expressed in terms of the distorted waves for each one of the constituents . to this end , we rewrite expression ( [ eq : wfin_ii ] ) as @xmath242 where we have introduced the momenta @xmath243 and @xmath244 . the scattering wave function corresponding to an incoming plane wave @xmath245 is given by the distorted wave @xmath246 . then , the total wave function in region ii , including both the incoming and scattered wave reads @xmath247 taking into account the asymptotic behaviour of the distorted waves , ( [ eq : distwaves ] ) , this wave function can be written beyond the range of the potentials as @xmath248 and @xmath249 , are smooth functions of the energy in the region where the integrand takes significant values , they can be evaluated at @xmath250 and @xmath251 , respectively . also , it is convenient to express the products of planes waves in terms of the relative and centre of mass momenta . considering the case of equal masses we can use the expressions @xmath252 and @xmath253 . then one can perform explicitly the integration with respect to @xmath228 , to obtain @xmath254 where @xmath255 is the fourier transform of @xmath256 . this wave function can be interpreted as follows . the incoming wave is decomposed as products of mls , @xmath255 , describing the internal evolution , times an incoming plane wave describing the centre of mass motion , @xmath257 . this wave scatters by the target giving rise to three terms . the second term in ( [ eq : wfii(r , r ) ] ) is just the conjugate of the incident wave , times the product of the s - matrices of the constituents . the remaining two terms comprise the product of the function @xmath258 , or its conjugate , times a plane wave in the variable @xmath12 . as it can be easily verified , the function @xmath258 vanishes for large values of @xmath13 and so , these two terms do not contribute to the asymptotic wave function . actually , these terms containing only one of the involved s - matrices can be physically regarded as the situation in which only one of the particle has scattered and the other has not yet . this is consistent with the fact that they both cancel at large distances . however , we remark that these vanishing terms are essential in order to reproduce the wave function at small distances . in this sense , it is also interesting to note that the wave function ( [ eq : wfii(r , r ) ] ) retains components associated to closed channels , even when they are explicitly omitted in the asymptotic region , according ( [ eq : wf_asym_usm ] ) . as noted in section ii , in the case of the wall potential the inclusion of these states is essential in order to reproduce the boundary conditions . for this particular problem , expression ( [ eq : wfii(r , r ) ] ) is valid for all the interaction region and , as can be easily verified , it identically fulfils the boundary conditions , vanishing for @xmath259 . thus , we can conclude that an incoming mls scatters with the product of the s - matrices of the fragments . this result is consistent with our previous discussion in which , using wave packet arguments , we concluded that the s - matrix is given by the product of the s - matrices of the fragments in the basis characterised by the momenta of the two particles . we now see that this property also holds for the mls basis which , in a sense , can be described as a wave packet of plane waves in terms of @xmath228 , centred around @xmath260 . therefore , writing explicitly the value of the coefficients @xmath233 the scattered wave in the interaction region behaves at large distances as @xmath261 by writing the mls appearing in this expression in terms of the original basis , and imposing the conservation of flux , one can easily obtain the scattered wave in the asymptotic region @xmath262 the expression between brackets provides the s - matrix element connecting an initial state @xmath28 with a final state @xmath32 @xmath263 as pointed before , the peculiarities of the problem treated in this work imply that , starting with the system in its ground state , only positive parity states are suitable to be populated . as far as expression ( [ eq : derived_snm ] ) concerns , this means that the matrix elements @xmath166 , with @xmath28 odd , are identically zero . under these considerations it is possible to exclude those inhibited states . then , one is left with a truncated basis of the form @xmath156 , where @xmath29 is now even . following the arguments above and using the results of the appendix one finally gets the following expression for the s - matrix in the case of equal masses @xmath264 with @xmath28 and @xmath32 even and the sum in @xmath70 restricted to the positive zeros of @xmath76 . the coefficients @xmath265 are equal to @xmath266 up to a factor @xmath267 ( see appendix ) . we remark the simplicity of the expression ( [ eq : derived_snm ] ) as compared to the exact solution of the problem . part of the payoff for this simplicity is the existence of two undetermined parameters , @xmath194 and @xmath184 . _ _ _ _ the optimal value of @xmath184 is nevertheless constrained by two considerations . on one side , it can not be too large , as at large distances the interaction between the two fragments would dominate over the interaction due to the target , and so ignoring the correlations between the particles for values just below @xmath184 would not be a good approximation . on other side , the value of @xmath184 should be large enough to allow us to ignore the interaction with the target for @xmath13 above @xmath184 . thus , the matching radius must be about the size of the system . moreover , @xmath194 must be of the order of the expectation value of @xmath23 on the ground state . in this section we analyse the reliability of usa by comparing its predictions with the exact results . in fig . [ fig : smat_usa ] the s - matrix coefficients for the usa model ( dot - dashed line ) with the phases @xmath268 given by ( [ eq : delta_ns ] ) are compared with the exact calculation ( solid line ) and the adiabatic approach ( dashed line ) . the calculations have been performed using the matching radius @xmath269 and an average potential @xmath270 . these values were determined by fitting the elastic s - matrix at high energies , where the model is expected to be more accurate . we notice that for these energy independent parameters a good description of the elastic and inelastic coefficients is achieved for energies above @xmath271 . by contrast , the adiabatic model does not seem to give a good description of the s - matrix at any scattering energy . a quantity closely related to the s - matrix coefficients is the _ average final excitation energy_. it gives an idea on the degree of excitation of the final system . it has been defined as @xmath272 we show the energy dependence of the average excitation energy , @xmath273 . as in the previous case , the usa model ( dot - dashed line ) agrees well with the exact calculation ( solid line ) for energies above @xmath274 . on the contrary , the usa model , as expected , does not describe properly the low energy regime . in particular , a spurious discontinuity is observed for the threshold at @xmath275 . in fig . [ fig:<n|wf>_e10 ] we compare the wave function given by the usa approach with the exact calculation at @xmath276 . the curve for the exact calculation has been obtained using expression ( [ eq:<n|phi > ] ) . in the usa model , represented by the dot - dashed line , one has to distinguish the asymptotic and the interaction regions . for the asymptotic wave function , the interaction with the wall is neglected ( @xmath277 ) and so the range of values of @xmath12 is unrestricted . then , the projection on a state @xmath28 is simply given by @xmath278 by contrast , in the interacting region , explicit account is taken for the wall and an expression similar to ( [ eq:<n|phi > ] ) should be used instead . thus , the total wave function in ii ( [ eq : wfii(r , r ) ] ) is projected on the different eigenstates , taking into account that the integration is restricted to the interval @xmath279 . at this collision energy , [ fig : smat_usa ] indicates that the scattering coefficients are well reproduced by ( [ eq : derived_snm ] ) with the phases @xmath280 . then , we have adopted this prescription ( with the same matching radius ) to describe the approximated wave function at this energy . as can be seen in fig . [ fig:<n|wf>_e10 ] , the agreement with the exact calculation is quite good for all the states . as a general rule , our calculations show a better agreement for the elastic and first excited states , and it tends to be worse for excited states of increasing energies . this is expected because within the usa model we restrict the basis to the set of open channels and so , the effect of this truncation becomes more evident as we explore excited states close to the cut off . we have explored in more detail the limits of the usa model as well as the validity of the prescription ( [ eq : delta_ns ] ) for the phases . in order to do that we start from the exact expression of the s - matrix for a fixed collision energy . making use of ( [ eq : derived_snm ] ) , we fit the phases @xmath237 and @xmath238 appearing in @xmath268 in order to reproduce the exact calculation . we find that the s - matrices can be accurately fitted for all the scattering energies ( even for very small values ) . although these phases , @xmath237 and @xmath238 , can not be written exactly as @xmath281 and @xmath282 respectively , one can define effective energy and channel dependent radii so that @xmath283 and @xmath284 . in fig . [ fig : rnrs_e9 - 10 ] we plot the values of the quantities @xmath285 versus @xmath28 and @xmath286 versus the absolute value of @xmath260 , i.e. , the internal momentum at which the state @xmath287 is peaked . the selected collision energies are @xmath288 and @xmath276 . in both cases we deal with a total of five ho states , namely , the ground state and the first four excited states with even parity . using the prescription ( [ eq : delta_ns ] ) both quantities are just @xmath184 for all values of @xmath28 or @xmath70 . note that the value of @xmath92 is rather constant and very close to the matching radius used in our calculations , i.e. , @xmath269 . this constant value has been also plotted in the figure for reference . the values of @xmath285 are also very close to @xmath289 for the lower values of @xmath28 , but they tend to deviate from our prescription for values of @xmath28 close to the threshold . it is remarkable that the values of @xmath285 and @xmath92 are mostly independent on the scattering energy and of the individual state considered . this indicates that for scattering energies large compared to @xmath149 , the usa works very well , and the matching radius can be taken as a constant , related to the size of the system , and independent on the energy or the internal state . for lower energies , the usa may still be used , but in this regime the radius @xmath184 depends on the energy and the state considered . we have formulated a one dimensional problem consisting on the scattering of two particles , interacting with a ho potential , that collide with an infinite potential . this problem contains some of the dynamical features of the scattering of composite systems in atomic , molecular and nuclear physics , which interact with a target through short range interactions . we have obtained the _ exact _ solution of the problem using two different procedures . the first one consists of imposing the adequate boundary conditions on the scattering wave function . the second procedure deals with a basis of configuration localized states ( cls ) , which are wave functions with a strong spatial localisation . both procedures converge , provided that a sufficiently large basis of states is used . the main characteristic of the exact solution is that , for large scattering energy , the elastic scattering dominates . in order to achieve convergence , the inclusion of closed channels , i.e. , states with internal energy larger than the scattering energy , is required . although these states do not contribute to the s - matrix they must be taken into account in the calculations to obtain accurate results . this fact indicates that for short - range interactions one should be very careful when truncating the basis of states used in continuum discretized calculations . we have compared our _ exact _ results with the adiabatic approach , that considers the relative coordinate frozen during the scattering process . the results disagree completely . this indicates that the adiabatic approximation could be inaccurate when the interactions of the fragments with the target have a very short range . these short range interactions could couple to highly excited internal states for which the adiabatic approximation is not valid . however , it should be reminded that our calculations make use of a sharp infinite wall . thus , they will be relevant when the range of the interaction is short compared not only with the size of the projectile but also with the wavelength describing the motion of the fragments with respect to the target . we have developed a model which describes the scattering of a composite object in terms of the scattering wave functions and the s - matrices of the fragments . the model , that we call uncorrelated scattering approximation ( usa ) , neglects the correlations between the fragments during the scattering . the application of the usa to our model problem gives an expression for the s - matrix of the composite system in terms of the product of the s - matrices of the fragments . similarly , the scattering wave function is given as a combination of the product of regular wave functions of the fragments . the particular superposition is determined by application of the asymptotic boundary conditions . the application of the usa to our model problem relies on the use of two parameters . the most important one is the distance @xmath184 at which the asymptotic and uncorrelated wave functions are matched . we have fixed this value to @xmath290 , in terms of the oscillator length , for all our calculations . it should be noticed that this parameter has a similar meaning to the matching radius that is used in r - matrix theory . the other parameter is the average potential @xmath194 that replaces the interaction between the fragments . we have set this value to @xmath291 . by considering the matching radius to be energy- and state - dependent , the exact s - matrix and scattering wave functions are accurately reproduced . moreover , for high scattering energies the elastic and inelastic s - matrices are well reproduced by taking a fixed value of the matching radius . in general , the agreement with the exact calculation is better for the observables associated with the ground and first excited states and they tend to be worse for states with excitation energies close to the total energy . this is attributed to the fact that for these states the relative velocity between the fragments is small and then the correlations are expected to be more important . the one - dimensional model presented in this work can be a useful test case to check the validity of different approaches used in the description of the scattering of composite systems . the present choice of a sharp infinite wall for the description of the interaction with the target , and harmonic oscillator for the interactions between the fragments has the advantage that sets @xmath1 as the unique scale for energies and @xmath0 as the unique scale for lengths . in this sense , our results , which are expressed in those units , are valid for any value of the mass or harmonic constant . however , the model could be done more realistic , and more complicated , by substituting the infinite wall for an exponential function , and substituting the harmonic oscillator by finite potentials . we are grateful to prof . r. c. johnson for his worthy comments on the adiabatic model . a.m.m . acknowledges a research grant from the fundacin cmara of the university of sevilla . this work has been partially supported by the spanish dgicyt , project pb98 - 1111 . in this work different sets of basis states have been used , based on harmonic oscillator ( ho ) wave functions . these obey the general form where @xmath25 is the oscillator length , @xmath293 a normalisation constant and @xmath27 the hermite polynomial of order @xmath28 . in all the calculations we restrict the infinite set of states to a finite family of @xmath29 states , denoted by @xmath294 , where @xmath295 . by diagonalizing the position operator in the truncated basis of ho wave functions , a new set of @xmath29 states is obtained which have the property of being localised in configuration space . they are called configuration localised states ( cls ) and are denoted by @xmath296 . these two families of states are related by means of the orthogonal transformation for ho wave functions , the state @xmath68 is localised around @xmath298 , where @xmath79 is the @xmath70-th zero of the hermite polynomial @xmath299 . in this case , the transformation coefficients are given by ( see @xcite ) in analogy with the cls , it is possible to define internal states localised in momentum space , known as momentum localized states ( mts ) . this is carried out by diagonalizing the momentum operator in the truncated ho basis . thus , starting with the basis of @xmath29 states , this procedure provides a new set of @xmath29 internal states , each one of them is peaked around a certain momentum . as in the case of the cls there is an orthogonal transformation relating both sets of states : where @xmath223 represents a mls . then @xmath302 is localised around @xmath303 where , due to the formal analogy of the ho wave functions in momentum and configuration space , @xmath304 are again the zeros of the hermite polynomial @xmath93 . this analogy provides also a simple relation between the transformation coefficients @xmath305 and @xmath306 . in this work , we take the internal wave function to be real in configuration space , and so they will be affected by a factor @xmath307 in momentum space . moreover , we take the cls to be real functions , so the coefficients @xmath306 are real numbers . therefore , although the formalisms presented in this work do not require the fragments to have equal masses , we have performed all the calculations for this particular situation . in this case , parity conservation implies that only positive parity states are suitable to be populated during the process . in fact , the coefficients @xmath166 derived from the exact calculations of sections ( [ sec : exact_direct ] ) and ( [ sec : exact_cls ] ) are found to be zero for odd values of @xmath28 . this is also satisfied by eq . ( [ eq : derived_snm ] ) , as can be easily verified . under these considerations one is allowed to exclude those inhibited states from the beginning . this permits to work with the truncated basis of states : \{@xmath309 , with @xmath29 even . this requires , however , some care in the evaluation of the localised states in the new basis . the formalism of cls can not be directly applied to this set , as many of its properties entails the sum over both even and odd states . the starting point for the construction of the cls formalism requires a set of functions of the form @xcite then , taking @xmath316 , @xmath317 and @xmath318 we can identify @xmath319 ( @xmath320 ) . the set of configuration localised states are now calculated for the new set of functions . it requires the calculation of the roots of the polynomial @xmath321 which , attending to ( [ eq : phi2n ] ) , are just the square of the zeros of @xmath76 . the new set of localised states are given by a linear combination of the states @xmath72 and so they are even functions with respect to the variable @xmath12 . therefore , we call them _ _ _ _ symmetric configuration localised states , scls . the transformation between the set of states ( [ eq : phi2n ] ) and the scls is then expressed as as an example , in fig . [ fig : cls ] the set of ho wave functions @xmath326 are plotted versus the adimensional variable @xmath327 ( upper figure ) . the corresponding symmetric localised states are also plotted ( lower figure ) and labelled with the index @xmath70 ( @xmath328 ) . notice that each one of these localised states is peaked around two symmetrical points , corresponding to symmetrical roots of the hermite polynomial @xmath329 . the state @xmath331 is localised around @xmath332 where @xmath304 are again the positive zeros of @xmath76 . in the case of the ho basis , the transformation coefficients are related to those in configuration space :

according to our current understanding @xcite , dark matter ( dm ) and dark energy ( de ) represent the two major components of the present universe . surprisingly , it is found that the dm and de energy densities , @xmath0 and @xmath1 , are today roughly the same ( differing only by a factor of two ) , while their ratio has been varying by several orders of magnitude in the past history of the universe . it seems quite natural , then , to explore the possibility of a dm de interaction which could account for this coincidence . this approach , however , is not free from problems if the de component is interpreted in terms of a dynamical quintessence @xcite scalar field . indeed , such a scalar is constrained to be extremely light in order to fit the data , giving rise to unwanted long - range forces which may represent a severe threat to the equivalence principle . in addition , couplings of the quintessence scalar with the gauge field strengths are potential sources of dangerous time variations of the fundamental constants . a possible way of coping with a very light scalar while avoiding these shortcomings , is to choose to work in the framework of a scalar - tensor gravity ( st ) theory @xcite , where by construction matter has a purely metric coupling with gravity . it has been shown @xcite that in this case , the scalar field can benefit from an attraction mechanism which , during the matter dominated era , makes st overlapping with standard general relativity ( gr ) . at the same time , st may possess a second `` attraction mechanism '' @xcite which will ensure the correct evolution of @xmath1 along a so - called ` tracking ' @xcite solution . while st can very closely reproduce standard gr at the present time , it may however lead to major differences in the past evolution of the universe , differences which may result in observable consequences for us today . for example , it has been shown @xcite that st theories may have a profund impact on nucleosynthesis . at the same time , a curious fact has recently come to attention : a non conventional dynamics of the quintessence scalar in the past history of the universe may remain ` hidden ' to the available cosmological observations , but manifest itself through the dm relic abundance @xcite . it is then worth studying if st may provide a viable quintessence candidate and at the same time have an impact on the dm relic abundance . we have considered this possibility and computed explicitly the differences from the standard case . when considering st theories , the departures from standard cosmology are mainly due to the different expansion rate ( @xmath2 ) which they determine . such deviation from the usual expansion rate of gr bears potentially relevant consequences in all those phenomena which are closely dependent on the timing in which they occur . the aim of the present work is to study the possible modifications of the expansion rate of the universe in st at the time of cold dark matter ( cdm ) freeze - out , focussing on weekly interacting massive particles ( wimps ) as the most natural candidates for cdm . as it is well known , their present relic density depends on the precise moment they decouple and , in turn , on the precise moment the wimp annihilation rate equals the expansion rate of the universe . we expect then that a variation of @xmath2 in the past may lead to measurable consequences on the wimp relic abundance . in order to assess the allowed departure of @xmath2 from its standard value at wimps freeze - out , we have to take into account the bounds imposed on st by phenomena at later epochs @xcite . we already mentioned the crucial test of nucleosynthesis : passing the bbn exam will imply rather severe restrictions on the coupling of the scalar field of st to matter and also on its initial conditions at temperatures much higher than the wimp freeze - out . coming to later epochs , we have to consider the photon decoupling and the consequent restrictions imposed by cmb data @xcite . interestingly enough , we will show that the bbn `` filter '' on st is so efficient that it drastically limits any visible effect on the cmb spectrum ( in particular , shifts in the peaks positions are forced to be smaller than the experimental error ) . more restrictive than the cmb probe turn out to be the gr tests , in particular after the tight bound on the post - newtonian parameter @xmath3 , recently provided by the cassini spacecraft @xcite . this constraint becomes quite relevant when combined with that on the scalar equation of state of @xmath4 coming from sne ia data @xcite . we will find that significant deviations from standard cosmology are possible only if the de equation of state differs appreciably from @xmath5 today . in other words , if de is a cosmological constant , it is unlikely that future cosmological observations will be able to discriminate between st and gr . the question we intend to explicitly tackle is the following : taking all the abovementioned restrictions ( bbn , cmb , gr tests ) into account , how much can the hubble parameter @xmath2 at the time of wimp freeze - out differ from its canonical value if st replaces gr ? in other words , how much is the wimp relic density allowed to vary , if we consider st instead of gr ? we find that in st theories the expansion rate of the universe at few gevs can profoudly differ from the usual value obtained in gr ( with variations up to five orders of magnitude ) and , yet , allow the correct light elements production at bbn . this situation is perfectly analogous to the ` kination ' effect studied in @xcite , where a modification of @xmath2 at wimp freeze - out was induced by a short period of dominance of the scalar kinetic energy , although with some deal of fine - tuning . in the case considered here , the effect of st on @xmath2 depends on the strength of the scalar - matter coupling , however no particular fine - tuning is needed to pass the severe nucleosynthesis test even when large modifications of @xmath2 at freeze - out occur . this means that the attraction of st towards gr proceeds very rapidly during the cooling of the universe from the few gevs of wimps freeze - out down to the mev range of nucleosynthesis . the overlap of st with gr can subsequently be very efficient leading to st scenarios which can hardly be disentangled from ordinary gr in present tests at the post - newtonian level . the fact that st strongly affects the number of cdm particles may turn out to be the major signature of these theories . > from the point of view of particle physics model building , these large variations in the wimps number density today is of utmost relevance . particles which were not considered suitable to play a significant role in cdm scenarios can be rescued because of their enhanced number density . on the other hand , particles ( or regions of the parameter space for certain wimps candidates ) , which in usual gr scenarios constitute promising cdm candidates , would be excluded because their boosted number would overclose the universe . these considerations become of particular interest if we focus on the case where the wimps correspond to the lightest supersymmetric particle . a complete analysis of the cosmologically excluded and cosmologically interesting regions of the susy parameter spaces in different susy contexts , when st is considered , is presently in progress @xcite . st theories represent a natural framework in which massless scalars may appear in the gravitational sector of the theory without being phenomenologically dangerous . in these theories a metric coupling of matter with the scalar field is assumed , thus ensuring the equivalence principle and the constancy of all non - gravitational coupling constants @xcite . moreover , as discussed in @xcite , a large class of these models exhibit an attractor mechanism towards gr , that is , the expansion of the universe during the matter dominated era tends to drive the scalar fields toward a state where the theory becomes indistinguishable from gr . st theories of gravity are defined by the action @xcite @xmath6 where & & s_g = d^4x . [ jordan ] the matter fields @xmath7 are coupled only to the metric tensor @xmath8 and not to @xmath9 , _ i.e. _ @xmath10 $ ] . @xmath11 is the ricci scalar constructed from the physical metric @xmath12 . each st model is identified by the two functions @xmath13 and @xmath14 . for instance , the well - known jordan - fierz - brans - dicke ( jfbd ) theory @xcite corresponds to @xmath15 ( constant ) and @xmath16 . the matter energy - momentum tensor is conserved , masses and non - gravitational couplings are time independent , and in a locally inertial frame non gravitational physics laws take their usual form . thus , the ` jordan ' frame variables @xmath12 and @xmath9 are also denoted as the ` physical ' ones in the literature . on the other hand , the equations of motion are rather cumbersome in this frame , as they mix spin-2 and spin-0 excitations . a more convenient formulation of the theory is obtained by defining two new gravitational field variables , @xmath17 and the dimensionless field @xmath18 , by means of the conformal transformation @xmath19 imposing the condition @xmath20 the gravitational action in the ` einstein frame ' reads @xmath21 , \ ] ] and matter couples to @xmath22 only through a purely metric coupling , @xmath23 \,\,\ , .\ ] ] in this frame masses and non - gravitational coupling constants are field - dependent , and the energy - momentum tensor of matter fields is not conserved separately , but only when summed with the scalar field one . on the other hand , the einstein frame planck mass @xmath24 is time - independent and the field equations have the simple form @xmath25 where @xmath26 \,,\ ] ] and @xmath27 is the matter energy - momentum tensor in the einstein frame . the relevant point about the scalar field equation in ( [ eomgen ] ) is that its source is given by the trace of the matter energy - momentum tensor , @xmath28 , which implies the ( weak ) equivalence principle . moreover , when @xmath29 the scalar field is decoupled from ordinary matter and the st theory is indistinguishable from ordinary gr . we next consider an homogeneous cosmological space - time @xmath30 where the matter energy - momentum tensor admits the perfect - fluid representation @xmath31 with @xmath32 . the friedmann - robertson - walker ( frw ) equations then take the form @xmath33 \\ \left ( \frac{\dot{a}}{a}\right ) ^{2}+\frac{k}{a^{2 } } & = & \frac{1}{3 \mp^2 } \left[\rho + \frac{\mp^2}{2}\dot\varphi^{2}+v(\varphi ) \right ] \\ \ddot{\varphi } + 3\frac{\dot{a}}{a}\dot\varphi & = & -\frac{1}{m_{\ast}^2}\left [ \frac{\alpha ( \varphi ) } { \sqrt{2}}(\rho-3p)+ \frac{\partial v}{\partial \varphi } \right ] \ , \end{aligned}\ ] ] with the bianchi identity @xmath34 the physical proper time , scale factor , energy , and pressure , are related to their einstein frame counterparts by the relations @xmath35 defining new dimensionless variables @xmath36 and setting @xmath37 ( flat space ) the field equation of motion takes the more convenient form @xmath38 \varphi ^{\prime}= \nonumber \\ & & \quad\quad\quad\quad\quad - \sqrt{2}\ ; \alpha ( \varphi ) \;(1 - 3 w)-2 \;\lambda \;\frac{d\log v(\varphi ) } { d\varphi } , \label{eom}\end{aligned}\ ] ] where primes denote derivation with respect to @xmath39 . this will be our master equation . the effect of the early presence of a scalar field on the physical processes will come through the jordan - frame hubble parameter @xmath40 : @xmath41 where @xmath42 is the einstein frame hubble parameter . in the flat space case ( @xmath37 ) , eq ( [ htilde ] ) finally gives : @xmath43 \,\,\,\ , . \label{htilde2}\ ] ] during radiation domination the scalar field energy density is suppressed , @xmath44 , and the first term in the rhs of eq . ( [ eom ] ) is proportional to & & 1 - 3w= = + & & = , [ rhs ] where the sum runs over all particles in thermal equilibrium , while @xmath45 is the contribution from the decoupled and pressureless matter abundance . during radiation domination , @xmath46 , where @xmath47 is the jordan - frame temperature , and the contribution from a single particle in equilibrium gives @xmath48\,,\ ] ] with @xmath49 , @xmath50 the number of degrees of freedom of a , @xmath51 the number of relativistic degrees of freedom and @xmath52 \equiv \int_0^\infty dx \frac{x^2}{\varepsilon_a [ \exp(\varepsilon_a ) \pm 1]}\ , , \label{ffm}\ ] ] where @xmath53 and the minus ( plus ) sign in the denominator of the integrand holds for bosons ( fermions ) . in fig . [ fm ] we plot @xmath54 $ ] . we see that it is different from zero only around @xmath55 , that is , for @xmath56 . for higher temperatures it is quadratically suppressed in @xmath57 , approaching the relativistic regime in which @xmath58 . for lower temperatures it is boltzmann- suppressed . then , as emphasized in @xcite , the field @xmath59 evolves even during radiation domination , receiving a ` kick ' each time a particle in equilibrium becomes non - relativistic . the second term in eq . ( [ rhs ] ) is suppressed as @xmath60 , so it becomes relevant only as equivalence is approached . in the following , we will consider the evolution of the field from the freeze out of the dm particles down to today , so we will have to take into account all particles of masses between the freeze - out and matter - radiation equivalence . = 3.2 in during matter domination @xmath61 and , as long as the field energy density is subdominant ( @xmath62 ) , the rhs of the equation of motion ( [ eom ] ) is given by @xmath63 and the field evolution depends on the form of the coupling function @xmath64 . as already mentioned , the jbd theory is given by a constant @xmath65 , and the value @xmath66 corresponds to gr . a very attractive class of models is that in which the function @xmath64 has a zero with a positive slope , since this point , corresponding to gr , is an attractive fixed point for the field equation of motion @xcite . it was emphasized in ref . @xcite ( see also @xcite ) that the fixed point starts to be effective around matter - radiation equivalence , and that it governs the field evolution until recent epochs , when the quintessence potential becomes dominant . if the latter has a run - away behavior , the same should be true for @xmath64 , so that the late - time behavior converges to gr . the evolution of the field during the last redshifts depends on the nature of de . we will consider two possibilities : a cosmological constant and a inverse - power law scalar potential for @xmath59 , which can be collectively represented by the potential @xmath67 @xmath68 corresponding to the cosmological constant . in general , a cosmological constant in the einstein frame does not correspond to a cosmological constant in the jordan frame , as one can read from eq . ( [ transf ] ) . however , present tests of gr ( see next section ) imply that at late times @xmath69 , so that the two frames are almost coincident and the expansion histories during the last few redshifts are practically indistinguishable . for the purpose of this paper , that is the analysis of the impact of de on st cosmology , the situation in which the quintessence field is different from @xmath59 and decoupled from it would be basically the same as that with a cosmological constant , since in both cases the second term in the rhs of eq . ( [ eom ] ) vanishes . the main feature of the potential in ( [ potential ] ) for @xmath70 is the existence of ` tracker ' solutions , which are attractors in field space @xcite . in the @xmath71 limit , the late - time behavior is completely determined by the two parameters @xmath72 and @xmath73 . a non - vanishing @xmath65 would modify the behavior of the field today , hopefully keeping the desirable property of insensitivity to the initial conditions . in this paper , we will consider the following choice for @xmath74 , a()=1 + b e^- , [ aphi ] corresponding to ( ) = - , [ alphaphi ] which has a run - away behavior with positive slope , as required by the discussion at the end of the previous subsection . the choice for the parameters @xmath75 and @xmath76 will be discussed in the following section . in fig . [ run ] we show the evolution of the background and field energy density . we see that field energy densities corresponding to different initial conditions converge to the same solution , driven by @xmath77 . notice also that the @xmath65-attractor becomes effective even before matter - radiation equivalence , due to the non - relativistic decoupling effect explained above . in fig . ( [ levels ] ) we show the region of parameter plane @xmath78-@xmath73 giving @xmath79 , where @xmath80 and @xmath81 . we see that in the st case , ( @xmath82 ) the region giving more negative values of the equation of state is somehow enlarged with respect to pure gr quintessence . however , in the observationally allowed region for @xmath83 the influence of the @xmath75 parameter is negligible . = 3.2 in = 3.2 in assuming @xmath84 in eq . ( [ htilde2 ] ) as we have checked numerically for the solutions relevant in this analysis the jordan frame expansion rate during nucleosynthesis may be approximated as ^2 a^2 ( ) . the above expression should be compared to the gr one , in which the planck mass at nucleosynthesis was the same as today . we obtain : .|_nuc . |_nuc = . the change in the expansion rate is completely analogous to that obtained by adding extra neutrinos to the standard gr case . using the bound @xmath85 ( which is more restrictive than those obtained for instance in @xcite ) , we get < 1.08 . [ bbn ] at the post - newtonian level , the deviations from gr may be parametrized in terms of an effective field - dependent newtonian constant ) as well . ] @xmath86 and two dimensionless parameters @xmath87 and @xmath88 which , in the present theories turn out to be @xcite @xmath89 where @xmath90 . a new constraint on the parameter @xmath87 has been obtained recently using radio links with the cassini spacecraft @xcite , _ pn-1 = ( 2.12.3 ) 10 ^ -5 . [ casbound ] present bounds on @xmath91 are @xmath92 and are less restrictive for our choice of @xmath77 , since @xmath93 . the bound from the cassini spacecraft turns out to be quite strong when used in connection with the one on the equation of state @xmath94 from sne ia . in fig . ( [ betaw ] ) we show the excluded region in the @xmath76-@xmath94 plane implied by eq . ( [ casbound ] ) . we see that an equation of state @xmath95 , as implied by sne ia data at @xmath96 c.l . @xcite , requires either a large value for @xmath76 , or a very small @xmath75 . since the last case corresponds to an expansion history of the universe practically indistinguishable from gr , any non - standard behavior induced by the st theories in the past should be likely accompanied by an equation of state different from @xmath5 today . if de is a pure cosmological constant , then the bound from cassini implies @xmath97 ( making @xmath98 practically indistinguishable from one at least since bbn on ) , or unnaturally large values of @xmath76 . = 3.2 in the impact of a cosmological constant or quintessence on the cmb power spectrum has been extensively analyzed in refs . @xcite . in the context of st theories , the problem has been studied in refs . the main change with respect to a theory for de based on gr is due to a difference in the expansion rate , which affects the angular scale of the anisotropies . the angle under which the first peak is seen goes as _ peak ~/l_peak ~v_s t_dec z_dec/ d(z_dec ) , where @xmath99 is the corresponding multipole , @xmath100 is the sound speed , @xmath101 and @xmath102 are the time and redshift of decoupling , and @xmath103 the distance to the last scattering surface . the latter is given by d(z_dec ) = _ 1/(z_dec+1)^1 , [ dlast ] and is thus dominated by the behavior of @xmath2 close to the upper limit of integration , where @xmath104 is smaller . for this reason , st theories passing the gr tests ( @xmath105 today , that is , @xmath106 ) imply a small deviation of the distance to the last scattering surface with respect to gr . on the other hand , the decoupling time might be significantly more perturbed . it is given by an expression analogous to eq . ( [ dlast ] ) with the upper ( lower ) limit of integration replaced by @xmath107 ( @xmath108 ) . as a result , since the universe expanded faster than in gr at early times , we expect @xmath101 to be smaller , and the peak to move towards higher multiples . we find , which is consistent with the numerical findings of ref . @xcite . due to the well known degeneracy of the cmb spectrum with respect to cosmological parameters , present measurements of the peak locations @xcite do not translate straightforwardly into a bound on @xmath109 . for instance , it is found that a shift in the peak multipole can be obtained also by varying the energy densities according to @xcite @xmath110 so that , in general , a full reanalysis of the cmb including the new parameter @xmath109 would be required . however , in the present case we find that , once the bbn bound on @xmath111 has been imposed , the resulting values for @xmath109 are so close to unity as to give shifts in the peak multiples smaller than the experimental error . thus , the cmb spectrum does not provide significant bounds to the present scenario . having in mind all the bounds discussed in the previous section , we can now go on to compute the cosmological evolution of the scalar field and its impact on the dm relic abundance . as a first step , we want to estimate if st can have a sizeable effect on the jordan - frame hubble parameter @xmath2 at the time of wimp decoupling , without violating any of the avaliable cosmological observations . we will consider the function @xmath74 of eq . ( [ aphi ] ) , imposing on the parameters @xmath75 and @xmath76 the phenomenological constraints already discussed . we will then compute the ratio @xmath112 at the decoupling time of a typical wimp of mass @xmath113 gev . in this way we will be able to get an estimate of the effect before going into further detail . the tightest bound is that coming from eq . ( [ bbn ] ) . it has an impact on both @xmath75 in eq . ( [ aphi ] ) and on the initial conditions of @xmath22 at temperatures higher than the wimp freeze - out . indeed , since on the tracker solution the scalar field is @xmath114 today , it should have been @xmath115 at nucleosynthesis , otherwise it would not have reached the attractor in time @xcite . this implies @xmath116 . = 3.2 in as already discussed , the equation for the dynamics of the scalar field @xmath59 is obtained by substituting the expression of eq . ( [ rhs ] ) in the rhs of eq . ( [ eom ] ) and choosing a coupling function @xmath64 as defined in eq . ( [ alphaphi ] ) . in the sum of eq . ( [ rhs ] ) only the terms corresponding to particles with @xmath117 have been considered , i.e. particles lighter than the critical temperature of the phase transition through which they acquire a mass ( see ref . in particular we have taken into account the top quark , the @xmath118 , the @xmath119 , the bottom quark , the tau quark , the charmed quark , the pions , the muon , the electron and a wimp particle of mass @xmath113 gev . numerical integration of the equation for @xmath59 has been carried out between approximately 500 gev and today . we have then computed , through eq.([htilde2 ] ) , the modified expansion rate in the st theory at a temperature @xmath120 gev corresponding to a typical time of wimp decoupling and compared it to the expansion rate of the standard case at the same temperature . in fig . [ contorno ] we plot the ratio @xmath112 at @xmath120 gev as a function of the initial value of @xmath98 and initial ratio of the scalar to background energy density @xmath121 . we have restricted the possible initial conditions to those regions of parameters values respecting the bbn bound of eq . ( [ bbn ] ) . we see that we have been able to produce an enhancement of the expansion rate up to @xmath122 at the time of wimp decoupling . it is then worth studying in more detail what happens to the wimp relic abundance . let us now consider the calculation of the relic abundance of a dm wimp with mass @xmath123 and annihilation cross - section @xmath124 . as already mentioned , laboratory clocks and rods measure the `` physical '' metric @xmath125 and so the standard laws of non - gravitational physics take their usual form in units of the interval @xmath126 . as outlined in ref.@xcite , the effect of the modified st gravity will enter the computation of particle physics processes ( like the wimp relic abundance ) through the `` physical '' expansion rate @xmath2 defined in eq . ( [ htilde ] ) . we have therefore to implement the standard boltzmann equation with the modified physical hubble parameter @xmath2 : @xmath127 where @xmath128 , @xmath129 is the entropy density and @xmath130 is the wimp density per comoving volume . a typical behaviour of the function @xmath74 defined in eq . ( [ aphi ] ) , calculated for parameters @xmath131 and @xmath132 . ] we have considered values of @xmath2 wich respect all the bounds discussed in section iv . specifically , we have considered the function @xmath74 as given in eq . ( [ aphi ] ) with parameters @xmath131 and @xmath132 . the function @xmath74 for this choice of parameters is plotted in fig . [ fig : aphi ] , which shows that @xmath74 is very large at large temperatures , and then , at a temperature @xmath133 , sharply drops to values close to 1 before nucleosynthesis sets in . a parametrization of the behaviour of @xmath74 for @xmath134 , that will be useful in the following discussion , is : @xmath135 where @xmath136 is the current temperature of the universe . we have numerically checked that , in the regime we are considering , a good approximation to the physical hubble parameter is given by : @xmath137 the solution of the boltzmann equation is therefore formally the same as in the standard case , with the noticeable difference that now the hubble parameter gets an additional temperature dependence , given by the function @xmath74 . this can be translated in a change in the effective number of degrees of freedom at temperature @xmath47 : @xmath138 an approximated solution of eq.([eq : boltzmann ] ) can be cast in a form analogous to the standard case : @xmath139 where @xmath140 and @xmath141 and @xmath142 are the wimp abundances per comovin volume today and at freeze out , respectively . numerical solution of the boltzmann equation eq . ( [ eq : boltzmann ] ) in a st cosmology for a toy model of a dm wimp of mass @xmath143 gev and constant annihilation cross - section @xmath144 gev@xmath145 . the temperature evolution of the wimp abundance @xmath146 clearly shows that freeze out is anticipated , since the expansion rate of the universe is largely enhanced by the presence of the scalar field @xmath59 . at a value @xmath147 a re annihilation phase occurs and @xmath146 drops to the present day value . ] the freeze out temperature is obtained by the following implicit equation : @xmath148 \label{eq : xf}\ ] ] where @xmath149 is the internal number of degrees of freedom of our wimp . clearly , when @xmath150 we recover the standard case . the relic abundance is then simply given by : @xmath151 where @xmath152 is the present entropy density and @xmath153 denotes the critical density . a numerical solution of the boltzmann equation eq . ( [ eq : boltzmann ] ) in a st cosmology with the function @xmath154 given in fig . [ fig : aphi ] is shown in fig . [ fig : abundance ] for a toy model of a dm wimp of mass @xmath143 gev and constant annihilation cross - section @xmath144 gev@xmath145 . the temperature evolution of the wimp abundance @xmath146 clearly shows that freeze out is anticipated , since the expansion rate of the universe is largely enhanced by the presence of the scalar field @xmath59 . this effect is expected . however , we note that a peculiar effect emerges : when the st theory approached gr ( a fact which is parametrized by @xmath155 at a temperature @xmath133 , which in our model is 0.1 gev ) , @xmath156 rapidly drops below the interaction rate @xmath157 establishing a short period during which the already frozen wimps are still abundant enough to start a sizeable re annihilation . this post - freeze out `` re annihilation phase '' has the effect of reducing the wimp abundance , which nevertheless remains much larger than in the standard case . for the specific case shown in fig . [ fig : abundance ] the wimp relic abundance is @xmath158 for gr , while for a st cosmology becomes @xmath159 , with an increase of a factor of 44 . the phenomenon of re annihilation can be conveniently discussed in terms of the relation between the expansion rate of the universe @xmath156 and the wimp interaction rate @xmath160 . a numerical calculation of these two quantities is plotted in fig . [ fig : rates ] as a function of the temperature . the departure from equilibrium occurs earlier than in the gr case , because @xmath161 . when decoupling is completed , the particles evolve with an approximately constant @xmath162 and @xmath163 , while the hubble rate evolves as @xmath164 , i.e. slower than in the standard case ( we have used here the approximate @xmath154 behavious of eq.([eq : aphi ] ) ) . at the transition temperature @xmath133 the hubble rate drops to its standard value @xmath165 and becomes smaller than the interaction rate : in this case the decoupled wimps start to annihilate again , for a short period . after this re annihilation phase , the particles continue to evolve with an approximately constant abundace @xmath166 and @xmath157 recovers the behaviour @xmath167 , while @xmath168 as usual . the expansion rate of the universe @xmath169 and the wimp interaction rate @xmath170 are plotted as a function of the temperature . the re - annihilation effect discussed in the text is outlined . the small drop in the rates at @xmath171 mev is due to the quark hadron phase transition . ] the ratio between the freeze out values of @xmath172 in st cosmology and in gr as a function of the wimp mass . the dashed , solid and dotted lines refer to @xmath173 gev@xmath145 , @xmath174 gev@xmath145 and @xmath175 gev@xmath145 , respectively . ] we notice that a re annihilation phase does not occur in the case of kination , i.e. in the case the energy density of the universe is dominated by the kinetic term of a scalar field @xcite . in this case the evolution of the expansion rate is @xmath176 during kination , and than evolves smootly into the standard behaviour @xmath177 . during kination both @xmath2 and @xmath157 have the same @xmath47dependence and closely follow each other , until kination ends and the standard behaviour is recovered . re annihilation is possibile if the phase during which the expansion rate has the transition toward its standard gr behaviour is faster than the post - freeze out evolution of the interaction rate , i.e. faster than @xmath167 . the change in the freeze out temperature is shown in fig . [ fig : freeze ] where we show the ratio between the freeze out values of @xmath172 in st cosmology and in gr . the freeze out temperature is anticipated about a factor of 2 , with a dependence also on the annihilation cross section , as is clear from eq . ( [ eq : xf ] ) : for very low values of @xmath178 the freeze - out temperature may be anticipated up to a factor of 5 . for these low cross sections the relic abundance is anyway largely overabundant : we can therefore quantify the reduction in @xmath179 in a factor which ranges between 10% and 40% for wimps which can provide abundances in the cosmologically acceptable range . increase in the wimp relic abundance in st cosmology with respect to the gr case . the solid curve refers to an annihilation cross section constant in temperature , i.e. @xmath180 gev@xmath145 , while the dashed line stands for an annihilation cross section which evolves with temperature as @xmath181 gev@xmath182 . ] the amount of increase in the relic abundance which is present in st cosmology is shown in fig . [ fig : ratio ] . the solid curve refers to an annihilation cross section constant in temperature , i.e. @xmath183 , while the dashed line stands for an annihilation cross section which evolves with temperature as : @xmath184 ( these two cases correspond to the two limiting situations of the usual non relativistic expansion of the thermally averaged annihilation cross section : @xmath185 ) . in the case of @xmath186wave annihilation the increase in relic abundance ranges from a factor of 10 up to a factor of 400 . for a pure @xmath187 dependence , the enhancement can be as large as 3 orders of magnitude . the behaviours shown in fig . [ fig : ratio ] , which have been obtained by a numerical integration of the boltzmann equation eq . ( [ eq : boltzmann ] ) , can be understood by employing the approximate analytical solution ( [ eq : y0 ] ) . in the case of @xmath183 , eq . ( [ eq : y0 ] ) gives : @xmath188 in the standard gr case , and @xmath189 \nonumber \\ & + & { \cal g}\,m \left[g(x_\varphi)\frac{a}{x_\varphi}\right ] \label{eq : yphi}\end{aligned}\ ] ] in our st model where the @xmath154 function is given in eq . ( [ eq : aphi ] ) for @xmath134 and @xmath190 otherwise ( @xmath191 ) . for the sake of simplicity , in both solutions we have dropped the term @xmath192 which adds a small correction , not relevant for the present approximate discussion . in both equations @xmath193 . the ratio @xmath194 of the relic abundances is : @xmath195 where we have approximated @xmath196 and we have defined @xmath197 . by making explicit the mass dependencies we obtain : @xmath198 where the mass is expresses in gev , @xmath199 , @xmath200 , and the numerical values have been obtained for @xmath201 and @xmath202 ( since in our case @xmath133 is smaller than the quark hadron phase transition which we have set at @xmath203 mev ) . the analytic approximation of eq . ( [ eq : ratio ] ) helps to explain the behaviour shown by the solid curve in fig . [ fig : ratio ] , which has been obtained by numerical calculations which employ the exact form of the function @xmath74 . from eq . ( [ eq : ratio ] ) we can in fact derive that , for low masses , the ratio @xmath194 has the behaviour : @xmath204 which shows that in this mass regime @xmath194 grows almost linearly with the wimp mass @xmath123 , and it is larger for lower values of @xmath133 . if we accept @xmath133 as low as the bbn scale , we can obtain a further increase in the relic abundance of a factor 100 on the top of the one showed in fig . [ fig : ratioa ] for low values of @xmath123 . when the wimp mass is very large , the ratio @xmath194 behaves as : @xmath205 with a slight drop with the mass . the position of the maximum and the maximal value of @xmath194 are given by : @xmath206 and : @xmath207 these expressions show that the maximal effect is also obtained for the lowest values of @xmath133 ; in this case the position of @xmath208 is shifted toward lower masses . for @xmath133 at the bbn scale , the maximal increase in the wimp relic abundance is of the order of 3000 , instead of about 400 obtained for @xmath209 gev and shown in fig . [ fig : ratioa ] . an interesting property shown by eq . ( [ eq : ratio ] ) is that @xmath194 does not depend explicitely on the annihilation cross section @xmath183 , which drops out in the ratio . an implicit dependence on the cross section is present through @xmath179 , as can be seen in eq . ( [ eq : xf ] ) . this dependence however is only logarithmic and does not spoil the general behaviour of @xmath194 shown in fig . [ fig : ratio ] . this is shown in fig . [ fig : ratioa ] : the largest difference occurs for very low annihilation cross sections , for which the deviation of @xmath179 is larger . however , for cross sections of interest , i.e. cross section which provide relic abundances below the cosmologically acceptable upper bound , the values of @xmath194 are stable to a relatively good extent . increase in the wimp relic abundance with an annihilation cross section @xmath183 , for different values of @xmath210 . the dot dashed , solid and dashed lines correspond to @xmath211 gev@xmath145 , @xmath174 gev@xmath145 and @xmath175 gev@xmath145 , respectively . ] increase in the wimp relic abundance with an annihilation cross section @xmath212 , for different values of @xmath213 . the dot dashed , solid and dashed lines correspond to @xmath214 gev@xmath145 @xmath174 gev@xmath145 and @xmath215 gev@xmath145 , respectively . ] the relic abundance in a st theory as a function of the wimp mass in the case of @xmath216 gev ( solid line ) and @xmath217 ( dashed line ) . the upper ( lower ) dotted lines corresponds to the gr case for @xmath218 gev and @xmath219 , respectively . ] a similar analysis holds in the case of @xmath220 . however , in this case the dependence of @xmath194 with @xmath179 is somehow stronger ( as obtained from the integration in eq . ( [ eq : y0 ] ) ) , and the effect of changing @xmath124 is slightely larger . this effect can be seen in fig . [ fig : ratiob ] , where @xmath194 is shown for different values of the parameter @xmath213 . notice that larger cross sections , which in the standard case provide lower values for the relic abundance , are the ones which get more enhanced in st cosmology . finally , as an example we show in fig . [ fig : omega ] the relic abundance as a function of the wimp mass in the case of @xmath221 gev and @xmath217 . we see that , in this case , the relic abundance can be at the level required to explain the cdm content of the universe ( @xmath222 @xcite ) for a st theory , while it is underabundant in the standard case . the models shown in fig . [ fig : omega ] represent a case in which we can explain at the same time both the dm and de contents of the universe , and the interplay of the two component is crucial in determinig the right abundances of both dm and de . an analysis of specific particle candidates of dm , in particular in supersymmetric models , will be examined elsewhere @xcite . the idea of exploiting primordial ( ultralight ) scalars in order to shed some light on a dynamical interpretation of de is by now a widespread research topic in the literature . in this paper we follow the promising proposal of considering the quintessence scalar as embedded in a scalar tensor theory of gravity . this approach is at variance with the usual interpretation of quintessence as a new light scalar whose interactions with matter are subject to the tight phenomenological constraints on the equivalence principle violation and time - variation of the fundamental coupling constants . identifying the quintessence field with the scalar component of a st theory , instead , does not pose any threat on the equivalence principle , since by construction matter has a purely metric coupling with gravity . we focus on quintessence st models which possess a double `` attraction mechanism '' , one to gr and the other ensuring @xmath1 to follow a tracking solution . these two simultaneous mechanisms act as a `` protection '' for the theory to prevent its fall into immediate troubles ( for instance , large departures from gr predictions ) . nevertheless , we still obtain important phnomenological signatures which might disentangle this theory from gr or other alternative proposals . the tests of our st scenario divide into two classes : deviations from gr and departures from standard cosmology , in particular concerning the expansion rate of the universe . the latter effect has a big impact on the most distant epoch of the universe for which we have `` direct '' information , i.e. nucleosynthesis . however , we pointed out that we can further extend the implications of a non - standard @xmath2 in the early universe to times prior to nucleosynthesis . sticking to the standard wimp picture of dm , one of the most relevant events before bbn is the wimp decoupling which is expected to have occurred at a temperature of a few gevs . our work shows that , despite the severe `` filters '' on st quintessence models which are provided by bbn , solar system tests of gravity and , to a lesser degree , by cmb , it is still possible to find remarkable enhancements on the expansion rate of the universe at wimp freeze - out , yielding to relic wimp abundances which can vary up to a few orders of magnitude with respect to the standard case . in this paper we pointed out some general features of the new `` wimp story '' around its decoupling temperature in the presence of st quintessence . in particular , we noticed that some unexpected effect can take place , such as a short phase of wimp `` re - annihilation '' when st approaches gr . needless to say , such potentially ( very ) large deviations entail new prospects on the wimp characterization both for the choice of the cdm candidates and for their direct and indirect detection probes . a thorough reconsideration of the `` traditional '' wimp identified with the lightest neutralino in susy extensions of the sm as well as the identification of other potentially viable cdm candidates in the st context is presently under way @xcite . we acknowledge research grants funded by the italian ministero dellistruzione , delluniversit e della ricerca ( miur ) and by the istituto nazionale di fisica nucleare ( infn ) within the _ astroparticle physics project_. fr is partially supported by the university of padova fund for young researchers , research project n. @xmath223 . s. m. carroll , phys . * 81 * ( 1998 ) 3067 . a. masiero , m. pietroni and f. rosati , phys . d * 61 * ( 2000 ) 023504 . t. damour , f. piazza and g. veneziano , phys . d * 66 * ( 2002 ) 046007 ; t. damour , f. piazza and g. veneziano , phys . * 89 * ( 2002 ) 081601 a. r. liddle and r. j. scherrer , phys . d * 59 * ( 1999 ) 023509 ; i. zlatev , l. m. wang and p. j. steinhardt , phys . rev . * 82 * ( 1999 ) 896 ; p. j. steinhardt , l. m. wang and i. zlatev , phys . rev . d * 59 * ( 1999 ) 123504 ; b. ratra and p. j. e. peebles , phys . rev . d * 37 * ( 1988 ) 3406 ; p. j. e. peebles and b. ratra , astrophys . j. * 325 * ( 1988 ) l17 . a. g. riess _ et al . _ [ supernova search team collaboration ] , astron . j. * 116 * , 1009 ( 1998 ) ; s. perlmutter _ et al . _ [ supernova cosmology project collaboration ] , astrophys . j. * 517 * , 565 ( 1999 ) ; j. l. tonry _ et al . _ , astrophys . j. * 594 * , 1 ( 2003 ) ; b. j. barris _ et al . _ , astrophys . j. * 602 * , 571 ( 2004 ) ; a. g. riess _ et al . _ , astrophys . j. * 607 * , 665 ( 2004 ) . t. damour , gr - qc/9606079 , lectures given at les houches 1992 , susy95 and corf 1995 . ; g. esposito - farese and d. polarski , phys . d * 63 * , 063504 ( 2001 ) ; b. boisseau , g. esposito - farese , d. polarski and a. a. starobinsky , phys . lett . * 85 * , 2236 ( 2000 ) . d. huterer and m. s. turner , phys . d * 64 * , 123527 ( 2001 ) ; s. hannestad and e. mortsell , phys . rev . d * 66 * , 063508 ( 2002 ) ; a. balbi , c. baccigalupi , f. perrotta , s. matarrese and n. vittorio , astrophys . j. * 588 * , l5 ( 2003 ) . x. l. chen and m. kamionkowski , phys . d * 60 * , 104036 ( 1999 ) ; f. perrotta , c. baccigalupi and s. matarrese , phys . d * 61 * , 023507 ( 2000 ) ; r. nagata , t. chiba and n. sugiyama , phys . d * 69 * , 083512 ( 2004 ) ; r. nagata , t. chiba and n. sugiyama , phys . d * 66 * , 103510 ( 2002 ) .

catastrophes as defined in the abstract are related to a class of phenomena sometimes called `` relaxation '' oscillations . the observation of two widely separated time scales makes relaxation oscillations a good a priori subject of theoretical investigation , because one may suspect that their formation results from the existence of a ( more or less hidden ) small parameter . this gives some hope of a `` general '' theory based upon the small size of the parameter . to take an example , some major earthquakes lasting a few tens of seconds occur in the same general area about every hundred years . they thus involve a ratio of ` typical ' times in the neighborhood of @xmath0 , which is a very small number . in the present study a laboratory experiment was devised which displays similar relaxation oscillations , but acting time - wise on a scale that is convenient for investigation . experimental observations are explained in light of a two - part _ dynamical bifurcation _ model of catastrophe . comprising the local form of a physical model shown to be valid when close to the catastrophe , the striking result is as follows . in response to an external source of noise the two models ( the local one and the physical one ) predict fluctuations with a correlation time that increases before the catastrophe , just as is observed experimentally . the important point is that this well known critical slowing down phenomenon occurs significantly _ before _ the transition and could be used to forewarn it . we study the creeping of a soft metal under constant stress . accurate time records show that this creeping actually displays the following time dependent noisy component . the wire typically lengthens slowly with background ( non - thermal ) noise , with sometimes a `` large '' sliding event , followed again by a noisy slow lengthening regime , etc . plastic deformation of solids is a complex phenomenon , not yet fully understood @xcite . it has long been observed to take place in a non - smooth manner . most studies have focused on the portevin - le chatelier effect observed under constant strain rate conditions , whereas the present experiment is concerned with creep at constant stress . in both cases large steps ( cyclic slips ) are embedded in a noisy background . there is a general agreement that this is due to the complex dynamics of networks of dislocations whose motion is a means for the stressed solid to flow . as reported in section [ sec : experiment ] below , careful observations of the creep in strained sn - pb quasi - eutectic material ( a solder wire ) show the following : _ i ) _ on average a sample under constant stress lengthens at constant rate . _ ii ) _ continuous monitoring shows time dependent fluctuations of this length superposed on its secular increase . _ iii ) _ from time to time the length jumps by steps . afterwards a noisy and steady ( on average ) length increase is recovered until the next jump , etc . explaining all this remains a challenge for the common models of creep . nevertheless it is of great interest because it can be seen as a laboratory model of other far less accessible phenomena like earthquakes , where on average there is also a continuous slow sliding with random microseismic noise , interrupted by large fast sliding steps characterizing major earthquakes . we recently introduced the idea @xcite that in such systems the slow to fast transition can be described by a saddle - node bifurcation in a dynamical system evolving slowly with time . we pointed out the interest in such a modelization , which allows one to theoretically predict the response of this dynamical system to a noise source . it was shown that the induced fluctuations drift toward low frequencies ( i.e. , toward large correlation times ) _ before _ the transition , and so could be used as a forewarning . here we show that the dynamical saddle - node bifurcation model is the reduced form of a set of equations previously used to describe the portevin - le chatelier effect in metals , or metallic alloys , i.e. , the ananthakrishna ( ak ) model @xcite . the saddle - node model describes fast transitions resulting from the intrinsic dynamics of the original system which can be described locally ( close to the step ) as a slowly rocking potential system . using the ak model as well as the saddle - node reduced model we show that , with an added external source of noise , the correlation time of the fluctuations increases before the transition , following the classical scenario of slowing down at bifurcation points . as was shown long ago by dorodnytsin @xcite , the same local dynamics describe the transition from slow manifold to fast transients in relaxation oscillations of dynamical models , like the van der pol equation in the strongly nonlinear limit . by looking at the fluctuations of the length of our samples , we found this behavior near the step - like transitions , with the characteristic drift to low frequencies before the transition . unlike recent publications that have presented the idea @xcite that precursors of earthquakes could be found in the mechanical response to external perturbations ( such as the increase of the fluctuations and their slowing down near transitions ) our idea goes further , by giving an order of magnitude of the precursor time the authors of @xcite do not introduce the effect of a given time dependence of the parameters and consider only systems with steady parameter values on both sides of the bifurcation . to quote reference @xcite `` the suggested approach to analytical study of any kind of catastrophes is based essentially on the solution of a stationary problem of the possibility and conditions of the unstable equilibrium state in the system in question '' . without taking into account explicitly the time dependence of the parameters sweeping the bifurcation set , it is impossible to get the time scale for predictions . as we show , this scale depends crucially on the rate of change of the parameters near the bifurcation , which may be estimated from the knowledge of the ratio of the two time scales ( fast and slow ones ) for saddle - node models . an important challenge lies in the difficulty of stating a suitable model for a given catastrophic event . actually there is more than one class of possible slow - to - fast transitions in dynamical systems . the dynamical saddle - node can be seen as belonging to the class of systems with an equilibrium point losing stability as a parameter changes . it is not a loss of stability , but rather a loss of existence of the equilibrium point , occurring at the folding point of the slow manifold . however slow - to - fast transition may happen without any folding of the slow manifold . as shown in @xcite exploring a very often used mathematical representation of stick - slip behavior , the dieterich - ruina equations , the slow - fast transition can originate from the finite time singularity of the slow dynamics itself . in that case , critical speed - up , or drift of fluctuations toward large frequencies , is found to replace the critical slowing - down effect . this nonlinear phenomenon is obviously outside the class of phenomena explainable by a stability analysis of equilibria of dynamical systems . we refer the interested reader to the paper on this subject @xcite . in the present paper we do not consider this case , because it is clearly not the one observed in the creeping experiments . note that at this time it is unknown if real earthquakes ( as well as other observed catastrophes ) belong to the saddle - node case with a slowing - down near the transition or to the finite time singularity case with speed - up expected . the creeping experiment is described in section [ sec : experiment ] together with the striking spectral observations . in the two sections following section [ sec : experiment ] , we present our theoretical approach . we introduce in section [ sec : dynsaddbif ] an abstract dynamical model showing fast jumps . in this model , one assumes that , as a dynamical system , the jump follows a `` saddle - node '' bifurcation . there a pair of fixed points , one locally stable , the other locally unstable , merge and disappear as a control parameter changes . afterwards , the system moves quickly to a new equilibrium state that is at finite distance ( in phase space ) from previous equilibria , whence the jump . after reviewing the saddle - node bifurcation in this light , we assume that the parameter changes with time , namely that the parameter defining the bifurcation is itself a slow function of time . when , by this change , the parameter crosses the bifurcation value , the dynamical system makes an abrupt transition and jumps `` generically '' from one equilibrium state to another . in section [ sec : ak ] we show that this `` abstract '' model is pertinent for describing the slow - fast transition in the relaxation regime of a set of equations derived by ananthakrishna for creeping in the relaxation regime . this set of equations describes the dynamics of populations of dislocations in the creeping solid . the model shows relaxation oscillations in which slow drift is interrupted by fast variations . near the slow - fast transition , we show that this model reduces to the generic equation quoted above , for a certain range of parameters . therefore it displays a typical critical slowing - down in its response to external noise , in agreement with the real data reported in section [ sec : experiment ] . of particular interest is the fact that , from the experimental data , one can predict in advance a `` large slip '' event . the event is preceded by a shift toward low frequencies in the random fluctuations of specimen length . choose a uniform``wire '' that creeps under small stress at room temperature , fabricated from a much studied material , sn - pb solder , alloyed to be nearly eutectic . to establish creep at virtually constant stress , let one end of the wire be fixed and at the other end establish a constant force of tension . by this means we found the length of the wire to increase on average at constant rate . by highly accurate monitoring of this length vs time we observed a small time dependent part with the following pattern : on a background of fluctuations , from time to time a large slip event was observed , after which the continuous lengthening with a small background noise was recovered . the observed noise is non - thermal , since thermal noise has a far too small amplitude to be relevant . it is assumed to result from rearrangement of defect structures in the poly - crystalline structure of the quasi - eutectic , involving dislocation dynamics . we assume that this noise originates from a source that is to first - order independent of the overall lengthening of the wire . this is equivalent to saying that it is due to ongoing micro - scale events triggered by the imposed stress , independent of the global creeping . therefore we analyzed the response to this noise source according to the ak equations subject to an external noise ( see below ) . the instrument used in these experiments , which is pictured in fig.[fig : solder ] , is an extensometer @xcite . young s modulus can be accurately measured with a wire specimen , by placing different size masses on the weight pan . the trace of the vertical wire holding up the boom / weight pan in fig.[fig : solder ] has been enforced ( colored in black ) to be visible in the image . though not presently used , the black clamp was for purpose of holding a power resistor that was employed to measure the specimen s thermal coefficient of expansion . to measure temperature changes , a solid state thermometer was placed down into the sample space , through the hole seen near the top knurled clamp . @xmath1 @xmath1 to calibrate the instrument a tungsten wire of diameter 0.1 mm was mounted in the extensometer , and signal output level changes were recorded as various gram - mass standards were placed on the weight pan . by using the known young s modulus for tungsten , the resulting measurements yielded a constant of 1.0 nm per analog to digital count , for the 24 bit adc employed , which is sold by symmetric research @xcite . this constant is applicable to the data presently reported . a different measurement technique yielded essentially the same calibration constant . a he - ne laser was used with a mirror , operating as an optical lever , to measure boom position change as different masses were placed on the pan . on the basis of two factors , ordinary solder was chosen for the present study . first of all , an unusual property of tin is well known , when an ingot of the metal is strained by large amounts . the sound which it then emits is well described by the german word zinngeschrei " , which translates `` tin cries '' . the initiative for this work was also influenced by the observation of unusual spectral features in the output from a novel seismograph @xcite . the unusual low - frequency motions of the earth s crust that were then observed to precede an earthquake @xcite are readily seen by the volksmeter . this is due to the instrument s use of a `` displacement '' sensor , rather than the `` velocity '' sensor used by conventional seismometers . it was therefore natural to consider an alloy of tin , with the expectation that its defect properties should be more like those of the earth than is possible for a pure metal . the tin alloy for our study was ordinary soft solder ( 60% sn/ 40% pb ) , used universally in the electronics industry . although the pure ` eutectic ' alloy is actually 63% sn/37% pb , we will nevertheless use this word to describe our specimen in the discussions that follow . the stress level due to the load placed on the pb - sn wire used in the present experiments was considerably smaller than the typically @xmath2 used in usual creep studies . the present load was due solely to the weight of aluminum comprising the boom plus empty weight pan of the extensometer . this stress value was estimated to be @xmath3 , based on considerations that include the distance of the wire s attachment point from the position of the ( fine ) roller bearing in the unpright housing , that supports the boom on its end opposite the pan . @xmath1 ( a ) . ( b ) filtered signal @xmath4 at the exit of the single pole high - pass filter ( [ eq : filt ] ) . , title="fig:",height=144 ] ( b ) . ( b ) filtered signal @xmath4 at the exit of the single pole high - pass filter ( [ eq : filt ] ) . , title="fig:",height=144 ] @xmath1 during a typical avalanche , the elongation of a @xmath5 long , @xmath6 diameter specimen , would be about @xmath7 . before the avalanche , typical lengthening velocity is @xmath8 and typical rms fluctuations observed in the wire length ( with secular term removed by high - pass filtering ) would be about @xmath9 . an example of creep record ( length growth versus time ) is shown in fig.[fig : creep - exp](a ) . the fluctuations of the raw signal are not visible because they are much smaller than the average length variations . to eliminate the average growth , we use a standard technique of filtering . at the exit of a low - pass single pole filter , the filtered signal @xmath4 is given by the recursive formula @xmath10 where @xmath11 stands for the raw creep signal . in other words the filtered signal is the convolution product of the raw signal derivative by an exponential response function @xmath12 , where @xmath13 is the corner frequency of the filter . in equation ( [ eq : filt ] ) the parameter @xmath14 is given by @xmath15 where @xmath16 is the sampling time of the record , equal to @xmath17 in the experiment . using a corner frequency @xmath18 , the filtered signal takes the form shown in fig.[fig : creep - exp](b ) . @xmath1 ( a ) before the second step ( the step starts at @xmath19 ) , for the two time intervals ( 1 - 2 ) indicated in figure ( [ fig : creep - exp])-a , at the abscissa @xmath20 and @xmath21 respectively , each for a time interval of @xmath22 counts . ( c - d ) cumulative spectra corresponding to figures ( a - b ) respectively ( e)width @xmath23 of the cumulative spectrum versus time ( at height equal to @xmath24 per cent of the maximum ) , the vertical arrows indicate the very beginning of the fast steps . , title="fig:",height=115 ] ( b ) before the second step ( the step starts at @xmath19 ) , for the two time intervals ( 1 - 2 ) indicated in figure ( [ fig : creep - exp])-a , at the abscissa @xmath20 and @xmath21 respectively , each for a time interval of @xmath22 counts . ( c - d ) cumulative spectra corresponding to figures ( a - b ) respectively ( e)width @xmath23 of the cumulative spectrum versus time ( at height equal to @xmath24 per cent of the maximum ) , the vertical arrows indicate the very beginning of the fast steps . , title="fig:",height=115 ] @xmath1 @xmath1 ( c ) before the second step ( the step starts at @xmath19 ) , for the two time intervals ( 1 - 2 ) indicated in figure ( [ fig : creep - exp])-a , at the abscissa @xmath20 and @xmath21 respectively , each for a time interval of @xmath22 counts . ( c - d ) cumulative spectra corresponding to figures ( a - b ) respectively ( e)width @xmath23 of the cumulative spectrum versus time ( at height equal to @xmath24 per cent of the maximum ) , the vertical arrows indicate the very beginning of the fast steps . , title="fig:",height=115 ] ( d ) before the second step ( the step starts at @xmath19 ) , for the two time intervals ( 1 - 2 ) indicated in figure ( [ fig : creep - exp])-a , at the abscissa @xmath20 and @xmath21 respectively , each for a time interval of @xmath22 counts . ( c - d ) cumulative spectra corresponding to figures ( a - b ) respectively ( e)width @xmath23 of the cumulative spectrum versus time ( at height equal to @xmath24 per cent of the maximum ) , the vertical arrows indicate the very beginning of the fast steps . , title="fig:",height=115 ] @xmath1 @xmath1 ( e ) before the second step ( the step starts at @xmath19 ) , for the two time intervals ( 1 - 2 ) indicated in figure ( [ fig : creep - exp])-a , at the abscissa @xmath20 and @xmath21 respectively , each for a time interval of @xmath22 counts . ( c - d ) cumulative spectra corresponding to figures ( a - b ) respectively ( e)width @xmath23 of the cumulative spectrum versus time ( at height equal to @xmath24 per cent of the maximum ) , the vertical arrows indicate the very beginning of the fast steps . , title="fig:",height=115 ] @xmath1 in order to analyze the spectral properties of the fluctuations , and see how they evolve with time , we must consider separate sequences of a given time duration . one significant difficulty in this analysis is the proper choice for the magnitude of this time duration . the window can not be too large , lest the spectrum becomes invariant with time . conversely , if the window is too narrow , there will be excessive noise in the spectral ( or correlation ) signal . we believe the time windows have been judiciously chosen ; so that there is a significant increase of the correlation time before a `` burst '' . such increase of the correlation time ( actually the decrease of the frequency width of the signal ) was found to happen before every sliding event , giving some hope that this critical slowing down " is pertinent for predicting the catastrophe " before it occurs . for the present work a time duration of order @xmath25 of the interval between adjacent bursts was chosen . the spectral density of a given sequence ( @xmath26 ) of the filtered signal @xmath27 changes significantly during the creep process . the striking effect is the shift of the spectral density toward low frequencies in the last stage of the slow regime , i.e. , just before the burst . examples of this phenomenon are presented in the figures ( [ fig : creep - exp2])(a - b ) , which show the spectra corresponding to the time intervals ( @xmath28 indicated in figure ( [ fig : creep - exp])-a . the two spectra clearly differ . the first spectral density ( a ) displays a large number of components ; whereas the second spectral density , corresponding to the time interval just before the burst ( figure b ) , is concentrated close to zero frequency . to quantify this effect , we calculate the cumulative spectrum @xmath29 which is a smooth curve whose asymptotic value ( for n = n ) gives the experimental variance of the fluctuations during the sequence considered @xmath30 @xmath1 ( a ) . the horizontal segments ending at each point indicates the time interval ( before this point ) over which the width is calculated.,title="fig:",height=115 ] ( b ) . the horizontal segments ending at each point indicates the time interval ( before this point ) over which the width is calculated.,title="fig:",height=115 ] @xmath1 the shift of the density spectrum toward low frequencies is visually more clear in its associated cumulative spectrum , as seen in figures ( c - d ) . we define the low frequency extension " @xmath31 of a sequence by the width of the cumulative spectrum corresponding to @xmath24 per cent of its maximum value , @xmath32 shown in figure 3(e ) is the evolution of the spectral width for the whole experimental record . a drop in @xmath31 is clearly seen to occur before each burst . the important point is that the decrease of the width before the burst is experimentally foreseeable because it occurs during a precursor " time which is noticeably larger than the duration of a sequence @xmath33 . this is illustrated in fig.[fig : creep - exp3](b ) which displays a zoom of the spectral width evolution before the burst . the two following sections are devoted to theoretical models , the first one shows the main features of what is measured in the experiment on creeping , namely the critical slowing down of the fluctuation spectrum occurring before the sliding event , the second one is a model of creeping found in the literature which consists in a set of nonlinear coupled ordinary differential equations , derived by ananthakrishna and collaborators @xcite . this section explains why a saddle - node bifurcation with a slow sweeping of the bifurcation parameter exhibits a slowing down in its response to a source of noise in a window of time extending well before the bifurcation itself . abstract model has no direct connection with the physical phenomenon of creeping . in the section afterwards , however , we explain that a model of creeping shows the same slowing down in a range of parameters , linked to a local saddle - node bifurcation . consider first the saddle - node bifurcation of a gradient flow " , that is a damped dynamical system such that a coordinate @xmath34 is a solution of the equation of motion of the form @xmath35 in this equation @xmath36 is a potential , and the dynamics tends to everywhere lower the value of @xmath36 . the catastrophe theory of thom and arnold @xcite studies how steady equilibria of equations like ( [ eq : grad ] ) change under smooth deformations of the potential @xmath36 . below we consider a different kind of question , namely what happens to the solution of equation ( [ eq : grad ] ) when the potential @xmath36 becomes itself a slowly varying function of time , and particularly when a pair of equilibrium points disappears by a saddle - node bifurcation . indeed this question of the sweeping across bifurcations has been already widely studied @xcite with various applications in mind ( by `` sweeping '' we mean crossing of a transition point with a time dependent parameter in the equation(s ) of motion ) . however , to the best of our knowledge the occurrence of an intermediate time scale in the case of slow sweeping has been overlooked , although we believe it to be crucial for a strategy of foretelling catastrophes in the real world . the equation ( [ eq : grad ] ) is too general to be very helpful . however , it may describe a saddle - node bifurcation where a stable equilibrium disappears , assuming that @xmath37 depends slowly on time in a prescribed way , to become a function @xmath38 . near the transition , one may use a mathematical picture which is correct for a short time around the transition if the potential @xmath38 is a smooth function ( see below for what happens afterwards ) . assume first that @xmath36 does not depend explicitly on time and takes the form @xmath39 with @xmath40 real constant ( for the moment ) . for @xmath40 negative @xmath36 has two real extrema ( _ i.e. _ the roots of @xmath41 ) , one @xmath42 is a stable equilibrium , the other , @xmath43 , is an unstable equilibrium . for @xmath44 the two equilibria merge and disappear for @xmath40 positive , see fig.[fig : pot](a ) . this is the saddle - node bifurcation . the shape of @xmath36 near @xmath45 and for @xmath40 small is universal : for a given smooth @xmath36 showing this saddle - node bifurcation , one can always rescale @xmath46 and the external parameter to obtain the " local problem in this form . the extension to a time dependent control parameter @xmath40 goes as follows . if @xmath40 is a smooth function of time , one can assume that @xmath47 crosses the critical value , _ i.e. _ zero in the present case , at time zero in such a way that @xmath48 with @xmath49 non zero constant , dots being for higher terms in the taylor expansion of @xmath47 . for @xmath50 and @xmath46 close to zero , after rescaling , one can represent the dynamical system ( [ eq : grad ] ) , close to the saddle - node bifurcation , by the `` universal '' parameterless equation @xmath51 @xmath1 ( a ) . ( b ) quartic potential , @xmath52,title="fig:",height=144 ] ( b ) . ( b ) quartic potential , @xmath52,title="fig:",height=144 ] @xmath1 outside of the neighborhood of @xmath53 , the solution of ( [ eq : grad ] ) depends on other parameters defining @xmath54 for finite values of @xmath46 , as explained below . although this is not obvious from its formulation , this model is also valid for the transition from slow to fast motion in the van der pol oscillator in the limit of large non linearities , as shown by dorodnitsyn @xcite . let us sketch the proof of this ( interesting ) point . actually we shall look at the formally more general situation of relaxation oscillations , namely at solutions of a set of coupled ode s ( ordinary differential equations ) with a large parameter in the form : @xmath55 and @xmath56 in this set of equations dots are for time derivatives and the functions @xmath57 and @xmath58 are smooth with values of order @xmath59 when their argument is also of order one . moreover , @xmath60 is a large parameter . the slow manifold is defined by the condition that , in the limit @xmath60 large , the function @xmath61 must be close to zero over at least part of the trajectory . the cartesian equation @xmath62 defines a curve in the plane @xmath63 which allows to find @xmath64 as a function of @xmath46 , at least locally . this defines the equation of motion ( along the slow manifold ) @xmath65 where @xmath66 is such that @xmath67 the slow trajectory so defined stops at `` folds '' where the function @xmath66 ceases to be well defined , namely for values of @xmath63 such that @xmath68 , @xmath62 , @xmath69 being not zero . this defines a discrete set of points . near those points , the equation of motion can be solved by taking @xmath70 and @xmath71 where @xmath72 are the cartesian coordinates of the point such that @xmath73 let us look at the solution of the coupled equations ( [ eq : relax1 ] ) , ( [ eq : relax2 ] ) near @xmath72 . the equation ( [ eq : relax2 ] ) is not singular at this point and so can be solved for small variation of @xmath64 and for @xmath50 small , like @xmath74 , @xmath75 being the ( arbitrary ) time where the trajectory is in @xmath76 and @xmath77 . consider now the first equation ( [ eq : relax1 ] ) , and expand its right - hand side for @xmath78 and @xmath79 small . because @xmath80 at @xmath72 , the first nontrivial term in the taylor expansion of @xmath81 near @xmath82 with respect to @xmath46 is @xmath83 . on the other hand the first term coming from the expansion with respect to @xmath79 is @xmath84 . therefore , for @xmath50 and @xmath78 small , the equation for @xmath85 derived from ( [ eq : relax1 ] ) reads : @xmath86 one can check that all terms not written explicitly there are effectively negligible compared to the ones kept . standard rescalings allows to transform the equation ( [ eq : relax3 ] ) into the `` universal '' equation ( [ eq : gradt ] ) , provided various constraints of sign are satisfied by the quantities @xmath87 , @xmath88 and @xmath89 all computed at @xmath72 . this is a short version of the classical calculation by dorodnycsin , showing that the `` universal '' equation ( [ eq : gradt ] ) is also relevant for the slow to fast transitions in relaxation oscillations . it is worth mentioning also that this derivation does not really assume that the overall motion is periodic , as it may be extended quite easily to a system of many more coupled ode s with only one fast variable . this shows that such large jumps are also possible in non periodic dynamics , because the dynamics on the slow manifold may be chaotic in between the jumps if this slow manifold has a number of dimensions sufficiently large . now we shall focus on the slowing down near the dynamical saddle - node bifurcation described by equation ( [ eq : gradt ] ) . we shall first give its explicit solution . we look for a solution of equation ( [ eq : gradt ] ) transiting from the `` stable '' fixed point at `` large '' negative times to the rolling down towards positive value of @xmath46 at positive times . this solution behaves like @xmath90 at large negative times . the equation ( [ eq : gradt ] ) is of the riccati type and can be integrated by introducing the function @xmath91 such that @xmath92 where @xmath93 and @xmath91 is a solution of airy s equation @xmath94 the solution of equation ( [ eq : gradt ] ) relevant with the given condition at @xmath95 is drawn on figure([fig : quartic]-a ) . in terms of the variable @xmath91 it is the airy function @xmath96 which writes @xmath97 yet we have only solved the transient problem near the saddle - node bifurcation . the transition ends - up when @xmath50 becomes equal to the first zero of the airy function @xmath96 , _ i.e. _ the smallest root of the equation @xmath98 , a pure number , about @xmath99 . it corresponds to a divergence of @xmath100 , which behaves as @xmath101 just before this transition , as derived by expanding @xmath102 close to @xmath103 . therefore the `` generic '' equation ( [ eq : gradt ] ) for the dynamical saddle - node bifurcation displays a finite time singularity . let us precise the following mathematical subtlety . this property of the _ local flow _ , which results from the folding of the slow manifold , differs qualitatively from the finite time singularity found in the dieterich - ruina equations @xcite , where it was a property of the flow _ reduced to the slow manifold _ which is everywhere convex , as discussed in the introduction . in the case of the dynamical saddle - node bifurcation the singularity requires one to consider both the dynamics on and off the slow manifold , and this happens because the geometry forbids the continuation of an exact trajectory on this folded slow manifold . actually the solution ( [ eq : singdn ] ) loses its physical meaning sometime before the singularity since the `` universal '' dynamical equation ( [ eq : gradt ] ) was derived under the assumption that @xmath46 remains close to zero . this local theory can not deal with finite variations away from the critical values , therefore we shall need to add finite amplitude effects to limit the growth of the instability after the transition , see the subsection [ sec : quarticpot ] . we shall study now two questions , first the response of this dynamical system to an external noise , then the dynamics of a system showing a saddle - node bifurcation of the type just studied and reaching a new stable fixed point after this bifurcation . we explore first the response of our system to a small external noise , and look for qualitative changes in this response which could be a signal that occurs before the transition . let us consider the equation ( [ eq : gradt ] ) with a small noise added , so that equation ( [ eq : gradt ] ) is replaced by @xmath104 where @xmath105 is a random function of time , and @xmath106 a small coefficient . in the limit @xmath106 small , one can solve equation ( [ eq : gradtn ] ) by expansion in powers of @xmath106 , @xmath107 where @xmath108 the linear response to the noise is @xmath109 because @xmath110 tends rapidly to zero as @xmath111 tends to minus infinity , one can take @xmath112 to get rid of the effect of the initial conditions . to make the developments above more concrete , let us take a delta - correlated ( or white ) noise , such that @xmath113 the pair correlation of @xmath114 is given by @xmath115 where @xmath116 is the smallest of the two real numbers @xmath50 and @xmath117 . the behavior of this pair correlation for large negative values of both @xmath50 and @xmath117 , is derived from the asymptotic expression of airy s function , @xmath118 . setting @xmath119 , and @xmath120 , the variance of the fluctuations writes @xmath121 in the limit @xmath122 the integral is concentrated near @xmath123 so that @xmath124 which shows that the amplitude of the fluctuations increases some time before the transition itself . as the transition approaches , the standard deviation of the fluctuations , @xmath125 increases close to the critical time @xmath103 , because @xmath126 . as the zeroth order solution diverges at @xmath127 , it does not make sense to describe the dynamical behavior of the fluctuations due to the external noise very close to @xmath103 , as shown above . as said before , this unbounded growth of the fluctuations is a consequence of the _ local _ cubic form of @xmath36 when expanded near @xmath45 , in obvious contradiction with the fact that @xmath34 tends to infinity . to suppress the divergence of @xmath34 after the saddle - node bifurcation we add a stabilizing ( positive ) term to the potential @xmath36 which becomes quartic , @xmath128 as drawn in fig.[fig : pot](b ) . because of the growth of @xmath129 at infinity , like @xmath130 , the solution of the differential equation @xmath131 does not diverge at finite time . notice that , formally this does not apply to cases where the local equation ( [ eq : gradt ] ) describes a slow - to - fast transition in a limit cycle , instead of a gradient flow dynamics . however , unless one insists to look at what happens well after the saddle - node bifurcation , the details of the dynamics after the fast slide are not significant , because one can consider that the point of landing on the slow manifold after the fast drift is like a new equilibrium for a gradient flow , this neglecting the slow motion on this manifold . the equation ( [ eq : dynquartic ] ) can be written in the given scaled form for any quartic potential provided the coefficient of @xmath130 is positive . for such a potential one parameter only remains . in equation ( [ eq : quartic ] ) , we choosed to keep as explicit coefficient the coefficient @xmath40 of the linear term in equation ( [ eq : quartic ] ) . for @xmath132 the dynamical system ( [ eq : dynquartic ] ) is exactly at the saddle - node bifurcation , because at @xmath133 both the first and second derivative of @xmath129 vanish , but not the third derivative . contrary to the case of the pure cubic potential , this system has always , that is for any value of @xmath40 , a stable fixed point beyond the pair of fixed points collapsing at the saddle - node bifurcation . this makes it a fair candidate for describing the dynamical saddle - node bifurcation without blow - up . as in the previous case , we shall take now a time dependent @xmath40 , that will be taken as @xmath134 with @xmath49 positive constant . because of the rescaling of the cubic and quartic term , the parameter @xmath49 can not be eliminated ( another possibility would be to put a parameter in front of the cubic term ) . for the potential @xmath135 we shall analyze the solution of the dynamical equation @xmath136 tending at large negative and positive times to the equilibrium point @xmath137 , @xmath50 being considered as a parameter , see fig.[fig : quartic](a ) . moreover we consider the limit @xmath49 small , @xmath49 being related to the ratio of small to large time scales , it can be estimated from experimental data . we shall prove that in this limit @xmath49 small , there are three characteristic time intervals , depending how @xmath46 is close to zero . the long time scale is the time lapse between successive major slips , typically of order @xmath138 in our experiments on creeping . in our model it is the time needed for the potential @xmath139 to change significantly , to move from a pair of fixed points to a saddle - node bifurcation . because time enters in @xmath139 through the combination @xmath140 , the a - dimensional time needed for a change of shape of @xmath141 is of order @xmath142 the short time @xmath143 is of order unity in units of our model equation ( [ eq : dynquarticscaling ] ) as shown in the next paragraph . it is the duration of the abrupt change of slope of the function of time @xmath34 , the rising of @xmath34 at @xmath103 . in the creep experiment reported above , it is about @xmath144 . therefore the ratio of these two time scales @xmath145 is as small as @xmath146 in our experiments . there is another time scale , @xmath147 , the time interval standing before the transition , and close to it , during which the potential is very flat , while the solution has not yet jumped . during this time , @xmath46 and @xmath148 are much smaller than unity , then the cubic term on the right - hand side of equation ( [ eq : dynquarticscaling ] ) is negligible . in this range one recovers the universal equation of the dynamical saddle - node bifurcation ( [ eq : gradt ] ) by taking @xmath149 and @xmath150 , with the boundary condition @xmath151 at @xmath152 tending to minus infinity . this property concerns the rectangular domain drawn on fig . [ fig : quartic]-(a ) , where @xmath46 is small , @xmath153 , and @xmath50 extends from @xmath154 to @xmath155 , located before the abrupt increase . therefore the time extension of this domain introduces the intermediate time scale , @xmath156 long compared to the short time @xmath143 ( which is of order unity , see below ) and small compared to @xmath157 the average time between slips . we now show that the short time is of order unity , by matching the solution @xmath158 of the universal equation to the solution of equation ( [ eq : dynquarticlate ] ) below , in the vicinity of the critical point @xmath159 . because @xmath158 behaves like@xmath160 before it diverges , it follows that the solution @xmath34 behaves as @xmath161 for `` large '' values of @xmath162 before the critical time . this becomes of order one when @xmath163 becomes also of order one . when this happens , the term @xmath148 in equation ( [ eq : dynquarticscaling ] ) is negligible , therefore the solution of this equation , which can be matched with the solution near the bifurcation , is the solution of the integrable equation @xmath164 with the asymptotic behavior for very large negative times @xmath165 . this equation shows that , in our model , the fast time scale is of order one , because it has no explicit dependence with respect to the small parameter @xmath49 . this result is confirmed by the numerics , see fig.[fig : quartic](a ) . more precisely defining the rising time of @xmath34 by @xmath166 of the width of the time derivative @xmath167 , we get @xmath168 independently of the value of @xmath49 , for @xmath49 small . in the experiment , the intermediate time scale is @xmath169 , which gives about @xmath170 for @xmath171 , @xmath172 . in summary , by matching the two solutions in the range @xmath173 , we show that the catastrophe takes place during this time @xmath143 which is of order one , because the displacement is then of order one , compared to the small displacement of order @xmath174 taking place during time @xmath175 typical of the `` universal '' transition process . from this understanding of the various scales in the deterministic part of the dynamical equations , we can now look at the response to noise of this system , particularly at the range of time where something like a `` critical slowing down '' could be observed , and which actually happens in our experiments . with a noise source added , the dynamical equation ( [ eq : grad ] ) becomes , @xmath176 actually the effective noise amplitude is not equal to @xmath106 close to the saddle - node , as it depends on the value of the parameter @xmath49 . indeed for @xmath177 , the cubic term in equation ( [ eq : xxquartic ] ) is negligible , and the equation reduces to @xmath178 which may be written as @xmath179 , by setting @xmath180 , @xmath181 , and @xmath182 . therefore the effective noise is larger than @xmath106 , by a factor @xmath183 in the rectangular domain of fig.[fig : quartic ] . @xmath1 ( a ) ) , with and without noise , @xmath184 ( smooth curve ) and large amplitude noise @xmath185 . ( b ) solution of equation ( [ eq : dynquarticscaling ] ) for @xmath171 . the rectangle around the origin defines the region @xmath186 and @xmath187 , with @xmath188 . the critical time is @xmath189 @xmath190 . the two vertical lines inserted between the two arrows delimitate the large slope time duration , of order unity . ( c ) solution of equation ( [ eq : xxquartic ] ) with large amplitude noise @xmath185,title="fig:",height=120 ] ( b ) ) , with and without noise , @xmath184 ( smooth curve ) and large amplitude noise @xmath185 . ( b ) solution of equation ( [ eq : dynquarticscaling ] ) for @xmath171 . the rectangle around the origin defines the region @xmath186 and @xmath187 , with @xmath188 . the critical time is @xmath189 @xmath190 . the two vertical lines inserted between the two arrows delimitate the large slope time duration , of order unity . ( c ) solution of equation ( [ eq : xxquartic ] ) with large amplitude noise @xmath185,title="fig:",height=144 ] ( c ) ) , with and without noise , @xmath184 ( smooth curve ) and large amplitude noise @xmath185 . ( b ) solution of equation ( [ eq : dynquarticscaling ] ) for @xmath171 . the rectangle around the origin defines the region @xmath186 and @xmath187 , with @xmath188 . the critical time is @xmath189 @xmath190 . the two vertical lines inserted between the two arrows delimitate the large slope time duration , of order unity . ( c ) solution of equation ( [ eq : xxquartic ] ) with large amplitude noise @xmath185,title="fig:",height=120]@xmath1 . [ fig : quartic ] let us consider now the fluctuations of the solution @xmath34 of equation ( [ eq : xxquartic ] ) . for a small noise source , the solution may be expanded in power of @xmath106 as above . at first order it gives @xmath191 x_1(t ) + \xi(t ) \mathrm { , } \label{eq : x1}\ ] ] whose solution is formally @xmath192 \mathrm{. } \label{eq : x1i}\ ] ] in general @xmath193 is the time integral of the second derivative of the potential @xmath194 , which yields with our choice of @xmath54 : @xmath195 \mathrm{.}\ ] ] the standard deviation @xmath196 has to be calculated numerically . we expect it to display the same behavior as for the cubic case in the whole domain where @xmath197 , i.e. a little before the transition and close to it , because the potential is cubic in this range . after the transition , we expect that the fluctuation decreases , because the solution without noise becomes quasi - steady . this is confirmed by the numerics : the amplitude of the fluctuations strongly increases close to the critical time @xmath103 , its maximum occurring at time @xmath103 , then it decreases . more precisely the standard deviation behaves exactly as @xmath198 , the red curve in fig.([fig : width](a ) , for small noise . therefore the strong increase of the variance of the signal fluctuations can not be used as a precursor for predicting the transition because it occurs simultaneously with the signal itself close to the transition . note that this observation seems to contradict the currently found statement that fluctuation enhancement precedes the transition and can be used as a precursor . in the case of the saddle - node bifurcation model , we have indeed observed a `` precursor '' growth of the fluctuations , but only in the case of `` large '' amplitude noise , see fig.[fig : quartic](b ) . let us focus on the case of small noise . in this case we show below that the correlation time of the fluctuations changes much earlier than the onset of amplitude growth . this slowing - down can be used to foretell the event itself , in considerable advance of its occurrence . it is seen to consistently happen in the creeping experiment described below . @xmath1 ( a ) for the saddle - node model ( [ eq : xxquartic ] ) with @xmath171 . the correlations functions are defined by equation ( [ eq : gamma ] ) and ( [ eq : gammap ] ) for curves ( a ) and ( b ) respectively ; ( c ) spectral width ( a. u. ) calculated with @xmath199 , for @xmath171 . ( d ) @xmath200 for @xmath171 ( blue),for @xmath201 ( red ) . , title="fig:",height=153 ] ( b ) for the saddle - node model ( [ eq : xxquartic ] ) with @xmath171 . the correlations functions are defined by equation ( [ eq : gamma ] ) and ( [ eq : gammap ] ) for curves ( a ) and ( b ) respectively ; ( c ) spectral width ( a. u. ) calculated with @xmath199 , for @xmath171 . ( d ) @xmath200 for @xmath171 ( blue),for @xmath201 ( red ) . , title="fig:",height=153]@xmath1 @xmath1 ( c ) for the saddle - node model ( [ eq : xxquartic ] ) with @xmath171 . the correlations functions are defined by equation ( [ eq : gamma ] ) and ( [ eq : gammap ] ) for curves ( a ) and ( b ) respectively ; ( c ) spectral width ( a. u. ) calculated with @xmath199 , for @xmath171 . ( d ) @xmath200 for @xmath171 ( blue),for @xmath201 ( red ) . , title="fig:",height=168 ] ( d ) for the saddle - node model ( [ eq : xxquartic ] ) with @xmath171 . the correlations functions are defined by equation ( [ eq : gamma ] ) and ( [ eq : gammap ] ) for curves ( a ) and ( b ) respectively ; ( c ) spectral width ( a. u. ) calculated with @xmath199 , for @xmath171 . ( d ) @xmath200 for @xmath171 ( blue),for @xmath201 ( red ) . , title="fig:",height=144 ] @xmath1 consider the case of small effective noise , where the correlation function and the spectrum of the fluctuations @xmath202 are well described by the correlation function and spectrum of @xmath114 , respectively . as noted in the previous section , the calculation of these functions requires some care ; because the system is not in a statistically steady state . therefore the spectral density of the fluctuations depends on time and the correlation function @xmath203 depends on both @xmath50 and on the time difference @xmath204 . more precisely , the correct definition of the correlation function is actually given by @xmath205 the latter definition of correlation is not readily accessible in experimental situations , because it requires knowledge of the time dependent variance , which is difficult to estimate from a single sample . a more accessible tool is often used ; it is given by the expression @xmath206 which coincides with the correct expression if the variance is the same at time @xmath50 and @xmath207 only . if the variance changes noticeably during the time interval of duration @xmath204 , the latter expression is biased , since @xmath208 . we expect such a discrepancy to manifest itself close to the catastrophe , since the variance increases there by a large amount . to illustrate this point we show in fig.[fig : width](a - b ) the correlation functions of the fluctuation @xmath209 for the solution of the saddle - node model , as defined by equations ( [ eq : gamma ] ) and ( [ eq : gammap ] ) respectively . the strong increase of the correlation time before the catastrophe is noticeably truncated when using the biased expression ( [ eq : gammap ] ) . using the correct definition of the correlation function ( [ eq : gamma ] ) , the correlation time increases by a factor of ten over a time interval of order @xmath147 before the catastrophe , while the enhancement is only about @xmath210 when using the biased expression ( [ eq : gammap ] ) . moreover the enhancement is followed by a drop in the latter case . nevertheless both curves display well the critical slowing down effect , which manifests as a growth of the correlation time close to the transition . the increase of the correlation length _ before _ the catastrophe is understood by looking at the formal expression ( [ eq : x1i ] ) . the second derivative of the potential vanishes at @xmath211 , and that leads to the flatness of @xmath193 in the time domain @xmath212 , as shown in fig.[fig : width](d ) . this time domain could therefore be identified as a `` precursor time '' , of order few @xmath147 . now let us consider the spectrum of the fluctuations in order to compare with the experimental results . a time dependent spectrum can be defined formally by the ( real ) wigner transform @xmath213 that has to be modified for numerical applications , either by using a filtering procedure like the one used in the previous section or by introducing a slipping window . in this section we use the latter process . choosing a gaussian window function of width @xmath214 , the numerical spectrum is given by @xmath215 where @xmath216 are the numerical integration time boundaries . this expression does not take into account the variability of the variance ; therefore we expect the observed change in the spectral properties to be biased , as was the case for the correlation function @xmath217 . the evolution of the spectral width @xmath218 ( half - height width ) is reported in fig.[fig : width](c ) in a range of a few times @xmath147 around the transition . the solution @xmath219 is hightlighted by the solid red line . the ( c ) part of the figure illustrates the same effect as the correlation function @xmath217 , but in the fourier space . the spectral width decreases noticeably from negative time of order few @xmath147 , until the time @xmath220 , where it grows , because of the bias due to the variance increase close to the burst . the decrease of @xmath218 corresponds to a shift of the spectrum towards low frequencies . the important result is that this shift occurs well _ before the transition_. it occurs over a time interval of order few @xmath147 . this result agrees with our experiment where @xmath147 was estimated to be about @xmath221 and the decrease of the spectral width keeps on for about @xmath222 ( see fig.[fig : creep - exp2 ] where @xmath223 corresponds to @xmath224 counts in the abscissa scale ) . the growth of the fluctuations and their shift to lower frequencies can be understood as follows . as the transition approaches , the potential @xmath38 becomes flatter and flatter , making weaker and weaker the restoring force toward equilibrium . therefore , at constant noise source , the amplitude of the fluctuations driven by this noise source will grow because the damping is ever less efficient . moreover , the typical time scale for this damping will get increasingly larger because of the decreasing stiffness of the potential , thus favoring noise at lower and lower frequencies . this section is devoted to an analysis of solutions of a set of equations devised for describing creeping in solids . more precisely our purpose is to show that , in a range of parameters this equation exhibits a dynamical saddle node bifurcation . as in the abstract " model of the previous section , this bifurcation is also preceded by a slowing - down of the fluctuations triggered by an external source of noise . creeping phenomena in real materials are complex and difficult to predict quantitatively , despite decades of efforts on theoretical models . we have chosen to consider a model developed by ananthakrishna for creeping in strained solids . to make things simpler , we have only used its version without space dependence in the quantities involved . note that the introduction of space variables would lead to an aperiodic creep signal more realistic than the periodic signals of the present model , however we conjecture that it should not affect the main result of our study ( the emphasis of a precursor signal over a given time interval ) . the ak model considered here is a set of three coupled non linear ordinary equations with three dimensionless parameters ( @xmath49 , @xmath40 and @xmath225 , where the letters @xmath49 and @xmath40 have no connection with the same symbols used previously ) . the unknown time dependent quantities are three scaled variables corresponding to three density types of dislocations , @xmath34 , @xmath91 and @xmath226 , representing respectively mobile , immobile , and those with clouds of solute atoms that mimic cottrell s idea . the model equations write , @xmath227 @xmath228 @xmath229 where the variable @xmath34 stands for the elongation rate . as one can check , if the three variables are positive at time zero , they remain so at later times if @xmath49 , @xmath40 and @xmath225 are positive , as assumed . the relative elongation ( or strain , or creep ) @xmath230 of the solder wire is the time integral of @xmath34 , @xmath231 . solutions of equations ( [ eq : x1])-([eq : z1 ] ) have been extensively studied @xcite , their shape and duration versus the parameter values are given in @xcite . recall that relaxation oscillations are depicted for small values of the parameter @xmath40 only . for @xmath225 larger than a certain critical value @xmath232 , depending on @xmath233 and not written here , the solution is stable . the fixed point coordinates are given by the expressions @xmath234 at lowest order for the small parameter @xmath40 . this fixed point becomes unstable by a poincar - andronov bifurcation for small values of @xmath40 , only under the condition that @xmath225 becomes smaller than @xmath232 . in a large range of parameter values the limit cycle associated with the variable @xmath34 displays relaxation oscillations characterized by slow steps and fast bursts . in this range of @xmath40 and @xmath225 small , we focus on the case @xmath235 , for two reasons . first because it allows the set of equations to be reduced to the generic saddle - node equation introduced in the previous section . secondly because it leads to @xmath34 solutions looking approximately like our experimental data ( the strain rate @xmath34 has to be compared with the data files labeled @xmath4 in section [ sec : experiment ] ) . an example of the numerical solution of equations ( [ eq : x1])-([eq : z1 ] ) is given in fig.[fig:2droite ] . in this case the burst amplitude drawn in figure ( b ) ) is noticeably larger than the experimental one shown in fig.[fig : creep - exp]-(a ) , nevertheless we have chosen the parameter values for pedagogical reasons , in order to give a clear representation of the full cycle of the relaxation oscillations . consider the evolution of the elongation rate @xmath34 . the limit cycle of duration @xmath236 displays four stages . in the first step , @xmath237 , the flow evolves slowly and @xmath34 takes values of order unity . this stage is followed by a fast jump of @xmath34 at @xmath238 , then by a short slow stage @xmath239 with high values of @xmath34 , of order @xmath240 , finally followed by a fast decrease of @xmath34 at @xmath241 . @xmath1 ( a))-([eq : z1 ] ) for @xmath242 , @xmath243 , and @xmath244 . the three curves are @xmath245 ( blue ) , @xmath91 ( purple ) and @xmath246 ( yellow ) , ( b ) creep signal @xmath230 . , title="fig:",height=144 ] ( b))-([eq : z1 ] ) for @xmath242 , @xmath243 , and @xmath244 . the three curves are @xmath245 ( blue ) , @xmath91 ( purple ) and @xmath246 ( yellow ) , ( b ) creep signal @xmath230 . , title="fig:",height=144]@xmath1 we first consider the linear stability of the 3d flow , especially along the first stage described just above , which precedes the fast jump . in the next subsection we derive the local form of the model close to the burst . we shall prove that both approaches provide the same information , namely an estimation of the precursor time value . in the slow regime , assuming a 3d flow of the form @xmath247 ( and similar expressions for @xmath248 ) close to the trajectory @xmath249 , the exponents @xmath250 are the eigenvalues of the jacobian matrix @xmath251 or solutions of the equation @xmath252 with @xmath253 , @xmath254 , and @xmath255 . @xmath1 ( a ) ( b)@xmath1 one of the eigenvalues is real and negative all along the trajectory , while the other two become complex conjugates in the slow regime , their real part crossing zero before the burst at time @xmath256 , see fig.[fig : lyap ] . the time interval @xmath257 depends on the values of the parameters @xmath258 . it is generally a small fraction of the limit cycle period . below it is shown that the numerical value of @xmath257 is nearly equal to the intermediate time scale @xmath147 for the model . we now consider the behavior of the 3d flow in the vicinity of the burst , and derive the normal form of the ak model close to @xmath259 . we focus on the first part of the limit cycle , preceding the burst of @xmath34 , and try to understand how the trajectory leaves the slow manifold ( sm ) that we define now . because @xmath260 , the slow stages are described by canceling the right - hand side of equations ( [ eq : x1])-([eq : y1 ] ) , that reduces the 3d flow to the 1d flow @xmath261 from the first equation the variables @xmath34 is an explicit function of @xmath91 , @xmath262 inserting this expression into the second equation ( [ eq : sm ] ) , allows to also express the variable @xmath226 in terms of @xmath91 , @xmath263 in the phase space @xmath264 expression ( [ eq : zs ] ) defines the slow manifold , which is illustrated in fig . ( [ fig : nullcline ] ) , red curve . on the positive slope parts of the slow manifold the 1d flow obeys the differential equation @xmath265 , which becomes singular at critical points defined ( on the sm ) by the relation @xmath266 the 3d flow ( closed blue curve ) follows the path @xmath267 . at the critical point @xmath259 , the trajectory leaves the slow manifold , jumps to point @xmath268 ( fast stage @xmath269 ) , then it follows the portion ( @xmath270 ) , and finally returns to the sm in @xmath271 . we consider below the exit of the sm close to the critical point @xmath259 . ( red curve ) for @xmath242 , @xmath272 , and parametric plot of the 3d flow ( blue curve , numerical solution @xmath273 of equations ( [ eq : x1])-([eq : z1 ] ) for @xmath244 . , height=144 ] the coordinates of the critical point @xmath259 in the phase space @xmath274 are solutions of equations ( [ eq : xs ] ) and ( [ eq : df ] ) . from equation ( [ eq : xs ] ) we derive an expression for @xmath275 @xmath276 which should be identical to the expression of @xmath277 taken from the relation ( [ eq : df ] ) , @xmath278 these two expressions are identical if @xmath279 is a root of the cubic polynomial equation @xmath280 where @xmath281 . the positive solution @xmath279 and the corresponding value of @xmath282 and @xmath283 satisfying equations ( [ eq : xs])-([eq : zs ] ) respectively are given at leading order with respect to the small parameter @xmath284 by @xmath285 valid in the parameter range @xmath286 . we next derive the normal form describing the dynamical behavior of the solution close to the critical point . assuming that the variable @xmath46 follows adiabatically the variable @xmath64 according to equation ( [ eq : xs ] ) , the original system reduces to @xmath287 defining the dynamics for the portion of the trajectory near @xmath259 , including the burst @xmath269 . this is confirmed by the numerics : starting from the point b , with numerical initial conditions ( [ eq : turn ] ) , the behavior of the solutions of equation ( [ eq:2eqb ] ) agree well with the 3d flow , see fig . [ fig : closet ] . @xmath1 ( a ) ) with relations ( [ eq : xs])-([eq : zs ] ) , ( b ) solution of the 3d flow equations ( [ eq : x1])-([eq : z1 ] ) . , title="fig:",height=120 ] ( b ) ) with relations ( [ eq : xs])-([eq : zs ] ) , ( b ) solution of the 3d flow equations ( [ eq : x1])-([eq : z1 ] ) . , title="fig:",height=144]@xmath1 since we are interested in the description of the solution before and at the burst , in the intermediate regime where @xmath46 remains close to @xmath279 , we shall pursue our analysis by canceling the terms @xmath288 in ( [ eq : xs ] ) and in @xmath289 , because @xmath290 . this leads to the system @xmath291 where @xmath292 . close to @xmath259 and during the burst , the variable @xmath293 is essentially constant , due to the small value of the ratio @xmath294 , whereas the variable @xmath64 jumps toward smaller values , as shown in fig . [ fig : nullcline ] . therefore at leading order with respect to the small parameter @xmath294 , the solution of the last equation ( [ eq:2eqb ] ) is given by @xmath295 in the vicinity of @xmath259 . inserting this local solution ( [ eq : zbt ] ) for @xmath226 into the system ( [ eq:2eq ] ) , the exit from the sm is then described by a single equation for the local variation @xmath296 of @xmath64 of the variable @xmath64 , which is of the form @xmath297 where @xmath298 and @xmath299 sets for the @xmath300 derivative of @xmath301 with respect to the variable @xmath64 , taken at point @xmath259 ( note that @xmath302 at the critical point @xmath259 ) . using the reduced system ( [ eq:2eq ] ) , for the first two derivatives of @xmath301 with respect to @xmath64 we write @xmath303 expanding all expressions close to @xmath259 at leading order with respect to the small parameter @xmath40 , one obtains @xmath304 at this stage we can limit the series expansion in equation ( [ eq : s1 ] ) . a rough approximation for the series to converge is given by @xmath305 . using expressions ( [ eq : derf2 ] ) and the first equation ( [ eq:2eq ] ) gives the range of variation @xmath306 in this range the local form of the ak equations close to @xmath259 becomes @xmath307 which is identical to equation ( [ eq : relax3 ] ) . finally , setting @xmath308 , and @xmath309^{1/3 } \delta y$ ] , the relation ( [ eq : generic ] ) takes the generic form ( [ eq : gradt ] ) without any parameter recently proposed as a possible description of a signal before a catastrophe @xcite . @xmath310 as shown in section [ sec : dynsaddbif ] the solution of equation ( [ eq : generic ] ) displays an intermediate time scale of order @xmath311 which could be used to predict the burst , because the critical slowing down effect occurs during this time interval . in the ak equations the intermediate time scale , or precursor time , is given by the relation @xmath312 where the parameter @xmath313 is given by ( [ eq : gamma ] ) . recall that @xmath314 was shown to be the time delay between the catastrophe and the time @xmath315 at which the instability builds up along the slow manifold . for the parameter values of the above figures , we have @xmath316 , which agrees well with the lyapunov analysis , see fig . [ fig : lyap ] , and also with the precursor time deduced from the spectral analysis presented in the next subsection . in summary we have proven that the ak model , although formally different from the van der pol model , has a normal form close to the critical point that is consistent with the dynamical saddle - mode model equation studied in ( @xcite ) , with an intermediate time scale given by expression ( [ eq : prect ] ) . consequently , as for the saddle - node model , the ak model should display a response to noise with a strong increase of the correlation time occurring with a few times @xmath147 before the burst . the response to noise of the system ( [ eq : x1])-([eq : z1 ] ) is studied by setting ( @xmath317 , @xmath318 , @xmath319 ) , where @xmath249 is the solution of the noiseless ak equations , and the vector @xmath320 characterizes the fluctuations of the response to a noise source . these fluctuations result from the introduction of noise terms ( either multiplicative or additive ) in the original system . in the case of multiplicative noise sources , the response of the ak equations is a solution of the system @xmath321(1+\epsilon_x f_x(t))\\ \dot y ( t ) = [ b x(t)^2-x(t)y(t)-y(t)+az(t)](1 + \epsilon_y f_y(t))\\ \dot z ( t ) = ( c / b)[x(t)-z(t)](1 + \epsilon_z f_z(t ) ) \mathrm{. } \end{array } \right . \label{eq : eqnoise}\ ] ] we report below the result of the numerical study of the correlation function given by expressions ( [ eq : gamma ] ) and ( [ eq : gammap ] ) for small amplitude noise . in this case the variance increases slowly along the sm , following approximately the evolution of the variable @xmath322 , as reported in fig . [ fig : correl ] ( a ) . figures ( b - c ) displays the evolution of the correlation function before the burst , along the path @xmath323 . the width of the correlation function @xmath324 clearly increases as the burst is approached , showing an evident critical slowing down effect . the maximum growth of the correlation time occurs close to the time @xmath325 where the sm becomes linearly unstable , persisting until the burst . this result is in agreement with the experimental result given in section [ sec : experiment ] showing a shift of the spectrum towards low frequencies before the burst . @xmath1 ( a ) ( blue curve ) , in log scale . ( b ) correlation functions @xmath326 at times @xmath327 -green , @xmath328-orange and @xmath329-black curve versus @xmath204 . ( c ) half - height - half - with of the correlation function @xmath330 versus time @xmath50 . the input data are those of figure ( [ fig:2droite ] ) , noise amplitudes are @xmath331 for the three independent noise sources @xmath332 . , title="fig:",height=120 ] ( b ) ( blue curve ) , in log scale . ( b ) correlation functions @xmath326 at times @xmath327 -green , @xmath328-orange and @xmath329-black curve versus @xmath204 . ( c ) half - height - half - with of the correlation function @xmath330 versus time @xmath50 . the input data are those of figure ( [ fig:2droite ] ) , noise amplitudes are @xmath331 for the three independent noise sources @xmath332 . , title="fig:",height=120 ] @xmath1 @xmath1 ( c ) ( blue curve ) , in log scale . ( b ) correlation functions @xmath326 at times @xmath327 -green , @xmath328-orange and @xmath329-black curve versus @xmath204 . ( c ) half - height - half - with of the correlation function @xmath330 versus time @xmath50 . the input data are those of figure ( [ fig:2droite ] ) , noise amplitudes are @xmath331 for the three independent noise sources @xmath332 . , title="fig:",height=144 ] @xmath1 for noise of additive type the width of the correlation function @xmath326 behaves similarly , as shown in fig . [ fig : correladd ] . as pointed out in a previous section , using the expression ( [ eq : gammap ] ) to calculate the correlation time of the solution gives a biased result close to the burst where the variance is time dependent . in this case the increase of the width before the burst is reduced as illustrated in fig . [ fig : correladd](b , d ) . @xmath1 ( a ) ( dashed , blue , red curves respectively ) , ( a ) @xmath326 , ( b ) @xmath333 versus @xmath204 . ( c - d ) evolution of the half - width of the correlation functions along the trajectory before the burst , ( c ) half - width of @xmath326 ( d ) half - width of @xmath334 both curves drawn versus @xmath50 . same parameters and input data as in figures ( [ fig : correl ] ) . noise amplitudes are @xmath335 for the three independent noise sources @xmath332 . ( e ) evolution of the variance along the path ( @xmath269 ) , title="fig:",height=120 ] ( b ) ( dashed , blue , red curves respectively ) , ( a ) @xmath326 , ( b ) @xmath333 versus @xmath204 . ( c - d ) evolution of the half - width of the correlation functions along the trajectory before the burst , ( c ) half - width of @xmath326 ( d ) half - width of @xmath334 both curves drawn versus @xmath50 . same parameters and input data as in figures ( [ fig : correl ] ) . noise amplitudes are @xmath335 for the three independent noise sources @xmath332 . ( e ) evolution of the variance along the path ( @xmath269 ) , title="fig:",height=120 ] @xmath1 @xmath1 ( c ) ( dashed , blue , red curves respectively ) , ( a ) @xmath326 , ( b ) @xmath333 versus @xmath204 . ( c - d ) evolution of the half - width of the correlation functions along the trajectory before the burst , ( c ) half - width of @xmath326 ( d ) half - width of @xmath334 both curves drawn versus @xmath50 . same parameters and input data as in figures ( [ fig : correl ] ) . noise amplitudes are @xmath335 for the three independent noise sources @xmath332 . ( e ) evolution of the variance along the path ( @xmath269 ) , title="fig:",height=120 ] ( d ) ( dashed , blue , red curves respectively ) , ( a ) @xmath326 , ( b ) @xmath333 versus @xmath204 . ( c - d ) evolution of the half - width of the correlation functions along the trajectory before the burst , ( c ) half - width of @xmath326 ( d ) half - width of @xmath334 both curves drawn versus @xmath50 . same parameters and input data as in figures ( [ fig : correl ] ) . noise amplitudes are @xmath335 for the three independent noise sources @xmath332 . ( e ) evolution of the variance along the path ( @xmath269 ) , title="fig:",height=120 ] ( e ) ( dashed , blue , red curves respectively ) , ( a ) @xmath326 , ( b ) @xmath333 versus @xmath204 . ( c - d ) evolution of the half - width of the correlation functions along the trajectory before the burst , ( c ) half - width of @xmath326 ( d ) half - width of @xmath334 both curves drawn versus @xmath50 . same parameters and input data as in figures ( [ fig : correl ] ) . noise amplitudes are @xmath335 for the three independent noise sources @xmath332 . ( e ) evolution of the variance along the path ( @xmath269 ) , title="fig : " ] @xmath1 in order to present the ak model results by using the same analytical tools as used for the experiment , we have calculated the cumulative spectrum of the response @xmath336 , @xmath337 where @xmath338 , is the spectrum of the fluctuations for the response signal sampled during the time interval @xmath339 , in analogy with equations ( [ eq : spec])-([eq : cumul ] ) . the characteristic spectral width @xmath340 defined by the relation ( [ eq : cumwidth ] ) evolves in time as illustrated in fig . [ fig : specum ] for the ak model with additive noise . the horizontal segments corresponding to abscissa ( @xmath341 ) , have ordinates @xmath340 , the numerical value of the spectral width calculated during this time interval . the decrease of @xmath340 close to the burst shows well the expected shift of the spectrum towards low - frequencies , observed also in the experiments . of the response to additive noise @xmath336 , versus time before the burst , same data as in other figures . , height=144 ] in conclusion the ak model displays a range of parameter values where one can see the critical slowing - down observed also in the experiments . we have studied the ananthakrishna model for the case of small @xmath40 and @xmath225 parameter values , with @xmath342 . in this case the burst occurs close to a critical point , as noticed in @xcite . close to this critical point , the 3d flow ( namely a set of _ three _ coupled ode s ) can be reduced to the 2d system ( [ eq:2eq ] ) with a sort of langevin - like source term ( noise term with small amplitude noise)of the form ( [ eq : relax1])-([eq : relax2 ] ) discussed in section [ sec : dynsaddbif ] . the present system differs from the van der pol equation ; however , close to the critical point both models take the form of the dynamical saddle - mode model equation studied in @xcite . the ananthakrishna model displays a response to noise with a strong increase of the correlation time occurring close to the burst , or a shift of the spectrum towards low - frequency components . for this model the precursor time " is a time interval of order several times @xmath343 , also equal to several times the intermediate time scale which depends on the values of the parameters @xmath258 as given by equation ( [ eq : scgeneric ] ) . a detailed quantitative comparison between the ak model and the creep experiment will require a more complete study , which is beyond the scope of the present work . in particular it should be noted that including spatial variables in the ak model leads to chaotic solutions , which is in better agreement with our experiment where the limit cycle period may indeed vary by a factor of @xmath210 from one recorded data set to another . we have shown that the dynamical model of saddle - node transitions recently proposed to foretell catastrophes is applicable to describe the physics of collective dislocations . the experimental signal of the plastic deformation of the eutectic mixture of sn - pb subjected to a constant stress presented above clearly displays the well - known critical slowing down effect , with a precursor time of order @xmath344 of the relaxation oscillation period . this observation is shown to agree with the response to noise of a physical model proposed by ananthakrishna for describing creeping in ductile materials . this physical model has @xmath210 parameters , therefore it would be quasi - impossible to make a quantitative comparison between the experimental data and the theoretical model . here we show the ak model has a range of parameter values for which the slow - fast transition is described by the dynamical saddle - node model mentioned above . when adding a small noise to the ak model , we show that it indeed displays the critical slowing down scenario , and we derive the expression for the precursor time in terms of the @xmath210 parameter values . this gives a relation between the parameter values of the ak model in which the precursor time ( i.e. the time during which the most intense spectral components clearly shift toward low frequencies ) is of order one tenth of the relaxation oscillation period , as observed in the experiment . the important result is the following : for such systems which display relaxation oscillations with saddle - node transition , we show that there exists an intermediate time scale between the slow and fast regimes , which can be used to foretell catastrophes . unlike the majority of precursor tools that have focused on signals in the time domain , we have looked at fluctuations in the frequency domain . these spectral precursors are prone to fewer false alarms . g. ananthakrishna , physics reports * 440*,113 ( 2007 ) . y. pomeau and m. le berre , http://arxiv.org/abs/1102.5637 . 1 ) ; y. pomeau , m. le berre , j .- l . le moul , c. narteau et p. fromy , _ comptes - rendus de la 14e rencontre du non linaire _ , dits par c. josserand , m. lefranc et c. letellier , non - linaire publications , saint - etienne du rouvray ( 2011),p.135 . isbn : 978 - 2 - 9538596 - 0 - 7 . dubrovskiy and v.n . sergeev , geos , moscow . * 1 * , 222 ( 2001 ) ( in russian ) ; doklady phys . , * 49 * , 231 ( 2004 ) . v.a . dubrovskiy and v.n . sergeev , italian journal of engineering geology and environment , special issue * 1 * , 183 ( 2006 ) . y.pomeau and m. le berre , submitted in situations similar to one of the van der pol equation in the limit of large nonlinearities , the slow to fast transition occurs at the point on the slow manifold where the large component of the velocity is tangent to the slow manifold . this tangency never happens in the dr equations because , there , the large component of he velocity crosses everywhere the slow manifold at finite angle . therefore the scenario of slow to fast transition is different in the dr equations than what it is in limit cycles of the linard - van der pol like equations . this product came from tel - atomic , inc . , a company that specializes in tools for teaching advanced physics " . it is no longer available as a commercial product . one reason this instrument is no longer manufactured was the loss , due to a factory fire , of readily available chips that constituted the key component ( ne5521n integrated circuit ) used in the support electronics for its displacement sensor@xcite . a same first example of fully differential capacitive sensors is also employed ( except employed as a pair , for mechanical common mode rejection of noise ) with different electronics , in another of the company s instruments , the computerized cavendish balance @xcite . `` symmetric differential capacitance transducer employing cross coupled conductive plates to form equipotential pairs '' , u. s. patent no . 5461319 ( 1995 ) . a description of the sensor is provided at http://telatomic.com/mechanics/sensor.html http://telatomic.com/mechanics/cavendish-balance.html . the support electronics for this balance employs a novel capacitive to digital converter ( ad7745 ) , which was marketed by analog devices soon after the demise of the ne5521n . sweeping bifurcation in dynamcal systems is a well - known topic of investigations . to the best of our knowledge it has been considered in continuous transitions , either from the point of view of bifurcation theory in the time direction inverse of ours ( namely in the direction where a pair of equilibria appears in forward time imperfect bifurcation with a slowly - varying control parameter t.erneux and p. mandel , siam journal on applied mathematics , * 46*(1986 ) , pp . 1 - 15 ) or for continuous thermodynamical phase transitions ( sweeping of an instability : an alternative to self - organized criticality to get powerlaws without parameter tuning , d. sornette , j. phys . l * 4*(1994 ) 209 - 221 didier sornette ) . bouchaud and cont ( j - p . bouchaud and r. cont , eur . j. b * 6 * , 543 ( 1998 ) ) studied also the saddle - node transition with noise , but by considering the biurcation as constant at different values near the transition value , without explicit time dependence . bouchaud and r. cont , eur . j. b * 6 * , 543 ( 1998 ) .

the _ knot complement problem _ , first proposed in 1908 by heinrich tietze ( * ? ? ? * section 15 ) , asks when a knot in a given @xmath0-manifold is determined by its complement . it had been open for eighty years until in 1989 gordon and luecke @xcite proved it to be true for every knot in @xmath1 . a first step in this proof was to translate this problem into a problem concerning _ dehn surgery_. in sections [ section : complement]-[section : proof ] we recall the basic facts about the knot complement problem and dehn surgery . here i want to consider the same problem for _ legendrian knots _ in _ contact @xmath0-manifolds_. the main result is a generalization of the result by gordon and luecke , roughly speaking it says the following ( see theorem [ thm : contact1 ] in section [ section : contactcomplement ] for the precise statement ) . [ legendrian knot exterior theorem ] [ main : main1 ] + a legendrian knot in the tight contact structure of @xmath1 is determined by the contactomorphism type of its exterior . this result implies that all invariants of legendrian knots are actually invariants of the contactomorphism type of its exterior . the legendrian knot complement problem was also mentioned in @xcite . in sections [ section : contactcomplement]-[section : contactproof ] i recall the definition of _ contact dehn surgery _ and explain how to generalize the proof of gordon - luecke to this setting . the main ingredient for this is the fact that if there is a non - trivial contact dehn surgery along a legendrian knot in @xmath2 resulting again in @xmath2 , then this legendrian knot has to be a legendrian unknot of a very special type . an easy way to prove this fact ( given in section [ section : appendix ] ) is looking at all other contact surgeries resulting topologically again in @xmath1 and finding a legendrian knot in its exterior that violates the bennequin inequality after the surgery . therefore , i present in section [ section : loss ] a new formula to compute the thurston - bennequin invariant of such legendrian knots in the surgered manifold . moreover , in section [ counterexample ] i explain how to do a crossing change in a legendrian knot diagram with help of a contact dehn surgery ( a so called _ contact rolfsen twist _ ) . with that it is easy to construct counterexamples to the legendrian link exterior problem in the tight contact structure of @xmath1 , i.e. two non - equivalent legendrian links with contactomorphic exteriors . finally , in section [ topgenmfd ] and [ section : other ] i give a short discussion about the legendrian knot exterior problem in other manifolds . it turns out that in the topological setting the knot complement problem in a general manifold is equivalent to the non - existence of an _ exotic cosmetic dehn surgery _ resulting in this manifold . in the contact setting it is not clear if this equivalence is true . but in section [ section : other ] i present examples of legendrian knots in lens spaces not determined by their complements and of _ exotic cosmetic contact dehn surgeries_. all links are assumed to be tame and considered up to ( coarse ) equivalence . + let @xmath3 and @xmath4 be two links in an oriented closed @xmath0-manifold @xmath5 . then @xmath3 is * ( coarse ) equivalent * to @xmath4 , if there exists a homeomorphism @xmath6 of @xmath5 @xmath7 that maps @xmath3 to @xmath4 . then one writes @xmath8 . + the ( coarse ) equivalence is a weaker condition than the equivalence of knots up to isotopy . for example there is a reflection of @xmath1 that maps the left handed trefoil to the right handed trefoil , so this two knots are ( coarse ) equivalent , but one can show that the left handed trefoil is not isotopic to the right handed trefoil ( i.e. there is no such homeomorphism isotopic to the identity ) . one also can consider the * oriented ( coarse ) equivalence * , that means equivalence where only orientation preserving homeomorphisms of @xmath5 are allowed . in @xmath1 oriented equivalence is equivalent to isotopy ( because in @xmath1 every orientation preserving homeomorphism is isotopic to the identity ) , but in general this is a weaker condition than isotopy . a first observation is , that if two links @xmath3 and @xmath4 are equivalent , then the complements are homeomorphic . the following question is called the * link complement problem * ( or for one component links , the * knot complement problem * ) : + are two links in the same manifold with homeomorphic complements equivalent ? link complements are non - compact , but often it is much easier to work with compact manifolds . therefore , pick some regular ( closed ) neighborhood @xmath9 of a link @xmath10 in @xmath5 and call the complement @xmath11 of the interior @xmath12 of these neighborhood the * exterior * of @xmath10 . the corresponding problem whether the equivalence class of a link is determined by the homeomorphism type of its exterior is called the * link exterior problem*. ( actually , this was the problem asked by tietze in ( * ? ? ? * section 15 ) . ) by work of edwards ( * ? ? ? * theorem 3 ) this two problems are equivalent . + the first counterexample was given in 1937 by whitehead @xcite . he considered the following two links @xmath3 and @xmath4 in @xmath1 , now called * whitehead links * , see figure [ fig : whitehead ] . if one deletes one component out of @xmath3 then the remaining knot is an unknot . but if one deletes the unknot @xmath13 out of @xmath4 then the remaining knot is a trefoil . so @xmath3 can not be equivalent to @xmath4 . but the exteriors of this links are homeomorphic , as one can see as follows : consider the exteriors @xmath14 of the unknot @xmath13 . then cut open this @xmath0-manifold along the seifert disk of the unknot @xmath13 , make a full @xmath15-twist and re - glue the two disks together again . this is called a * rolfsen twist * and describes a homeomorphism of the link exteriors . by twisting several times along @xmath13 one can even get infinitely many non - equivalent links all with homeomorphic exteriors . the next natural question would be to ask if this holds on the level of knots . this is the so called knot complement theorem by gordon - luecke @xcite . [ knot complement theorem by gordon - luecke ] [ thm : gordon - luecke1 ] + let @xmath16 and @xmath17 be two knots in @xmath1 with homeomorphic complements , then @xmath16 is equivalent to @xmath17 . a starting point for proving this theorem , was to translate it into a problem concerning dehn surgery . in this section i recall the definition of dehn surgery , which is a very effective construction method for @xmath0-manifolds ( for more information see ( * ? ? ? * chapter vi ) or ( * ? ? ? * chapter 9 ) ) . roughly speaking one cuts out the neighborhood of a knot and glues a solid torus back in a different way to obtain a new @xmath0-manifold . more precisely : + let @xmath18 be a knot in a closed oriented @xmath0-manifold @xmath5 . take a non - trivial simple closed curve @xmath19 on @xmath20 and a homeomorphism @xmath21 , such that @xmath22 then define @xmath23 one says that @xmath24 is obtained out of @xmath5 by * dehn surgery * along @xmath18 with * slope * @xmath19 . one can easily show that @xmath24 is again a @xmath0-manifold independent of the choice of @xmath21 ( see ( * ? ? ? * chapter 9.f ) ) . so to specify @xmath24 one only has to describe the knot @xmath18 ( in the @xmath0-manifold @xmath5 ) together with the slope @xmath19 . to do this effectively one observes that there are two special kinds of curves on @xmath25 : * the * meridian * @xmath26 : a simple closed curve on @xmath27 , that is non - trivial on @xmath27 , but trivial in @xmath28 . * the * longitudes * @xmath29 : simple closed curves on @xmath27 , that are non - trivial on @xmath27 and intersect @xmath26 transversely exactly once . the curves shall be oriented in such a way , that the pair @xmath30 represents the positive orientation of @xmath27 in @xmath5 . one can show that the meridian @xmath26 is up to isotopy uniquely determined . but for the longitudes there are different choices . for a given longitude @xmath29 there are infinitely many other longitudes given by @xmath31 , for @xmath32 . given such a longitude @xmath29 one can write @xmath19 uniquely as @xmath33 + i want to describe surgeries along the unknot @xmath13 in @xmath1 . in this case there is a special longitude on @xmath34 . to see this observe that the exterior @xmath35 is again homeomorphic to a solid torus @xmath36 , this corresponds to the trivial genus-@xmath37 heegaard splitting of @xmath1 ( see for example ( * ? ? ? * example 8.5 . ) ) . write @xmath38 then choose the so called * surface longitude * @xmath39 of @xmath40 to be the meridian @xmath41 of @xmath42 ( see for example ( * ? ? ? * figure 8.7 . ) ) . so for surgeries along the unknot in @xmath1 one can write the slope @xmath19 uniquely as @xmath43 . now i want to show that @xmath44 is homeomorphic to @xmath1 . therefore , one first considers the so called * trivial dehn surgery * @xmath45 , where one cuts out a neighborhood of the knot and glues it back in the same way as before . so the manifold is not changing , and in this case it is again @xmath1 . then the idea is to do again a rolfsen twist along the unknot to obtain a homeomorphism from @xmath46 to @xmath47 . therefore , consider the diagram that will be specified in the following . + + ( 5,40)*+s^3=``a'';(11,40)*+=``a1 '' ; ( 20,40)*+s^3_u(_1)=``b '' ; ( 32,40)*+:==``b1 '' ; ( 45,40)*+s^1d^2=``c '' ; ( 75,40)*++=``d '' ; ( 102,40)*+t_2=``e'';(115,40)*+/_~=``e1'';(47,35)*+_0=``f '' ; ( 100,35)*+_1=``g'';(47,30)*+_0=``h '' ; ( 100,30)*+_1=``i'';(47,10)*+_0=``j '' ; ( 100,10)*+_1=``k'';(75,20)*+=``k1'';(47,5)*+_0=``l '' ; ( 95,5)*+_1+q_1=``m'';(20,0)*+s^3_u(_1+q_1)=``n'';(34,0)*+:==``n1 '' ; ( 45,0)*+s^1d^2=``o '' ; ( 75,0)*++=``p '' ; ( 102,0)*+t_2=``q'';(115,0)*+/_~=``q1'';@ > _ `` b'';``n'';@->@/_1pc/ _ `` c'';``o'';@->@/^1pc/^h `` e'';``q'';@|->__1 `` f'';``g'';@|- > `` h'';``i'';@|->__2 `` j'';``k'';@|- > `` l'';``m '' ; + + first one chooses the gluing maps @xmath48 such that they map the meridians @xmath49 as determined by the slope . by this the manifolds are fixed , but the maps @xmath48 are not . there are many possibilities to what a longitude @xmath50 can map , but the homeomorphism type of the resulting manifold is not affected by this . in this example one can choose the maps @xmath48 such that they map @xmath50 to @xmath39 . to construct a homeomorphism between the two resulting manifolds , one uses on the @xmath36-factor the identity map . if one finds a homeomorphism @xmath51 of the @xmath42-factor such that the diagram commutes , then this two maps fit together to a homeomorphism of the whole manifolds . for the map @xmath51 one can choose a @xmath52-fold dehn - twist of the solid torus @xmath42 , i.e. @xmath53 so the diagram gives rise to a homeomorphism between the two manifolds . + for all other surgeries along the unknot one computes the homology as @xmath54 so the surgeries from the above example are the only surgeries along the unknot that lead again to @xmath1 . in fact , these are the only non - trivial surgeries along an arbitrary knot in @xmath1 that yield again @xmath1 , as the following deep theorem shows . [ surgery theorem by gordon - luecke][thm : gordon - luecke2 ] + let @xmath18 be a knot in @xmath1 . if @xmath55 is homeomorphic to @xmath1 for @xmath56 , then @xmath18 is equivalent to the unknot @xmath13 . see @xcite the knot complement theorem [ thm : gordon - luecke1 ] now follows from the surgery theorem [ thm : gordon - luecke2 ] . the idea is to see a dehn surgery not as cutting out a solid torus and gluing it back in a different way . now one sees a dehn surgery as cutting out the complement of a solid torus , then doing a homeomorphism of this complement and then glue this complement back in a different way . to be more precise choose a homeomorphism @xmath57 and write @xmath58 where the gluing map is @xmath59 . then consider the surgery along @xmath17 with respect to the composition of maps @xmath60 to determine the homeomorphism type of this new manifold @xmath61 look at the following diagram : + + ( 10,40)*+s^3=``a'';(15,40)*+=``a1 '' ; ( 25,40)*+s^3_k_1 ( _ 1)=``b '' ; ( 35,40)*+:==``b1 '' ; ( 45,40)*+s^1d^2=``c '' ; ( 75,40)*++=``d '' ; ( 102,40)*+s^3=``e'';(115,40)*+/_~=``e1'';(47,35)*+_0=``f '' ; ( 100,35)*+_1=``g'';(75,20)*+=``k1'';(47,5)*+_0=``l '' ; ( 94,5)*+r_2=h(_1)=``m'';(25,0)*+s^3_k_2(r_2)=``n'';(35,0)*+:==``n1 '' ; ( 45,0)*+s^1d^2=``o '' ; ( 75,0)*++=``p '' ; ( 102,0)*+s^3=``q'';(115,0)*+/_~=``q1'';@>_f `` b'';``n'';@->@/_1pc/ _ `` c'';``o'';@->@/^1pc/^h `` e'';``q'';@|->__1 `` f'';``g'';@|->^h_1 `` l'';``m '' ; + + with similar arguments as in the example in section [ section : surgery ] this induces a homeomorphism @xmath6 from @xmath62 to @xmath61 and with the surgery theorem [ thm : gordon - luecke2 ] it follows that @xmath63 is equal to @xmath41 or @xmath17 is equivalent to the unknot @xmath13 . if @xmath64 , then the surgery @xmath61 is the trivial surgery , so the spines @xmath65 of the new solid tori are equal to the knots @xmath66 . therefore @xmath6 sends @xmath16 to @xmath17 . in the other case ( @xmath67 ) one does the same thing again but with @xmath16 and @xmath17 reversed , then it follows that @xmath68 . @xmath69 + exactly the same works also with orientations . the dehn surgery theorem [ thm : gordon - luecke2 ] also holds for oriented homeomorphism from @xmath55 to @xmath1 and oriented equivalence from @xmath18 to @xmath13 . then exactly the same proof as before shows that two knots with orientation preserving homeomorphic complements are orientation preserving equivalent . for @xmath1 this is the same as isotopic knots ( see also section [ section : complement ] ) . now i want to generalize this proof to the case of * legendrian knots * in the unique tight contact structure @xmath70 on @xmath1 , i.e. the tangent line to the knots lies always in the @xmath71-plane field given by the contact structure ( see @xcite for all basics about contact geometry and legendrian knots ) . i want to consider legendrian links up to ( coarse ) equivalence like in section [ section : complement ] and i only consider cooriented contact structures . + let @xmath3 and @xmath4 be two legendrian links in a closed contact @xmath0-manifold @xmath72 . then @xmath3 is * ( coarse ) equivalent * to @xmath4 if there exists a contactomorphism @xmath6 of @xmath5 @xmath73 that maps @xmath3 to @xmath4 . then one writes @xmath8 . + the ( coarse ) equivalence is in general a weaker condition than the equivalence given by legendrian isotopy ( for example in overtwisted contact structures on @xmath1 ) . but one can show that in @xmath2 this two concepts are the same ( see ( * ? ? ? * section 4.3 ) ) . it is a standard fact that every legendrian knot @xmath18 in a general contact @xmath0-manifold @xmath72 has a so called * standard neighborhood * @xmath28 in @xmath72 which is contactomorphic to @xmath74 with @xmath75-coordinate @xmath76 and cartesian coordinates @xmath77 on @xmath78 . this contactomorphism maps @xmath18 to @xmath79 ( see ( * ? ? ? * example 2.5.10 ) ) . it is easy to show that the boundary of this standard neighborhood is a convex surface . when i write @xmath28 for a legendrian knot @xmath18 , i always mean that @xmath28 is such a standard neighborhood and analogously to section [ section : complement ] i call the complement @xmath80 of the interior of such a standard neighborhood the * exterior * of the legendrian knot @xmath18 . but for contact manifolds there is the following problem . if @xmath81 is contactomorphic to @xmath82 , is then also @xmath83 contactomorphic to @xmath84 ? again , two equivalent legendrian links have contactomorphic exteriors . i want to show that for legendrian knots in @xmath2 the converse also holds . but this time it is not clear if this also holds for the complements . [ legendrian knot exterior theorem ] [ thm : contact1 ] + let @xmath16 and @xmath17 be two legendrian knots in @xmath2 with contactomorphic exteriors . then @xmath16 is equivalent to @xmath17 . + here @xmath16 and @xmath17 are understood to be unoriented legendrian knots , because the exterior of a knot can not see its orientation . but if one fixes an oriented longitude of the knot in its exterior , the same result holds also for oriented legendrian knots . for legendrian links the theorem is in general wrong . in section [ counterexample ] i will give some examples of legendrian links in @xmath2 that are * not * determined by the contactomorphism type of their exteriors . to generalize the proof from the first sections to legendrian knots in contact manifolds , i first want to recall the definition of contact dehn surgery along legendrian knots developed by ding and geiges in @xcite . + let @xmath18 be a legendrian knot in a contact @xmath0-manifold @xmath72 . then there is a distinguished , so called * contact longitude * @xmath29 on @xmath20 , given by pushing @xmath18 in a direction transverse to the contact planes , for example in the direction of the reeb vector field . when nothing else is said , from now on @xmath29 denotes always this contact longitude . it is easy to show that for a slope @xmath85 the surgered manifold @xmath24 is determined by the rational number @xmath86 , called the * ( contact ) surgery coefficient * ( see for example ( * ? ? ? * section 9.g ) ) . sometimes it will be necessary to express the slope @xmath19 with respect to the surface longitude @xmath87 , i.e. @xmath88 , then the rational number @xmath89 is called the * ( topological ) surgery coefficient*. ( recall that the surface longitude for a nullhomologous knot is obtained by pushing the knot into one of its seifert surfaces . ) from now on , depending on the context , i will denote by @xmath19 the slope or the corresponding surgery coefficient . now one wants to do dehn surgery along legendrian links with respect to this contact longitude . one can show that the contact structure of the old manifold extends to a contact structure of the resulting manifold . + let @xmath18 be a legendrian knot in a contact @xmath0-manifold @xmath72 . + ( 1 ) then @xmath24 carries a contact structure @xmath90 , which coincides with the old contact structure @xmath91 on @xmath92 . + ( 2 ) for @xmath93 one can choose @xmath90 to be tight on the new glued - in solid torus . + ( 3 ) for @xmath94 this tight contact structure on this new solid torus is unique . + one says @xmath95 is obtained from @xmath72 by * contact dehn surgery * along the legendrian knot @xmath18 with slope @xmath19 . [ sketch of proof ] + the neighborhood @xmath28 is chosen such that the boundary is a convex surface . the germ of a contact structure along a convex surface is determined by some simple data on that surface . therefore , it is easy to glue contact structures along convex surfaces . tight contact structures on solid tori with prescribed boundary conditions have been classified by honda @xcite . in particular , such a contact structure always exists if @xmath93 and is unique if @xmath94 . but in general there are more possibilities to choose this contact structure , so for other slopes the result of contact dehn surgery is in general not unique . for details see @xcite or @xcite . + consider the contact surgery along the legendrian unknot with @xmath96 and surgery coefficient @xmath97 ( see figure [ fig : exoticsurgery ] ) . the surgery coefficient @xmath97 corresponds then to the slope @xmath98 , where @xmath29 denotes the contact longitude . by expressing this slope with respect to the surface longitude @xmath39 given by the meridian of the complementary solid torus one gets @xmath99 . so the resulting manifold is by the example from section [ section : surgery ] topologically again @xmath1 . next i want to show that the resulting contact structure is unique and leads again to @xmath70 ( if one requires the contact structure on the new glued - in solid torus to be tight ) . in general a contact dehn surgery with surgery coefficient not of the form @xmath100 is not unique . but actually in this example it is . too see this , one first uses the algorithm in ( * ? ? ? * section 1 ) ( see also @xcite ) to change the contact surgery diagram into contact surgeries along a link with only @xmath101 surgery coefficients ( see figure [ fig : exoticsurgery ] ) . observe that different choices of stabilizations lead in general to different contact structures ( and correspond exactly to the different contact structures on the glued - in solid torus ) , but in this case the resulting contact structures are contactomorphic . the contactomorphism of the resulting manifold is induced by the contactomorphism @xmath102 of the old @xmath2 , that maps one link to the other ( see also ( * ? ? ? * section 9 ) ) . to see that this contact structure is really @xmath70 , it is enough to show that the contact structure is stein fillable . in ( * lemma 4.2 . ) it is shown , that @xmath103-contact dehn surgery along the legendrian unknot with @xmath96 leads to @xmath104 with the unique stein fillable contact structure on it . because @xmath105-contact dehn surgery preserves stein fillability the claim follows . to prove the legendrian knot exterior theorem [ thm : contact1 ] i generalize the surgery theorem by gordon - luecke [ thm : gordon - luecke2 ] . [ contact dehn surgery theorem][thm : contact2 ] + let @xmath18 be a legendrian knot in @xmath2 . if @xmath106 is contactomorphic to @xmath2 for @xmath56 then @xmath18 is equivalent to a legendrian unknot @xmath107 with @xmath108 and @xmath109 . + ( 1 ) the classification of legendrian unknots in @xmath2 by eliashberg - fraser @xcite says , that two such legendrian unknots are equivalent ( with orientations of the knots ) if and only if they have the same @xmath110 and @xmath111 . of course the exterior of a knot can not see its orientation , which for legendrian knots is given by the sign of @xmath111 . so the knots @xmath107 are unique up to equivalence ( of unoriented legendrian knots ) . a front projection of a legendrian unknot of type @xmath107 is shown in figure [ fig : legendreunknoten ] . + ( 2 ) for general slopes the contact structure @xmath90 is not unique . so one should read theorem [ thm : contact2 ] as follows : if there is a contact structure @xmath90 on @xmath55 , such that @xmath106 is obtained from @xmath72 by contact dehn surgery along @xmath18 with slope @xmath19 and if @xmath106 is contactomorphic to @xmath2 for @xmath56 , then the conclusion holds . the proof of this contact dehn surgery theorem [ thm : contact2 ] is given in section [ section : appendix ] . the proof of the legendrian knot exterior theorem [ thm : contact1 ] is now similar to the topological case . pick a contactomorphism @xmath112 and then consider again the following diagram : + + ( 0,40)*+(s^3,_st)=``a'';(8,40)*+=``a1 '' ; ( 24,40)*+(s^3_k_1 ( _ 1),_k_1 ( _ 1))=``b '' ; ( 40,40)*+:==``b1 '' ; ( 52,40)*+(s^1d^2,_1)=``c '' ; ( 75,40)*++=``d '' ; ( 100,40)*+(s^3,_st)=``e'';(114,40)*+/_~=``e1'';(53,35)*+_0=``f '' ; ( 99,35)*+_1=``g'';(75,20)*+=``k1'';(53,5)*+_0=``l '' ; ( 92,5)*+r_2:=h(_1)=``m'';(24,0)*+(s^3_k_2 ( r_2),_k_2 ( r_2))=``n'';(40,0)*+:==``n1 '' ; ( 52,0)*+(s^1d^2,_1)=``o '' ; ( 75,0)*++=``p '' ; ( 100,0)*+(s^3,_st)=``q'';(114,0)*+/_~=``q1'';@>_f `` b'';``n'';@->@/_1pc/ _ `` c'';``o'';@->@/^1pc/^h `` e'';``q'';@|->__1 `` f'';``g'';@|->^h_1 `` l'';``m '' ; + + here the contact structure @xmath113 denotes the unique tight contact structure on @xmath36 with convex boundary corresponding to the slope @xmath114 . because the contactomorphisms @xmath115 and @xmath51 on the two factors agree on the boundary convex surfaces , which determine the germ of the contact structures , these two maps glue together to a contactomorphism @xmath6 of the whole contact manifolds . from the contact dehn surgery theorem [ thm : contact2 ] it follows that @xmath63 is equal to @xmath41 or @xmath17 is equivalent to @xmath107 for some @xmath116 . if @xmath64 then this is the trivial contact dehn surgery , and so the contactomorphism @xmath6 maps @xmath16 to @xmath17 . in the other case one makes the same argument with @xmath16 and @xmath17 reversed and concludes that @xmath16 is equivalent to @xmath117 for some @xmath118 . again the classification result of eliashberg - fraser @xcite implies that @xmath119 is equivalent to @xmath120 if and only if @xmath121 . to show the last statement one observes that @xmath122 is a solid torus with tight contact structure and convex boundary . therefore , one can compute @xmath123 also as half the number of intersection points of the seifert disk of @xmath107 with the dividing set of the convex boundary ( see ( * ? ? ? * theorem 2.30 ) ) . because the seifert disks of @xmath107 and @xmath117 are both given by the @xmath78-factors of the exterior solid tori and because the exteriors are contactomorphic , the number of intersection points stays the same . to prove the contact dehn surgery theorem [ thm : contact2 ] one first determines with the surgery theorem by gordon - luecke [ thm : gordon - luecke2 ] all surgeries that lead again to @xmath1 . the main problem is then that there are always many different choices for extending the contact structure over the new glued - in solid torus . but a very simple proof can be given by finding some new legendrian knots in the exteriors of the surgery knots , that violates the bennequin inequality in the new contact manifold . for doing this i want to present in this section a formula for computing the thurston - bennequin invariant of a legendrian knot in a surgered manifold . the main problem when doing this , is that the thurston - bennequin invariant is only defined for nullhomologous ( or rationally nullhomologous ) knots and a surgery in general destroys this property . but as long as the resulting manifold is a homology sphere ( or a rationally homology sphere ) every knot has to be nullhomologous ( or rationally nullhomologous ) . in this case formulas for computing the new thurston - bennequin invariants out of the old ones and out of the algebraic surgery dates are given in ( * ? ? ? * lemma 6.6 ) , ( * ? ? ? * lemma 2 ) and ( * ? ? ? * lemma 6.4 ) . here i first want to give an easy ( and easy to check ) condition ( out of the algebraic surgery data ) when such a knot is nullhomologous in the new manifold . if this is the case i secondly show how to compute out of this dates the new thurston - bennequin invariant . + let @xmath124 be an oriented link ( where the choice of orientation is not important ) . and let @xmath5 be the @xmath0-manifold obtained out of @xmath1 by dehn surgery along @xmath10 with topological surgery coefficients @xmath125 , for @xmath126 . denote by @xmath127 an oriented knot in @xmath1 and @xmath5 depending on the context . for simplicity write the linking numbers as @xmath128 , for @xmath129 . set also @xmath130 the matrix @xmath131 is a generalization of the linking matrix , because for @xmath132 the matrix @xmath131 is the linking matrix . the knot @xmath133 is called * nullhomologous * in @xmath5 if @xmath134=0\in h_1(m;\mathbb{z})$ ] . one can show ( see for example ( * ? ? ? * page 123 ) ) that this is equivalent to the existence of a seifert surface for the knot , so the surface longitude for a knot is defined if and only if the knot is nullhomologous . with the following lemma one can decide from the algebraic surgery data if such a knot is nullhomologous in the surgered manifold . [ nullhomologous knots ] + @xmath133 is nullhomologous in @xmath5 if and only if there exists an @xmath135 such that @xmath136 . + it is easy to compute the homology of @xmath5 ( see for example ( * ? ? ? * proposition 5.3.11 ) ) as @xmath137 where the generators of the @xmath138-factors are given by right - handed meridians @xmath139 corresponding to the components @xmath140 . next i express @xmath133 as a linear combination of the @xmath139 . one can show that the coefficients are the linking numbers @xmath141 , i.e. @xmath142=\sum_{i=1 } ^n l_{i0}\mu_i.\end{aligned}\ ] ] so @xmath133 is nullhomologous if and only if one can express @xmath134=\sum l_{i0}\mu_i$ ] as a linear combination of the relations , i.e. if there exists integers @xmath143 , @xmath126 , such that @xmath144 by comparing the coefficients and by writing the corresponding equations in vector form one sees that this is true if and only if there exists a vector @xmath135 such that @xmath136 . now assume @xmath133 is a legendrian link in @xmath145 . and let @xmath91 be a contact structure on @xmath5 that coincides with @xmath70 outside a tubular neighborhood of @xmath10 . for example if @xmath10 is also a legendrian link in @xmath2 and @xmath72 is the result of a contact dehn surgery along @xmath10 . but it is important to notice that the setting here is a more general one , one can also use contact surgery along transversal knots or surgery along a knot that is not adapted to the contact structure . here all surgery coefficients are understood to be topological surgery coefficients , i.e. with respect to the surface longitude @xmath87 which has linking number zero with the knot . ( so if one has a legendrian surgery diagram one first has to change the contact surgery coefficients to topological surgery coefficients , for example with the formula @xmath146 . ) [ computing the thurston - bennequin invariant ] + if @xmath133 is nullhomologous in @xmath5 , then one can compute the new thurston - bennequin number @xmath147 of @xmath133 in @xmath72 from the old one @xmath148 of @xmath133 in @xmath2 as @xmath149 where @xmath150 is a vector given by the formula from the previous lemma . + let @xmath87 be the surface longitude of @xmath133 in @xmath1 , i.e. @xmath151 . and let @xmath152 be the contact longitude of @xmath133 in @xmath2 . then @xmath148 is given by @xmath153 because the contact longitude is defined by the contact structure along @xmath133 and the contact structure does not change near @xmath133 by doing the surgery along @xmath10 the contact longitude @xmath152 represents also the contact longitude in @xmath72 . but in general the surface longitude @xmath87 changes . since the knot @xmath133 is nullhomologous in @xmath5 , there is a unique @xmath154 such that @xmath155 ( this is the new surface longitude ) . then @xmath147 is given by @xmath156 putting this together leads to @xmath157 so the only thing left is to compute @xmath6 . as in the proof of the previous lemma one computes the homology of @xmath158 as @xmath159 and expresses @xmath87 as @xmath160 so @xmath161 is zero in @xmath162 if and only if there exists integers @xmath163 , @xmath164 , such that @xmath165 by comparing the coefficients this is equivalent to the existence of a vector @xmath166 such that @xmath167 if one chooses for @xmath168 a solution @xmath150 from the first lemma the formula follows . [ computing @xmath110-numbers with this formulas ] + ( 1 ) consider the surgery diagram from figure [ fig : computing - tb - ex1 ] ( i ) . the old thurston - bennequin invariant of @xmath133 is @xmath169 . and because the surgery along the unknot @xmath3 with topological framing @xmath170 leads again to @xmath1 the knot @xmath133 is again nullhomologous in the new manifold . this can be checked also with the formula from above . for @xmath150 one gets the equation @xmath171 . and therefore @xmath172 . observe again that this result does not depend on the explicit contact structure chosen for the surgery . + ( 2 ) consider the surgery diagram from figure [ fig : computing - tb - ex1 ] ( ii ) . again the surgery leads to @xmath1 so the resulting knot is again nullhomologous . the new thurston - bennequin invariant can be computed as follows . for @xmath150 one gets the equation @xmath173 . and therefore @xmath174 . in the next section we will see that the contact surgery theorem is an easy corollary out of these two examples . + ( 3 ) there are also examples where the solution @xmath175 of @xmath176 is not unique . but the result of @xmath177 , of course , is not affected by this . for example consider the surgery diagram from figure [ fig : computing - tb - ex3 ] ( i ) . first one has to check if the knot @xmath133 is again nullhomologous in the surgered manifold . for this , one has to look if there exists a solution @xmath178 of @xmath179 two obvious solutions are @xmath180 so the new knot is nullhomologous and the new thurston - bennequin invariant can be computed out of the old one as follows @xmath181 observe that this result does not depend on the choice of @xmath150 . + [ subsection : computtingrationallytb ] this results can be easily generalized to rationally nullhomologous knots . a knot @xmath133 in @xmath5 is called * rationally nullhomologous * if there exists a natural number @xmath182 such that @xmath183=0\in h_1(m;\mathbb{z})$ ] . for rationally nullhomologous legendrian knots in contact @xmath0-manifolds one can generalize the thurston - bennequin invariant to the so called * rational thurston - bennequin invariant * @xmath184 ( see for example ( * ? ? ? * definition 6.2 ) or ( * ? ? ? * section 2 and 3 ) , for the fact that this is in general well defined see ( * ? ? ? * section 5 ) ) . with the same notation as from the first sections one gets the following generalizations of these results . [ computing the rationally thurston - bennequin invariant ] + ( 1 ) @xmath133 is rationally nullhomologous in @xmath5 if and only if there exists a natural number @xmath182 and a vector @xmath135 such that @xmath185 . + ( 2 ) if @xmath133 is rationally nullhomologous in @xmath5 then one can compute the new rationally thurston - bennequin number @xmath186 of @xmath133 in @xmath72 from the old one @xmath148 of @xmath133 in @xmath2 as follows @xmath187 + the proofs are similar to the ones in the foregoing subsections . for the first part one has to change the condition @xmath134=0 $ ] to @xmath183=0 $ ] and do the same computations again . the second part works similar . if @xmath133 is only rationally nullhomologous in @xmath5 , then the thurston - bennequin numbers are related as follows ( see also ( * ? ? ? * proof of lemma 2 ) ) . @xmath188 then same computations as in the first part lead to the result . [ computing @xmath110-numbers of rationally legendrian unknots in lens spaces ] this formula for computing the rational thurston - bennequin invariant is very useful to calculate @xmath184 in lens spaces . for example , consider the surgery diagram from figure [ fig : computing - tb - ex3 ] ( ii ) . the @xmath189-surgery along @xmath3 leads to the lens space @xmath190 . to check if the knot @xmath133 is nullhomologous one has to solve the equation @xmath191 . for @xmath192 this equation has no solution in @xmath138 and therefore @xmath133 is not nullhomologous in the surgered manifold . but ( for @xmath193 ) the equation @xmath194 has a solution , for example @xmath195 and @xmath196 . so @xmath133 is rationally nullhomologous in @xmath190 . the rationally thurston - bennequin invariant is computed as follows @xmath197 observe again that the result is independent of the chosen solution @xmath150 and independent of the chosen contact structure on the new glued - in solid torus . + one can also study the same problem for a surgery in a general contact manifold ( not on @xmath2 ) . this is motivated by ( * ? ? ? * lemma 6.4 ) . consider now @xmath198 an oriented nullhomologous link in some contact @xmath0-manifold @xmath199 . denote by @xmath72 some result of contact surgery along @xmath10 and by @xmath200 an oriented nullhomologous legendrian knot in @xmath199 and @xmath72 , depending on the context . then one gets exactly the same formulas as before . the only part changing in the proof is that the homologies are different , for example @xmath201 for more details see ( * ? ? ? * proof of lemma 6.4 ) . if one has a surgery diagram with also @xmath37-handles included , then one can use the above methods as well . the first possibility is to change all @xmath37-handles into @xmath202-surgeries along unknots and the second possibility is to think of the surgery diagram as a surgery diagram in @xmath203 ( represented by @xmath116 @xmath37-handles ) and then use the above extension . in further work @xcite there will also appear a formula like this for the rotation number , that generalizes the formulas from ( * ? ? ? * lemma 6.6 ) , ( * ? ? ? * lemma 2 ) and ( * ? ? ? * lemma 6.4 ) , together with a similar formula for the self - linking number of transversal knots . let @xmath18 be a legendrian knot in @xmath2 such that @xmath106 is again a contact @xmath1 ( with any contact structure ) for @xmath56 . from the ( topological ) surgery theorem [ thm : gordon - luecke2 ] it follows , that @xmath18 is topologically equivalent to an unknot @xmath13 . ( so i will write @xmath13 instead of @xmath18 . ) in the example from section [ section : surgery ] and the following remark i explained that the topological surgery coefficient ( with respect to the surface longitude @xmath39 ) has to be of the form @xmath170 , for @xmath204 . now i want to show that every resulting contact structure @xmath205 is overtwisted if @xmath13 is not coarse equivalent to an unknot of the form @xmath107 . to do this i will show that in the resulting contact @xmath0-spheres there exists legendrian knots that can not be realized in @xmath70 . for this , one considers the first two examples from subsection [ subsection : computtingtb ] . by doing a rolfsen twists along @xmath13 one sees that the legendrian knot @xmath133 from the first example remains an unknot in the new surgered manifold . for @xmath206 one gets @xmath207 ( independent of the choice of the contact structure on the new glued - in solid torus ) . according to the bennequin inequality this knot can not lie in a tight contact structure . so for @xmath206 it is not possible to have a non - trivial contact surgery from @xmath2 to itself . for @xmath208 one looks at the second example from subsection [ subsection : computtingtb ] . in figure [ fig : proofofcontactsurgerytheorem21 ] it is shown ( again by doing a rolfsen twist ) that @xmath133 becomes in the new surgered manifold a negative @xmath209-torus knot @xmath210 . in ( * section 1 ) it is shown that the maximal @xmath110-number of such a knot in @xmath2 is given by @xmath211 , which is smaller than @xmath212 . for @xmath13 not coarse equivalent to an legendrian unknot of the form @xmath107 this example can be realized as a contact surgery along a legendrian knot , which proves the result . the next natural question is if there exists ( like in the topological case ) legendrian links not determined by the contactomorphism type of their exteriors . if one wants to generalize the examples of whitehead one needs a contact analogon of a rolfsen twist . in the example from section [ section : contactsurgery ] i gave a contact surgery from @xmath2 to @xmath2 . topologically this surgery represents a @xmath103-dehn surgery along an unknot . deleting such components from contact surgery diagrams leads to a contact analogon of a @xmath105-rolfsen twist . but it is not clear how the rest of the diagram changes then . + figure [ fig : rolfsentwist ] shows two moves for contact dehn surgery diagrams . there , @xmath10 is an arbitrary legendrian link with surgery coefficients @xmath101 , that can go arbitrarily often through the unknot ( even one knot several times is allowed ) . + of course one can do such a contact rolfsen twist more than once , this corresponds to doing @xmath213-contact dehn surgeries along @xmath116 legendrian push - offs of the unknot . it is easy to see that this is the same as doing a single contact dehn surgery along the unknot with contact surgery coefficient @xmath214 . + first of all one observes that the first row in figure [ fig : rolfsentwist ] maps to the second row under the contactomorphism @xmath102 . so it is enough to prove the second row . for this , one observes that one can put the whole legendrian link @xmath10 together with the legendrian unknot @xmath13 on the pages of an abstract open book for @xmath2 ( see for example ( * ? ? ? * proof of theorem 5.5 ) ) . a part of this abstract open book is shown in figure [ fig : rolfsenopenbook ] . the page is pictured in gray and a part of the monodromy is described as a right handed dehn twist along the blue curve . one has to read the red curve in the open book , as necessarily many parallel copies . observe that the abstract open book determines the contact manifold only up to contactomorphism , as the surgery diagram does . next one constructs from this an abstract open book for the surgered manifold as explained in @xcite . first , one puts the legendrian link @xmath10 together with the legendrian unknots @xmath215 and @xmath216 from the @xmath217-surgery diagram on the page of an abstract open book for @xmath2 . this is shown in figure [ fig : rolfsenproofneu ] on the upper right side . the additional stabilization of the legendrian unknot @xmath216 corresponds to a stabilization of the open book as shown in ( * ? ? ? * figure 1 and 2 ) . the monodromy of the open book for @xmath2 is given by right handed dehn twists along the blue curves . the monodromy of the surgered manifold is obtained by composing this old monodromoy with right handed dehn twists along the @xmath105-surgery knot and left handed dehn twists along the @xmath103-surgery knot ( see ( * ? ? ? * proposition 8) ) . two dehn twists cancel each other and one gets the open book in the bottom right corner of figure [ fig : rolfsenproofneu ] . that open book represents the surgery diagram of the once stabilized legendrian link @xmath10 shown in the bottom left corner of figure [ fig : rolfsenproofneu ] . observe also that the topological surgery coefficients of @xmath10 change exactly as prescribed by the topological rolfsen twists . [ counterexamples to the legendrian link complement problem ] + with this contact rolfsen twist one can give counterexamples to the legendrian link complement problem . but it is not as easy as in the topological setting . the main point there was that the new glued - in solid tori is again a tubular neighborhood of the spine of this torus . this is not the case in the contact setting . to see this , consider a legendrian knot @xmath18 in a contact @xmath0-manifold @xmath72 and the result of contact surgery along this knot @xmath95 . then the slope of the new glued - in solid torus is @xmath19 . but if this new glued - in solid torus would be the standard neighborhood of a legendrian knot then the slope would be of the form @xmath218 . so for general @xmath19 this is not the case . therefore , one looks at the contact surgery diagram with one @xmath103-surgery along @xmath215 and one @xmath105-surgery along @xmath216 . the new glued - in solid tori are again in a canonical way standard neighborhoods of the legendrian spines @xmath219 and @xmath220 . from the proof of the lemma it follows that this spines @xmath219 and @xmath220 lie in the resulting surgery diagram as shown in figure [ fig : rolfsentwistmitseele ] . + ( 1 ) consider the two legendrian links @xmath221 and @xmath222 in @xmath2 as depicted in figure [ fig : counterexample1 ] . these two links are not equivalent because their triples of @xmath110-numbers are different : @xmath223 ; @xmath224 . but their exteriors are contactomorphic , as one can see as follows . one does a @xmath103-surgery along @xmath215 and a @xmath105-surgery along @xmath216 . the resulting manifold is again @xmath2 , in which the legendrian knot @xmath10 looks now like @xmath225 in figure [ fig : counterexample1 ] on the right . so in the exteriors of @xmath226 and @xmath227 the legendrian knots @xmath10 and @xmath225 are the same . + ( 2 ) one can also get examples with different topological types and the same @xmath110-numbers . for that consider the legendrian links in figure [ fig : counterexample2 ] similar to the whitehead links as in the example from section [ section : complement ] . in the left legendrian link in figure [ fig : counterexample2 ] all three knots are unknots , but in the right link the knot @xmath225 is non - trivial , so they can not be equivalent . but their exteriors are contactomorphic with the same argument as in the foregoing example . + ( 3 ) by doing contact dehn surgeries corresponding to an @xmath116-fold rolfsen twist , for @xmath208 , one gets in both foregoing examples infinitely many different legendrian links with contactomorphic exteriors . instead of looking at legendrian knots in @xmath2 one can also look at legendrian knots in general contact @xmath0-manifolds and study the legendrian knot exterior problem for these knots . before studying the contact case in the next section , i first recall some basic facts about this problem in the topological setting . of course , one can transfer this problem in an arbitrary manifold to a problem concerning dehn surgery like before in the proof of theorem [ thm : gordon - luecke1 ] . + let @xmath18 be a knot in a @xmath0-manifold @xmath5 , such that there is no non - trivial dehn surgery along @xmath18 resulting again in @xmath5 . then the equivalence type of @xmath18 is determined by the diffeomorphism type of its complement . but to proof the hypothesis of this lemma turns out to be very difficult and in general , both , the conclusion and the hypothesis of the lemma are not true . to see this , consider the following counterexample ( see also @xcite ) . [ two non - equivalent knots with the same complements ] + consider two different dehn surgeries along the unknot @xmath13 in @xmath1 with the ( topological ) surgery coefficients @xmath228 and @xmath229 , leading to the lens spaces @xmath230 and @xmath231 . by the classification of lens spaces ( * ? ? ? * exercise 5.3.8 . ( b ) ) this two lens spaces are orientation preserving homeomorphic and the homeomorphism is given by interchanging the two solid tori ( see also ( * ? ? ? * section 2 ) ) . from this dehn surgery example it is easy to find two non - equivalent knots with the same exteriors . for this write + + ( 0,5)*+l(5,2)=``a'';(10,5)*+=``a1 '' ; ( 24,5)*+s^3_u ( r_i)=``b '' ; ( 39,5)*+:==``b1 '' ; ( 52,5)*+s^1d^2=``c '' ; ( 70,5)*++=``d '' ; ( 90,5)*+s^3=``e'';(104,5)*+/_~=``e1'';(52,0)*+_0=``f '' ; ( 89,0)*+r_i=``g'';@|- > `` f'';``g '' ; + + and consider the knots @xmath232 the knots @xmath16 and @xmath17 given as the spines of the new glued - in solid tori represents the spines of the genus-1 heegaard splitting of @xmath230 . as tubular neighborhood of @xmath66 one chooses the whole new glued - in solid tori @xmath36 , therefore the exterior is in both cases @xmath233 . it remains to show that these two knots are not equivalent . therefore , assume that there is an orientation preserving homeomorphism @xmath234 by restricting @xmath6 to the complementary solid tori @xmath235 one gets a homeomorphism @xmath236 which sends the slope @xmath237 to the slope @xmath238 . but such a map can not exists because all orientation preserving homeomorphisms of solid tori are isotopic to dehn twists along meridians . so @xmath16 is not orientation preserving equivalent to @xmath17 in @xmath230 . with the same methods as above it is easy to show ( see @xcite ) that if @xmath16 and @xmath17 are the cores of the two solid tori in the standard heegaard splitting of @xmath190 , then they have homeomorphic complements , but there is an orientation preserving ( reversing ) homeomorphism of @xmath190 sending @xmath16 to @xmath17 if and only if @xmath239 ( @xmath240 ) . the key point in the foregoing examples is that there is a so called * exotic cosmetic surgery * , that means two surgeries along the same knot resulting in the same manifold but with different slopes , such that there is no homeomorphism of the knot exterior mapping one slope to the other ( see @xcite ) . one can show that every knot in a given @xmath0-manifold is determined by its complement if and only if this manifold can not be obtained by exotic cosmetic surgery from another manifold . [ thm : cosmetic ] + the following two claims are equivalent . 1 . let @xmath16 and @xmath17 be knots in a closed @xmath0-manifold @xmath5 with homeomorphic complements . then @xmath16 is equivalent to @xmath17 . 2 . let @xmath241 be a knot in a closed @xmath0-manifold @xmath242 , such that @xmath243 and @xmath244 are both homeomorphic to @xmath5 ( for @xmath245 ) . then there exists a homeomorphism of the knot exterior @xmath246 mapping one slope to the other . + ( 1 ) the statement holds in the oriented and unoriented case . + ( 2 ) one can take the manifold @xmath242 to be homeomorphic to @xmath5 , this explains the name cosmetic surgery . [ proof of theorem [ thm : cosmetic ] ] + the proof of @xmath247 works exactly as in the foregoing example . the implication @xmath248 is similar to the proof of theorem [ thm : gordon - luecke1 ] . so the study of the knot complement problem is equivalent to the study of exotic cosmetic surgeries . in the topological setting some is known , but much remains open . beside the discussed knot complement theorem for knots in @xmath1 , gabai showed in @xcite that knots in @xmath104 ( or more generally in a connected sum of arbitrary @xmath249- or @xmath250-bundles over @xmath75 ) are determined by their complements . but in general manifolds this will not hold . building up on work by mathieu @xcite , rong classified in @xcite all knots in @xmath0-manifolds with seifert fibered complements , that are not determined by their complements . this knots are given by the spines of the solid tori in the standard heegaard splitting of some special lens spaces @xmath190 as described at the beginning of this section or as exceptional fibers of index @xmath71 in special seifert fibered manifolds . but all this homeomorphisms sending one of the exceptional fibers to another one have to be orientation reversing . later matignon @xcite proved that all non - hyperbolic knots in atoroidal irreducible seifert fibered @xmath0-manifolds are determined by their complements ( except the cores of the standard heegaard splittings in lens spaces ) . for hyperbolic knots there is until now only one counterexample . in @xcite bleiler , hodgson and weeks construct two non - equivalent hyperbolic knots in @xmath251 with orientation reversing homeomorphic complements . they also give very good reasons for the conjecture that all knots in hyperbolic @xmath0-manifolds are determined by their oriented complements . recently ichihara and jong @xcite found examples of knots in hyperbolic manifolds with orientation reversing homeomorpic complements . all together this leads to the still open oriented knot complement conjecture in general manifolds : + if @xmath16 and @xmath17 are knots in a closed oriented @xmath0-manifold @xmath5 with orientation preserving homeomorphic complements ( not homeomorphic to @xmath36 ) , then the knots are orientation preserving equivalent . with theorem [ thm : cosmetic ] and the following remark this is equivalent to the cosmetic surgery conjecture formulated in @xcite : + exotic cosmetic surgeries ( not resulting in a lens space ) are never orientation preserving ( or * truly * ) cosmetic . many of the mentioned results rely on the classification of all dehn surgeries in a solid torus resulting again in a solid torus by berge @xcite and gabai @xcite . but in the last years also heegaard floer homology turned out to be a very useful tool to study such questions . for example , wang @xcite showed that there are no exotic dehn surgeries along seifert genus-@xmath37 knots in @xmath1 . the same holds for non - trivial algebraic knots in @xmath1 by @xcite . finally , in @xcite and @xcite it is shown that knots in @xmath10-space homology spheres are determined by their complements . in particular , knots in the poincar sphere are determined up to orientation preserving equivalence by the oriented homeomorphism types of their complements . as far as i know , nothing is known about this in the contact setting . the question is which results from the topological setting generalize to the contact setting and where are the differences . in the foregoing section i presented examples of non - equivalent knots in lens spaces with homeomorphic exteriors . the first interesting question is if one can generalize this examples to the contact setting . [ non - equivalent legendrian knots with the same complements ] + consider a legendrian knot @xmath18 with @xmath252 . a standard neighborhood of @xmath18 is given by @xmath253 the surface longitude is given by @xmath254 and the contact longitude by @xmath218 . take two copies @xmath255 and @xmath256 of this standard neighborhood and glue them together along their boundaries to obtain the lens space @xmath190 as follows : + + ( 20,10)*+l(p , q)=``a'';(30,10)*+==``a1 '' ; ( 40,10)*+v_1=``c '' ; ( 60,10)*++=``d '' ; ( 80,10)*+v_2=``e'';(95,10)*+/_~=``e1'';(40,5)*+_1=``f '' ; ( 80,5)*+-q_2+p_2,=``g'';(40,0)*+_1=``h '' ; ( 80,0)*+r_2+s_2,=``i'';@|- > `` f'';``g '' ; @|- > `` h'';``i '' ; + + where @xmath257 . it is a standard fact that the contact structures on the solid tori fit together to a contact structure in the new lens space @xmath190 if the gluing map sends the contact longitude of @xmath258 to the contact longitude of @xmath259 . this leads to the conditions @xmath260 and @xmath261 . putting this together it follows that @xmath262 . in particular , it follows that if a contact lens space is obtained by gluing together two standard neighborhoods of legendrian knots , then the lens space is of the form @xmath263 . but in this case the spines @xmath16 and @xmath17 of the heegaard tori are topologically equivalent , as was explained in the foregoing section . ( 20,10)*+l(p , p-1)=``a'';(32,10)*+==``a1 '' ; ( 40,10)*+v_1=``c '' ; ( 60,10)*++=``d '' ; ( 80,10)*+v_2=``e'';(95,10)*+/_~=``e1'';(40,5)*+_1=``f '' ; ( 80,5)*+(1+pn)_2+p_2,=``g'';(40,0)*+_1=``h '' ; ( 80,0)*+-pn^2_2+(1-pn)_2.=``i'';@|- > `` f'';``g '' ; @|- > `` h'';``i '' ; + + this gluing map sends the contact longitude of @xmath258 to the contact longitude of @xmath259 and leads therefore to a contact structure on the lens space . the legendrian knots @xmath16 and @xmath17 represent the spines of the heegaard tori in this lens space . therefore they both have order @xmath264 in @xmath265 . the extended seifert longitude @xmath63 of @xmath17 is given by @xmath266 ( it is nullhomologous in the exterior of @xmath17 and @xmath267 . ) so one computes @xmath268 to compute @xmath269 one first computes the inverse gluing map as : + + ( 20,10)*+l(p , p-1)=``a'';(32,10)*+==``a1 '' ; ( 40,10)*+v_2=``c '' ; ( 60,10)*++=``d '' ; ( 80,10)*+v_1=``e'';(95,10)*+/_~=``e1'';(40,5)*+_2=``f '' ; ( 80,5)*+(1-pn)_1-p_1,=``g'';(40,0)*+_2=``h '' ; ( 80,0)*+pn^2_1+(1+pn)_1.=``i'';@|- > `` f'';``g '' ; @|- > `` h'';``i '' ; + + then one sees that the extended seifert longitude of @xmath16 is given by @xmath270 and so @xmath271 so the rational thurston - bennequin invariants are different , therefore @xmath16 and @xmath17 are not legendrian isotopic and then also not coarse equivalent . so in contrast to the topological case one gets obvious counterexamples to the legendrian knot exterior conjectures exactly in the lens spaces @xmath263 . it would be interesting if one can generalize the examples from this section to a general classification result in lens spaces . a first step to study the legendrian knot exterior problem in general contact manifolds would be a generalization of theorem [ thm : cosmetic ] . the generalization of the implication @xmath248 one can prove similarly to the proof of theorem [ thm : contact1 ] . + let @xmath18 be a legendrian knot in a contact manifold @xmath72 , such that there is no non - trivial contact dehn surgery along @xmath18 resulting again in @xmath72 . then the equivalence type of @xmath18 is determined by the contactomorphism type of its exterior . but the other implication does not generalize . the problem is again that the new glued - in solid torus is in general not a standard neighborhood of a legendrian knot as explained in section [ counterexample ] . so it is not directly clear if exotic cosmetic contact surgeries leads to counterexamples to the legendrian knot exterior problem . but their existence is interesting in its own . the following example was obtained earlier by geiges and onaran ( see also @xcite ) . [ exotic contact surgeries ] + consider the two different contact dehn surgeries along the legendrian unknot @xmath13 with @xmath96 in @xmath2 with respect to the contact surgery coefficients @xmath272 and @xmath273 ( see figure [ fig : lensspace ] ) . by expressing the surgery coefficients with respect to the surface longitude @xmath39 given by the meridian of the complementary solid torus one sees that the resulting manifolds are topologically the manifolds from the example from the foregoing section , so the homeomorphic lens spaces @xmath230 and @xmath231 . the next thing i want to show is that the two contact surgeries are unique and represent the same contact manifold . therefore one uses the same approach as in the example in section [ section : contactsurgery ] . the same argument as in that example shows that the two surgery diagrams both represent a unique contact structure @xmath91 on @xmath230 . the from the algorithm ( * ? ? ? * section 1 ) resulting @xmath217-contact surgery diagrams are shown in figure [ fig : lensspace ] in the middle ( for both contact surgeries is only one of the two possible stabilizations drawn ) . both surgery diagrams contain only @xmath105-contact surgeries , so the resulting contact structures have to be tight . but the classification of tight contact structures on lens spaces ( * ? ? ? * theorem 2.1 ) says that on @xmath230 there are two non - contactomorphic tight contact structures , so in this case this is not enough to conclude that these surgery diagrams represent the same contact manifold . to show that the contact structures of the two different surgeries are really contactomorphic one changes the contact dehn surgery diagrams into compatible open book decompositions ( for details see for example @xcite ) . the resulting open books are shown in figure [ fig : lensspace ] on the right . all colored curves represent right - handed dehn twists , the blue ones correspond to the monodromy of the open book decomposition of @xmath2 and the red and orange curves represent the dehn twists corresponding to the same colored legendrian links . all dehn twists together represent the monodromy of @xmath274 . by interchanging the holes , these two open books are the same , and therefore represent the same contact manifolds . these cosmetic contact surgeries are exotic because the corresponding topological cosmetic surgeries are exotic ( i.e. there is no homeomorphism of the exterior of the unknot that maps one slope to the other ) . with exactly the same methods one can get many other examples of this kind . for example , by contact dehn surgery along @xmath13 with contact surgery coefficients @xmath275 and @xmath276 one gets contactomorphic contact structures on the lens spaces @xmath277 and @xmath278 . an easy consequence of the contact dehn surgery theorem [ thm : contact2 ] is that the only possible candidates for legendrian knots in @xmath2 admitting a ( non - trivial ) contact surgery resulting again in @xmath2 are the legendrian unknots of type @xmath107 . that cosmetic contact surgeries along such knots really exists was shown in the example in section [ section : contactsurgery ] . it is also possible to show that every legendrian unknot of type @xmath107 admits ( infinitely many ) cosmetic contact surgeries resulting again in @xmath2 . by computing explicitly the @xmath279-invariants ( see @xcite ) of all other contact surgeries one can show that these are the only cosmetic contact surgeries from @xmath2 to @xmath2 . moreover , in section [ counterexample ] i showed that all topological examples of links not determined by their complements , where one does a composition of @xmath105-rolfsen twist along unknot components , works also for the contact case . but in the topological category there are also examples that do not arise in this way . teragaito @xcite and ichihara @xcite construct , building up on unpublished work of berge , links in @xmath1 with no unknot components , such that non - trivial dehn surgery along that link leads again to @xmath1 . similar to the arguments in theorem [ thm : cosmetic ] this leads to links not determined by their complements . gordon proves in @xcite that for every such link ( not determined by its complement , but without unknot components ) there exists only finitely many other links with homeomorphic complements . finally in @xcite the reverse question is considered . they study links that * are * determined by their complements . i would like to thank sebastian durst , hansjrg geiges , mirko klukas and sinem onaran for reading earlier versions of this article and for many useful discussions and suggestions . part of this work was supported by the dfg graduiertenkolleg 1269 `` global structures in geometry and analysis '' . r. avdek , contact surgery and supporting open books , _ algebr . _ * 13 * ( 2013 ) , 1613 - 1660 . k. baker and j. grigsby , grid diagrams and legendrian lens space links , _ j. symplectic geom . _ * 7 * ( 2009 ) , 415 - 448 . j. berge , the knots in @xmath280 which have nontrivial dehn surgeries that yield @xmath280 , _ topology appl . _ * 38 * ( 1991 ) , 1 - 19 . s. bleiler , c. hodgson and j. weeks , cosmetic surgery on knots , in _ proceedings of the kirbyfest _ , 23 - 34 , geometry and topology monographs * 2 * , coventry ( 1999 ) . j. conway , transverse surgery on knots in contact 3-manifolds , preprint available on : + http://arxiv.org/abs/1409.7077 f. ding and h. geiges , symplectic fillability of tight contact structures on torus bundles , _ algebr . _ * 1 * ( 2001 ) , 153 - 172 . f. ding and h. geiges , a legendrian surgery presentation of contact @xmath0-manifolds , _ math . cambridge philos . * 136 * ( 2004 ) , 583 - 598 . f. ding and h. geiges , the diffeotopy group of @xmath104 via contact topology , _ compos . _ * 146 * ( 2010 ) , 1096 - 1112 . f. ding , h. geiges and a. i. stipsicz , surgery diagrams for contact @xmath0-manifolds , _ turkish j. math . _ * 28 * ( 2004 ) , 41 - 74 . s. durst and m. kegel , computing rotation and self - linking numbers , preprint s. durst , m. kegel and m. klukas , computing the thurston - bennequin invariant in open books , preprint j. edwards , concentricity in @xmath0-manifolds , _ trans . * 113 * ( 1964 ) , 406 - 423 . y. eliashberg and m. fraser , topologically trivial legendrian knots , _ j. symplectic geom . _ * 7 * ( 2009 ) , 77 - 127 . j. etnyre , convex surfaces in contact geometry : class notes , lecture notes available on : http://people.math.gatech.edu/~etnyre/preprints/papers/surfaces.pdf j. etnyre and k. honda , knots and contact geometry i : torus knots and the figure eight knot , _ j. symplectic geom . _ * 1 * ( 2001 ) , 63 - 120 . j. etnyre , legendrian and transversal knots , in : _ handbook of knot theory _ ( w. menasco and m. thistlethwaite , eds . ) , elsevier , amsterdam ( 2005 ) , 105 - 185 d. gabai , foliations and the topology of @xmath0-manifolds ii , _ j. differential geom . _ * 26 * ( 1987 ) , 461 - 478 . d. gabai , surgery on knots in solid tori , _ topology _ * 28 * ( 1989 ) , 1 - 6 . f. gainullin , heegaard floer homology and knots determined by their complements , preprint available on : http://arxiv.org/abs/1504.06180 h. geiges , _ an introduction to contact topology _ , cambridge stud . adv . math . * 109 * ( cambridge university press , 2008 ) . h. geiges and s. onaran , legendrian rational unknots in lens spaces , _ j. symplectic geom . _ * 13 * ( 2015 ) , 17 - 50 . r. e. gompf and a. i. stipsicz , _ @xmath281-manifolds and kirby calculus _ , * 20 * ( american mathematical society , providence , ri , 1999 ) . c. gordon , links and their complements , in : `` topology and geometry : commemorating sistag '' , 71 - 82 , contemp . math . * 314 * ( american mathematical society , 2002 ) . c. gordon and j. luecke , knots are determined by their complements , _ * 2 * ( 1989 ) , 371 - 415 . k. honda , on the classification of tight contact structures i , _ geom . * 4 * ( 2000 ) , 309 - 368 . k. ichihara , on tunnel number one links with surgeries yielding the @xmath0-sphere , _ osaka j. math . _ * 47 * , number 1 ( 2010 ) , 189 - 208 . k. ichihara and i. jong , cosmetic banding on knots and links , preprint available on : + http://arxiv.org/abs/1602.01542 m. kegel , kontakt - dehn - chirurgie entlang legendre - knoten , bachelorarbeit , universitt zu kln ( 2011 ) , available online : http://www.mi.uni-koeln.de/~mkegel/downloads.html p. lisca , p. ozsvth , a. i. stipsicz and z. szab , heegaard floer invariants of legendrian knots in contact three - manifolds , _ j. eur . math . soc . _ ( jems ) * 11 * ( 2009 ) , 1307 - 1363 . y. mathieu , closed @xmath0-manifolds unchanged by dehn surgery , _ j. knot theory ramifications _ * 1 * ( 1992 ) , 279 - 296 . d. matignon , on the knot complement problem for non - hyperbolic knots , _ topology appl . _ * 157 * ( 2010 ) , 1900 - 1925 . h. matsuda , m. ozawa and k. shimokawa on non - simple reflexive links , _ j. knot theory ramifications _ * 11 * ( 2002 ) , 787 - 791 . b. zbaci an open book decomposition compatible with rational contact surgery , _ proceedings of gkova geometry - topology conference _ ( 2005 ) , 175 - 186 . v. v. prasolov and a. b. sossinsky , _ knots , links , braids and @xmath0-manifolds _ , transl . * 154 * ( american mathematical society , providence , 1997 ) . h. ravelomanana , exceptional cosmetic surgeries on @xmath1 , preprint available on : + http://arxiv.org/abs/1505.00238 h. ravelomanana , knot complement problem for @xmath10-space @xmath282 , preprint available on : http://arxiv.org/abs/1505.00239 d. rolfsen , _ knots and links _ , mathematics lecture series * 7 * ( publish or perish , berkeley , 1976 ) . y. rong , some knots not determined by their complements , in : `` quantum topology '' , ser . knots everything * 3 * , ( world sci . publishing , river edge , nj , 1993 ) , 339 - 353 . m. teragaito , links with surgery yielding the @xmath0-sphere , _ j. knot theory ramifications _ * 11 * ( 2002 ) , 105 - 108 . h. tietze , ber die topologischen invarianten mehrdimensionaler mannigfaltigkeiten , _ monatshefte fr mathematik und physik _ * 19 * ( 1908 ) , 1 - 118 . j. wang , cosmetic surgeries on genus one knots , _ algebr . _ * 6 * ( 2006 ) , 1491 - 1517 . j. h. c. whitehead , on doubled knots , _ j. london math . * 12 * ( 1937 ) , 63 - 71 .

collinder 419 is an open cluster in the constellation cygnus . it was first designated as barnard 794 ( barnard 1927 ) , but the accepted name comes from the work of collinder ( 1931 ) who classified the cluster as a @xmath3 normae cluster . this is an intermediate type of grouping between multiple star systems and open clusters . the prototype cluster ngc 6169 contains the bright star @xmath3 normae = hd 149038 and a host of surrounding fainter stars . in the case of collinder 419 , the bright central star is hd 193322 = hr 7767 = hip 100069 = wds 20181 + 4044 . hd 193322 is a multiple star system with a close visual companion stf 2666 ab ( separation of @xmath4 and probably orbitally bound ) and two more distant companions , stf 2666 ac ( @xmath5 ) and tar 5 ad ( @xmath6 ) , which , if bound have very long orbital periods . mcalister et al . ( 1987 ) discovered that the a component is a speckle binary , chr 96 aa , ab . an approximately 31 year orbital period for this pair was computed by hartkopf et al . in addition , one of the components of the speckle binary is a spectroscopic binary with a 311.03@xmath70.25 day period ( fullerton 1990 ; mckibben et al . fullerton ( 1990 ) also obtained a spectrum of the b component that suggests that it may also be a spectroscopic binary . thus , the central ab pair may contain as many as five individual stars . the primary of the system , aa1 , has a spectral type of o9 v:((n ) ) ( walborn 1972 ) , with the suffix indicating slightly broadened lines . mckibben et al . ( 1998 ) tentatively identify the close spectroscopic companion aa2 and the b component all as b - type stars . the magnitude difference of the speckle pair aa , ab is @xmath8 ( mason et al . 2009 ) which suggests that ab is also an early b - type star . both the aa2 and ab companions offer us the means to determine mass for the o star component aa1 through measurement of the orbital motion . by themselves , neither an orbit of the aa , ab pair nor the spectroscopic aa1,aa2 pair will yield a mass , but , in conjunction with a distance determination , a mass can be determined . the _ hipparcos _ catalogue ( perryman et al . 1997 ) has become the standard source for parallaxes , but it may be unreliable for some double stars ( shatskii & tokovinin 1998 ) and also for distant o stars ( schrder et al . 2004 ) . until a new space - based parallax engine is available sometime in the next decade we are left with classical techniques as the best way to determine o - star distances . we are currently pursuing efforts to determine the orbits of the central triple through speckle and optical long baseline interferometry ( turner et al . 2008 ) . here we present a new assessment of the distance to the cluster that we will use in the forthcoming orbit and mass determination . in 2 we present spectral classifications for the brighter members , and we show fits of their spectral energy distributions that provide estimates of reddening and angular size . we then describe in 3 an astrometric and photometric study of cluster membership based upon data in the ucac3 catalog ( zacharias et al . 2010 ) and the 2mass catalog ( cutri et al . we also present new @xmath0 photometry of the fainter cluster members in 4 . our results are summarized in 5 . we obtained spectra for five of the brighter stars in the cluster in order to determine spectral classifications and to estimate the intrinsic colors of the stars . one of us ( mvm ) obtained spectra of the four blue stars of the central multiple system using the kitt peak national observatory ( kpno ) 2.1 m telescope and cassegrain focus , goldcam ccd spectrograph . these spectra were made during 2005 october and november with the no . 47 grating ( 831 grooves mm@xmath9 ) in second order to obtain a resolving power @xmath10 . the observed wavelength range was 4050 4950 , a region that includes numerous h balmer , , and lines in o- and b - type spectra . the data were reduced using standard routines in iraf . the flux rectified spectra were compared directly to spectra from the atlas of valdes et al . ( 2004 ) in order to estimate the spectral classification based upon the strengths of the h balmer , , , and lines . our results are given in table 1 for the b , c , and d components of hd 193322 . see figures [ fig_hd193322b ] [ fig_hd193322d ] for the comparision between the target stars and the template spectra . our spectra for the combined flux of aa1 , aa2 , and ab agree with the o9 v:((n ) ) classification from walborn ( 1972 ) , which is listed in table 1 . these classifications all agree within a subtype with earlier results ( burnichon 1975 ; hoffleit 1982 ; abt & cardona 1983 ) . the very red star iras 20161 + 4035 is the brightest object in the @xmath11-band in this vicinity , and we were curious to determine if indeed it is related to the cluster . one of us ( edg ) obtained three spectra of the object with the kpno 0.9 m coude feed telescope in 2004 october . these spectra cover the range from 6460 to 7140 with a resolving power @xmath12 . the average spectrum is compared to several other cool star spectra from valdes et al . ( 2004 ) ( also made with the kpno coude feed telescope ) in figure [ red_spectra ] . we made a digital comparison with all the m2 m5 spectra in the valdes et al . spectral atlas , and the best fit ( based mainly on the appearance of the strong tio bands in this spectral region ) was the spectrum of the m3 ii star hd 40239 ( although there were a number of other m3 iii spectra which made an almost equally good fit ) . figure [ red_spectra ] shows the good overall match made with the m2 ii and m3 iii giants compared to the dwarf and supergiant spectra . we also measured the h@xmath13 equivalent width in these and similar stars to help establish the luminosity class ( eaton 1995 ) . we found that among stars of similar spectral type that the mean equivalent width of h@xmath13 ranged from @xmath14 for 2 main sequence stars , to @xmath15 for 4 giant stars , and up to @xmath16 for 3 supergiant stars . the measured h@xmath13 equivalent width of iras 20161 + 4035 , 1.38 , clearly places the star among those in the giant luminosity class , and we adopted this assignment in table 1 . note that the spectrum of iras 20161 + 4035 shows the strongest @xmath17 feature among the spectra illustrated in figure [ red_spectra ] . we can use these spectral classifications to estimate the intrinsic colors of the targets , compare these with observed multi - color observations , and find the reddening @xmath18 in each case . however , the available near - ir 2mass photometry for hd 193322 corresponds to the total flux for the entire aa1 , aa2 , ab , and b complex . thus , we need to make some assumptions about the relative flux contributions of each component in each filter band to derive magnitudes and fluxes for the individual stars . the magnitude differences of the ab pair were determined in the johnson @xmath19 bands through adaptive optics imaging by ten brummelaar et al . unfortunately , much less is known about the a sub - components . mason et al . ( 2009 ) found a magnitude difference of @xmath8 mag for the aa , ab speckle pair , and mckibben et al . ( 1998 ) used a statistical method to estimate the mass ratio and magnitude difference @xmath20 mag for the aa1,aa2 pair in the spectroscopic binary . these magnitude differences ( plus @xmath21 for the ab pair from ten brummelaar et al . 2000 ) lead to @xmath22-band flux ratios relative to star aa1 that are given in table 2 . to derive the flux ratios in other bands , we estimated their colors according to their spectral classification and the color calibration from wegner ( 1994 ) . we assumed that the aa2 and ab stars are main sequence objects and then estimated their spectral types ( b1.5 v and b1 v , respectively ) based on their magnitude differences relative to aa1 ( using the absolute magnitude versus spectral type relation for main sequence stars from lesh 1979 ) . table 2 lists the adopted flux ratio estimates for the sub - components ( which agree with those from ten brummelaar et al . for the johnson @xmath23 bands , but not for their @xmath24-band result for which component b is unusually faint in @xmath24 relative to the other bands ) . we used these flux ratios to determine the magnitudes of aa1 and b , the stars with reliable spectral classifications , in the following analysis . note that the current flux ratio uncertainties probably introduce an error of @xmath25 mag into the error budget for the magnitudes of these two stars . we estimated the reddening and angular sizes of the five bright stars listed in table 1 by comparing their observed fluxes with reddened model spectral energy distributions . the optical magnitudes were taken from the work of ten brummelaar et al . ( 2000 ; for components aa1 and b ) , burnichon & garnier ( 1976 ; for components c and d ) and droege et al . ( 2006 ; plus @xmath0 magnitudes given in 4 for iras 20161 + 4035 ) , and these were combined with 2mass magnitudes for the near - ir ( cutri et al . 2003 ) . the magnitudes were transformed to fluxes using calibrations from colina et al . ( 1996 ; johnson @xmath26 ) , bessell et al . ( 1998 ; cousins @xmath27 ) , and cohen et al . ( 2003 ; 2mass @xmath28 ) . we used kurucz model atmospheres ( assuming @xmath29 , solar abundances , and a microturbulent velocity of 2 km s@xmath9 ) for temperatures from the spectral calibration of bhm - vitense ( 1981 ; see table 1 ) to create low resolution versions of the unreddened spectral energy distributions ( seds ) . however , for the case of iras 20161 + 4035 ( m3 iii ) , we adopted a flux distribution from the marcs code ( gustafsson et al . 2008 ) for a model with @xmath30 k , @xmath31 , solar abundances , and a microturbulent velocity of 5 km s@xmath9 . we then fit the observed fluxes with model seds reddened according to the formulation from fitzpatrick ( 1999 ) . because the stars have low reddening and limited short - wavelength coverage , it is difficult to obtain an independent estimate of the ratio of total - to - selective extinction @xmath32 , so we adopted the default value @xmath33 throughout . the derived reddening @xmath18 and limb - darkened angular diameter @xmath34 are given in table 1 . the blue stars yield a consistent reddening of @xmath35 mag , which is significantly less than that found for iras 20161 + 4035 , @xmath36 mag . the sed for this red star is shown in figure [ fig_sed ] , and we see that there appears to be an ir flux - excess of @xmath37 near @xmath38 m , based upon data from the msx ( egan & price 1996 ) and iras missions . we argue below ( 5 ) that the higher reddening and ir - excess may be related to the youth of this star . with the recent release of the third u. s. naval observatory ccd astrograph catalog ( ucac3 ; zacharias et al . 2010 ) , we now have access to a large collection of astrometric and photometric data that can be used to search for additional cluster members . we began by selecting all the stars in ucac3 with complete sets of proper motion and photometric data that are found within a @xmath39 radius of hd 193322 , the nominal center of cr 419 . we show in figure [ fig_stellar_density ] the average surface density of star counts within annuli centered on the position of hd 193322 . the larger surface density associated with the cluster extends outwards to at least @xmath40 , and we set the outer boundary on cluster membership at @xmath41 to avoid including too many field stars at large distance . figure [ dss_image ] shows a @xmath39 field centered on hd 193322 and shows the low stellar density of the cluster . the image is from the near - ir plates of second palomar observatory sky survey ( poss2-ir ) @xcite . figure [ fig_proper_motion ] shows the distribution of proper motions of stars found within this radius . the circle in the diagram is centered on @xmath42 mas y@xmath9 and @xmath43 mas y@xmath9 , the average values for the bright cluster stars ( hd 193322a , c , d , and hd 228810 ) . there appears to be a bimodal distribution in proper motion in this direction , so we selected only those stars within the circle , i.e. , those with a combined proper motion difference from the cluster mean of less than 11 mas y@xmath9 ( the rms of the ucac3 proper motion errors and the expected range within the cluster itself ) . the constraints of position and proper motion reduced the potential sample to a group of 96 stars . the histogram in figure [ fig_stellar_density ] also demonstrates that there will be a significant number of field stars in this sample . to estimate this number , we applied the same proper motion constraint to the ucac3 sample within a separation range of @xmath41 to @xmath44 , and this indicates that the surface density of field stars in this direction with similar proper motions to the bright cluster stars is @xmath45 stars per @xmath46 . thus , we might expect to find some @xmath47 field stars in the @xmath41 radius circle centered on hd 193322 . this exercise suggests that the actual number of cluster stars is small ( perhaps @xmath48 total - @xmath49 field @xmath50 cluster stars ) and that the sample will be dominated by other field stars . we next considered the photometric data in ucac3 , which consists of 2mass photometry and the ucac3 red - band photometry that we will refer to as @xmath51 ( to distinguish it from the johnson and cousins counterparts ) . this filter has a bandpass from 579 to 642 nm that falls between the johnson @xmath22 and cousins @xmath52 bands . the ucac3 catalog reports magnitudes from centroid fit models and aperture photometry , and we used the latter in our analysis . the ucac3 listing for hd 193322 corresponds to the total flux from components aa1 , aa2 , ab , and b , and we used the flux ratios given in table 2 to determine the corresponding magnitudes for stars aa1 and b only . our goal is to compare the magnitudes of the candidate member stars with model predictions for given assumptions about the reddening @xmath18 and distance modulus dm , in order to find values of these parameters that lead to the best fit of the photometry . we adopted model isochrones for the cluster from the evolutionary models of bertelli et al . ( 1994 ) and marigo et al . we selected an isochrone for solar metallicity and a cluster age of 7 myr , the value proposed by kharchenko et al . ( 2005 ) and one that is consistent for a time of flight from cr 419 of the runaway star 68 cyg ( schilbach & rser 2008 ) . the isochrones were produced for a number of photometric bands including johnson @xmath22 , cousins @xmath52 , and 2mass @xmath28 . we need to transform the model optical magnitudes into the @xmath51 system in order to compare them with the ucac3 values . we used a sample of 402 photometric standard stars from landolt ( 1992 ) that appear in ucac3 to arrive at a transformation equation from johnson @xmath22 and cousins @xmath52 to ucac3 @xmath51 , @xmath53 which successfully fits the observed ucac3 @xmath51 magnitudes over the range @xmath54 with a standard deviation of 0.116 mag . the color term in the equation shows how the @xmath51 magnitude changes with stellar color because of the difference in the effective wavelengths of the @xmath51 and @xmath52 bands . the final term on the right hand side reflects a small systematic difference between the landolt and ucac3 systems that is approximately equal to the scatter in the relation at the bright and faint ends of the sample . we can apply a few other constraints to limit the number of field stars that may exist within the astrometric group of cluster candidates . figure [ fig_ir_colors ] shows the near - ir color - color diagram for the sample together with unreddened ( _ solid line _ ) and reddened ( _ dashed line _ ) versions of the cluster isochrone . the numbers beside the unreddened version indicate the stellar mass corresponding to the color pair , and massive stars have colors in the lower part of the diagram . the reddening vector direction is indicated by the dotted line . comern et al . ( 2002 ) and others have argued that the reddening - free parameter @xmath55 ( fitzpatrick 1999 ; indebetouw et al . 2005 ) provides an excellent means to identify massive , reddened stars that typically have @xmath56 . we decided to apply this criterion by restricting the sample to those massive stars with @xmath57 , the model value for a star of mass @xmath58 . imposing this limit will mean that we lose fainter , red cluster members , but , on the other hand , we will also remove cool background giants and cooler , low mass , foreground stars that do not belong to the cluster ( for example , those stars found near the unreddened isochrone between 0.8 and @xmath59 ) . we also applied a faint limit constraint , @xmath60 , to avoid stars with large magnitude errors . since our initial sample was drawn from ucac3 stars with complete photometric coverage , the result is that this sample tends towards stars with equal magnitude in all filters at the faint end . this group probably includes foreground a - type stars , so applying the faint limit will help us remove such stars from consideration . we show two versions of the color magnitude diagram in figures [ fig_ucac3_color_mag ] & [ fig_2mass_color_mag ] . both of these show evidence of a main sequence group at the blue end , but both also show a large population of redder stars that lie above the expected main sequence . in order to remove such targets , we restricted the sample to those with @xmath61 . this constraint , together with the @xmath62 limit described above , will result in a sample without any low mass , main sequence members . our final constraint is based on how closely the magnitudes match the model isochrone values . for any given choice of @xmath18 and dm , we formed an average @xmath63 statistic between all the observed and model magnitudes ( reddened according to the relations given by fitzpatrick 1999 ) for each mass point along the model isochrone . the best fit ( lowest @xmath63 ) model point was taken as an appropriate mass for the specific star . those stars with magnitudes and colors very different from the main sequence values will have a large minimum @xmath63 , and we restricted our final sample of main sequence cluster members to those with @xmath64 , a value high enough to include those cases where the observational photometric errors are underestimated . the final sample consists of 16 stars that are listed in table 3 and are indicated in figures [ fig_ir_colors ] [ fig_2mass_color_mag ] by plus signs . note that we make the tacit assumption that the same reddening can be applied to all cluster members , and this can only be true in some average sense . there may be a reddening gradient across this field ( schlegel et al . 1998 ) and some of the youngest objects may suffer from circumstellar dust reddening . nevertheless , until spectroscopy is available for many stars and their intrinsic colors found , we must accept this working hypothesis . we performed a grid search over a range of @xmath18 and dm to find those values that minimized the average @xmath63 value for the selected cluster stars . the best fit was made with @xmath1 mag and dm @xmath65 mag , yielding an average photometric statistic of @xmath66 . the isochrones for these parameters are plotted in figures [ fig_ir_colors ] [ fig_2mass_color_mag ] , and the corresponding main sequence masses from the isochrone are listed in the final column of table 3 . our derived reddening estimate is mainly consistent with earlier estimates for the cluster and star ( @xmath67 , kharchenko et al . 2005 ; @xmath68 , burnichon 1975 ; @xmath69 , schrder et al . 2004 ; @xmath70 , cardelli et al . the reddening is slightly larger than that found by sed fitting for the blue stars ( table 1 ) , but we suspect that the differences arise in the variable reddening across the cluster . we also repeated the analysis using an isochrone for a younger cluster ( age 3 myr ) , but the results were unchanged because the predicted colors and magnitudes were only significantly different for the most massive star hd 193322aa1 ( the predicted colors are marginally closer to the observed ones for the 7 myr isochrone ) . there are many stars that are plotted above the main sequence in figures [ fig_ucac3_color_mag ] & [ fig_2mass_color_mag ] . some of these stars may belong to a massive and more distant population . for example , in the infrared color - color diagram ( fig . [ fig_ir_colors ] ) , there appears to be a reddened group of ob stars near @xmath71 and @xmath72 . an independent analysis of the 2mass magnitudes of stars in the vicinity of cr 419 by david turner ( priv . communication ) suggests that these correspond to objects with @xmath73 at a distance of 1.6 kpc that may be associated with the nearby cluster berkeley 87 @xcite . we show how the 7 myr isochrone would appear for such reddened and distant stars as a dotted line in figures [ fig_ucac3_color_mag ] and [ fig_2mass_color_mag ] , and many of the stars appearing above the cr 419 main sequence are close to the predicted colors and magnitudes of a distant population . cr 419 is probably young enough that it may host pre - main sequence stars that would also appear above the main sequence . dashed lines in figure [ fig_ucac3_color_mag ] and [ fig_2mass_color_mag ] indicate the pre - main sequence isochrone for an age of 7 myr from the work of @xcite transformed to the distance and reddening of cr 419 , and indeed we find many examples of stars close to the predicted track . we suspect that redder stars appearing in these color - magnitude diagrams probably include some cluster pre - main sequence stars and large numbers of distant massive stars and other line - of - sight field stars . we also obtained new @xmath0 photometry of the central region of the cluster to explore the optical colors of the fainter stars . one of us ( jrp ) collected ccd images of collinder 419 in 2006 october with the emory university observatory , dfm 0.6 m cassegrain telescope . the detector was an apogee ap47 @xmath74 pixel ccd camera . in order to concentrate the signal for fainter sources , the camera was set to a @xmath75 binning mode with a net pixel scale of @xmath76 pixel@xmath9 and a @xmath77 field of view . the images of the brightest stars were saturated in individual frames in order to obtain a stronger signal on the faintest sources . thirty frames were collected through each johnson @xmath0 filter . in addition , standard stars from charts 140 , 141 , and 148 of landolt ( 1992 ) were observed over a large range in air mass to transform the instrumental magnitudes to the standard system . all the frames were debiased , dark subtracted , and flat fielded using standard routines in iraf . the images were combined using imcentroid and imcombine to build a master image in each filter . photometric measurements of the standard stars were made with aperture photometry routines in iraf . the transformation coefficients were computed using the idl procedure transf1.pro , written by marc w. buie . this procedure uses the general formula @xmath78 where @xmath79 is the standard magnitude , @xmath80 is the instrumental magnitude , @xmath81 is the first order extinction coefficient , @xmath82 is the airmass , @xmath83 is the second order extinction coefficient , @xmath84 is a color index , @xmath85 is a color coefficient , and @xmath86 is the magnitude zero point . the instrumental magnitudes of the stars in the collinder 419 frames were measured using psf fitting routines in iraf , and the relation between the psf and aperture measurements was determined for ten well isolated stars in the field . the transformation equations were then used to determine johnson magnitudes for the cluster stars . the errors in the transformation coefficients were added in quadrature with the instrumental magnitude errors to obtain the net observational errors ( approximately 0.04 , 0.03 , and 0.02 mag in @xmath0 , respectively ) . the results are listed in table 4 . we show in figure [ fig_optical_color_mag ] the color magnitude diagram @xmath87 for 43 of the 79 stars measured that meet the proper motion limits for cluster membership described above ( 3 ) . most of the proposed cluster members ( 3 ; table 3 ) were too bright and/or outside of the field of view for our @xmath0 photometric measurements , and we have @xmath0 photometry for only 5 of the 16 proposed members ( indicated by plus signs in fig . 8 and by `` yes '' in the final column of table 4 ) . figure [ fig_optical_color_mag ] also shows the main sequence relation from the models of bertelli et al . ( 1994 ) and marigo et al . ( 2008 ) for our derived estimates of @xmath88 mag and dm = 9.35 mag , and this relation appears to be consistent with our optical photometry . there are a number of stars that have magnitudes indicative of lower mass , main sequence stars , but there are many objects with a position above the main sequence with a distribution similar to that seen in the red and near - infrared color magnitude diagrams ( figs . [ fig_ucac3_color_mag ] and [ fig_2mass_color_mag ] ) . figure [ fig_color_color ] shows the optical color - color diagram @xmath89 for the subset with complete @xmath0 measurements and within the proper motion limits for membership . we also show both unreddened and reddened versions of the model colors . again , there is evidence for some lower main sequence members ( towards the middle and lower sections of the dashed line ) , but there is also a group of stars with colors near @xmath90 that are characteristic of b - stars with a reddening @xmath91 mag . most of these stars also have positions in the near - ir color color diagram associated with reddened , intermediate mass stars ( fig . [ fig_ir_colors ] ; @xmath92 ) . the distance we obtained from the photometric analysis of @xmath93 pc ( formal error ) does not account for possible systematic errors in the model isochrone magnitudes . by comparing the @xmath94 color - magnitude diagrams for a 7 myr cluster from marigo et al . ( 2008 ) with that from lejeune & schaerer ( 2001 ) , we estimate that the model uncertainties amount to @xmath95 mag over the main sequence range of interest , so we add this error in quadrature with the fitting errors to obtain @xmath96 pc . we can check this with distances derived from the angular sizes given in table 1 . this distance estimate depends on the ratio of the physical diameter to the angular diameter , and for this comparison , we estimated the stellar radii for hd 193322b , c , and d from the model isochrone values for main sequence stars with our adopted effective temperatures ( table 2 ) . we could not use this relation for hd 193322aa1 because the model isochrone did not extend to stars quite this hot , so we instead adopted a radius of @xmath97 from the spectral calibration for main sequence stars from martins et al . we estimated a radius of @xmath98 for the m3 iii star iras 20161 + 4035 from the calibration for giants by van belle et al . the resulting distances are listed in the final column of table 1 , and the average , @xmath99 pc , agrees with the photometric result . our results confirm the only previous distance estimate for the cluster from kharchenko et al . ( 2005 ) of 740 pc ( no error quoted ) . the cluster distance also agrees within errors with the _ hipparcos _ distance for hd 193322 of @xmath100 pc ( van leeuwen 2007 ) . we suspect that some of the red stars that meet the spatial and proper motion constraints of cluster membership but lie above the main sequence are very young objects . a similar population of bright , red objects has been reported for other young star clusters containing o stars ( kumar et al . 2004 ; comern & pasquali 2005 ; carlson et al . 2007 ; negueuela et al . 2007 ; comern et al . these stars may include ir - excess objects with remnant disks like the herbig ae / be stars . although it is possible that the red object , iras 20161 + 4035 , is an old star that is coincidentally passing through the vicinity of cr 419 , we suspect that it may also be a very young member of the cluster , despite its evolved appearance . the star s large reddening and ir - excess both suggest that it is immersed in natal dust , and its strong @xmath17 and high luminosity may be indicative of youth . kumar et al . ( 2004 ) identified three other cases of luminous , late - type giants that are found in young clusters . these cool giants may represent another kind of spectroscopic manifestation of young stellar objects . we thank todd henry , adric riedel , and russel white for helpful comments about m - star spectra , and we also thank bill binkert of the kpno staff for his assistance at the kpno coude feed telescope . we are grateful to david turner for discussions about field contamination by distant massive stars . this publication made use of data products from the two micron all sky survey , which is a joint project of the university of massachusetts and the infrared processing and analysis center / california institute of technology , funded by the national aeronautics and space administration and the national science foundation . it also used images from the second palomar observatory sky survey ( poss - ii ) , which was made by the california institute of technology with funds from the national science foundation , the national geographic society , the sloan foundation , the samuel oschin foundation , and the eastman kodak corporation . we also made use of the washington double star catalog , maintained at the u. s. naval observatory , and the simbad database , operated by the cds in strasbourg , france . this material is based upon work supported by the national science foundation under grant no . institutional support has been provided from the gsu college of arts and sciences and from the research program enhancement fund of the board of regents of the university system of georgia , administered through the gsu office of the vice president for research . a portion of the research was carried out at the jet propulsion laboratory , california institute of technology , under a contract with the national aeronautics and space administration . abt , h. a. , & cardona , o. 1983 , , 272 , 182 barnard , e. e. 1927 , a photographic atlas of selected regions of the milky way , ed . e. b. frost & m. r. calvert ( carnegie inst . washington publ . 247 ) ( washington dc : carnegie inst . ) bertelli , g. , bressan , a. , chiosi , c. , fagotto , f. , & nasi , e. 1994 , , 106 , 275 bessell , m. s. , castelli , f. , & plez , b. 1998 , , 333 , 231 bhm - vitense , e. 1981 , , 19 , 295 burnichon , m. l. 1975 , , 45 , 383 burnichon , m. l. , & garnier , r. 1976 , , 24 , 89 cardelli , j. a. , clayton , g. c. , & mathis , j. s. 1989 , , 345 , 245 carlson , l. r. , et al . 2007 , , 665 , l109 cohen , m. , wheaton , w. a. , & megeath , s. t. 2003 , , 126 , 1090 colina , l. , bohlin , r. , & castelli , f. 1996 , hst instrument science report cal / scs-008 ( baltimore : stsci ) collinder , p. 1931 , ann . obs . lund , 2 , 1 comern , f. , & pasquali , a. 2005 , , 430 , 541 comern , f. , pasquali , a. , figueras , f. , & torra , j. 2008 , , 486 , 453 comern , f. , et al . 2002 , , 389 , 874 cutri , r. , et al . 2003 , 2mass all - sky catalog of point sources ( pasadena : ipac / caltech ) droege , t. f. , richmond , m. w. , & sallman , m. 2006 , , 118 , 1666 eaton , j. a. 1995 , , 109 , 1797 egan , m. p. , & price , s. d. 1996 , , 112 , 2862 fitzpatrick , e. l. 1999 , , 111 , 63 fullerton , a. w. 1990 , ph.d . dissertation , univ . of toronto gustafsson , b. , edvardsson , b. , eriksson , k. , jrgensen , u. g. , nordlund , a. , & plez , b. 2008 , , 486 , 951 hartkopf , w. i. , gies , d. r. , mason , b. d. , bagnuolo , w. g. , jr . , & mcalister , h. a. 1993 , baas , 25 , 872 hoffleit , d. 1982 , the bright star catalogue , 4th ed . ( new haven : yale univ . indebetouw , r. , et al . 2005 , , 619 , 931 kharchenko , n. v. , piskunov , a. e. , rser , s. , schilbach , e. , & scholz , r .- d . 2005 , , 438 , 1163 kumar , b. , sagar , r. , sanwal , b. b. , & bessell , m. s. 2004 , , 353 , 991 landolt , a. u. 1992 , , 104 , 340 lejeune , t. , & schaerer , d. 2001 , , 366 , 538 lesh , j. r. 1979 , in spectral classification of the future ( iau coll . 47 ) , ricerche astronomiche , 9 , ed . m. f. mccarthy , a. d. g. philip , & g. v. coyne ( citta del vaticano : specola vaticana ) , 81 marigo , p. , girardi , l. , bressan , a. , groenewegen , m. a. t. , silva , l. , & granato , g. l. 2008 , , 482 , 883 martins , f. , schaerer , d. , & hillier , d. j. 2005 , , 436 , 1049 mason , b. d. , hartkopf , w. i. , gies , d. r. , henry , t. j. , & helsel , j. w. 2009 , aj , 137 , 3358 massey , p. , degioia - eastwood , k. , & waterhouse , e. 2001 , , 121 , 1050 mcalister , h. a. , hartkopf , w. i. , hutter , d. j. , shara , m. m. , & franz , o. g. 1987 , , 92 , 183 mckibben , w. p. , et al . 1998 , , 110 , 900 negueruela , i. , marco , a. , israel , g. l. , & bernabeu , g. 2007 , , 471 , 485 perryman , m. a. c. , et al . 1997 , , 323 , l49 reid , i.n . , 1991 , , 103 , 661 schilbach , e. , & rser , s. 2008 , , 489 , 105 schlegel , d. j. , finkbeiner , d. p. , & davis , m. 1998 , , 500 , 525 schrder , s. e. , kaper , l. , lamers , h. j. g. l. m. , & brown , a. g. a. 2004 , , 428 , 149 siess , l. , dufour , e. , & forestini , m. 2000 , , 358 , 593 shatskii , n. i. , & tokovinin , a. a. 1998 , astronomy letters , 24 , 673 ten brummelaar , t. , mason , b. d. , mcalister , h. a. , roberts , l. c. , jr . , turner , n. h. , hartkopf , w. i. , & bagnuolo , w. g. , jr . 2000 , , 119 , 2403 turner , d. g. , & forbes , d. 1982 , , 94 , 789 turner , n. h. , ten brummelaar , t. a. , & mason , b. d. 2008 , baas , 40 , 208 valdes , f. , gupta , r. , rose , j. a. , singh , h. p. , & bell , d. j. 2004 , , 152 , 251 van belle , g. t. , et al . 1999 , , 117 , 521 van leeuwen , f. 2007 , , 474 , 653 walborn , n. r. 1972 , , 77 , 312 walborn , n.r . , & fitzpatrick , e.l . 1990 , , 102 , 379 wegner , w. 1994 , , 270 , 229 zacharias , n. , et al . 2010 , aj , 139 , 2184 ) and declination ( @xmath101 ) for stars within @xmath41 of the cluster center . the position of the circle marks the cluster average proper motion and its circumference represents the boundary for possible cluster membership . [ fig_proper_motion],title="fig:",height=453 ] signs are used for the rest of the sample . the solid and dashed lines represent the values for the cluster isochrone for no reddening and @xmath88 , respectively . the numbers beside the unreddened isochrone indicate the stellar masses at the terminal points and at the origin of the reddening trajectory ( _ dotted line _ ) . stars below the dotted line are candidate massive stars ( @xmath102 ) . [ fig_ir_colors],title="fig:",height=453 ] and 2mass @xmath103 magnitudes . plus signs mark probable , main sequence members , while @xmath104 symbols are used for the remainder of the sample that meets the positional and proper motion criteria for cluster membership . the solid line shows the best fit isochrone . the dashed line indicates the pre - main sequence isochrone for an age of 7 myr . the dotted line is the isochrone for distant ob stars that may be related to the nearby cluster berkeley 87 . [ fig_ucac3_color_mag],title="fig:",height=453 ] lcccc hd 193322aa1 & o9 v:((n ) ) & @xmath105 & @xmath106 & 0.83 + hd 193322b & b1.5 v & @xmath107 & @xmath108 & 0.73 + hd 193322c & b8 v & @xmath109 & @xmath110 & 0.79 + hd 193322d & b9 v & @xmath111 & @xmath112 & 0.70 + iras 20161 + 4035 & m3 iii & @xmath113 & @xmath114 & 0.78 + lcccccccc hd 193322aa1 & 33.2 & 1.00 & 1.00 & 1.00 & 1.00 & 1.00 & 1.00 & 1.00 + hd 193322aa2 & 23.0 & 0.29 & 0.30 & 0.31 & 0.33 & 0.33 & 0.35 & 0.35 + hd 193322ab & 25.0 & 0.41 & 0.43 & 0.44 & 0.46 & 0.47 & 0.48 & 0.48 + hd 193322b & 23.0 & 0.21 & 0.21 & 0.21 & 0.21 & 0.23 & 0.24 & 0.25 + cccccccc 262@xmath115202014 & & 12.29 & 1.06 & 0.14 & 0.08 & 11.02 & 2.2 + 262@xmath115202033 & & 12.90 & 1.20 & 0.23 & 0.12 & 11.35 & 1.7 + 262@xmath115202070 & hd 228810 & 10.01 & 0.55 & 0.10 & 0.05 & 9.31 & 5.0 + 262@xmath115202090 & & 12.44 & 0.94 & 0.08 & 0.09 & 11.34 & 1.9 + 262@xmath115202099 & & 13.66 & 1.33 & 0.26 & 0.09 & 11.98 & 1.4 + 262@xmath115202103 & & 12.91 & 1.14 & 0.21 & 0.12 & 11.44 & 1.7 + 262@xmath115202123 & hd 193322c & 11.14 & 0.51 & 0.02 & @xmath1150.02 & 10.63 & 2.8 + 262@xmath115202132 & hd 193322d & 11.26 & 0.58 & 0.03 & 0.08 & 10.57 & 2.8 + 262@xmath115202133 & hd 193322aa1 & 6.59 & 0.07 & @xmath1150.07 & @xmath1150.04 & 6.64 & 14.9 + 262@xmath115202133 & hd 193322b & 8.29 & 0.17 & @xmath1150.03 & 0.01 & 8.14 & 9.0 + 262@xmath115202155 & & 12.27 & 0.71 & 0.07 & 0.02 & 11.46 & 1.8 + 262@xmath115202269 & & 12.25 & 0.72 & 0.06 & 0.07 & 11.40 & 1.8 + 262@xmath115202314 & & 12.83 & 1.09 & 0.23 & 0.06 & 11.45 & 1.7 + 262@xmath115202350 & & 12.37 & 0.88 & 0.06 & 0.10 & 11.34 & 1.9 + 262@xmath115202365 & & 11.63 & 0.74 & 0.07 & 0.05 & 10.77 & 2.5 + 262@xmath115202402 & & 11.67 & 0.78 & 0.11 & 0.02 & 10.76 & 2.5 + ccccc 262 - 201944 & 12.51@xmath70.04 & 12.68@xmath70.02 & 11.58@xmath70.01 & + 262 - 201945 & 14.26@xmath70.04 & 14.42@xmath70.02 & 13.59@xmath70.02 & + 262 - 201948 & & 15.92@xmath70.03 & 13.91@xmath70.02 & + 262 - 201953 & & 16.52@xmath70.04 & 15.21@xmath70.01 & + 262 - 201963 & 16.25@xmath70.05 & 16.09@xmath70.03 & 14.91@xmath70.02 & + 262 - 201973 & 16.22@xmath70.06 & 16.13@xmath70.04 & 15.01@xmath70.02 & + 262 - 201978 & 16.60@xmath70.06 & 16.11@xmath70.04 & 15.02@xmath70.02 & + 262 - 201981 & 15.30@xmath70.04 & 15.27@xmath70.02 & 14.43@xmath70.01 & + 262 - 201993 & 16.04@xmath70.05 & 15.76@xmath70.02 & 14.83@xmath70.02 & + 262 - 201997 & & 16.66@xmath70.03 & 15.53@xmath70.03 & + 262 - 202009 & & 17.15@xmath70.05 & 15.86@xmath70.01 & + 262 - 202014 & 12.90@xmath70.04 & 12.83@xmath70.01 & 12.41@xmath70.03 & yes + 262 - 202017 & 15.74@xmath70.04 & 15.23@xmath70.02 & 14.18@xmath70.01 & + 262 - 202018 & & 14.99@xmath70.04 & 14.67@xmath70.03 & + 262 - 202025 & & 16.47@xmath70.03 & 14.62@xmath70.01 & + 262 - 202026 & & 16.31@xmath70.03 & 15.22@xmath70.01 & + 262 - 202027 & & 16.49@xmath70.03 & 15.61@xmath70.01 & + 262 - 202032 & 14.72@xmath70.04 & 14.64@xmath70.01 & 13.74@xmath70.01 & + 262 - 202033 & 13.65@xmath70.04 & 13.59@xmath70.01 & 13.08@xmath70.02 & yes + 262 - 202034 & & 16.12@xmath70.02 & 14.04@xmath70.02 & + 262 - 202039 & 15.17@xmath70.04 & 14.92@xmath70.01 & 14.08@xmath70.01 & + 262 - 202042 & & 15.95@xmath70.02 & 15.32@xmath70.02 & + 262 - 202045 & 14.40@xmath70.04 & 14.20@xmath70.03 & 13.55@xmath70.03 & + 262 - 202046 & & 16.88@xmath70.04 & 15.73@xmath70.01 & + 262 - 202047 & & 17.26@xmath70.05 & 15.96@xmath70.01 & + 262 - 202050 & & 16.52@xmath70.03 & 15.22@xmath70.01 & + 262 - 202052 & 15.33@xmath70.04 & 14.99@xmath70.02 & 14.20@xmath70.03 & + 262 - 202054 & 16.21@xmath70.05 & 15.74@xmath70.02 & 14.76@xmath70.02 & + 262 - 202059 & 14.90@xmath70.04 & 13.90@xmath70.01 & 11.51@xmath70.04 & + 262 - 202061 & & 17.47@xmath70.05 & 16.64@xmath70.01 & + 262 - 202066 & 14.98@xmath70.04 & 14.73@xmath70.02 & 13.88@xmath70.03 & + 262 - 202067 & 14.64@xmath70.04 & 14.46@xmath70.01 & 13.75@xmath70.02 & + 262 - 202070 & 10.19@xmath70.04 & 10.57@xmath70.01 & 10.33@xmath70.01 & yes + 262 - 202088 & 13.09@xmath70.04 & 13.03@xmath70.01 & 12.48@xmath70.02 & + 262 - 202089 & 16.76@xmath70.07 & 16.24@xmath70.03 & 15.13@xmath70.02 & + 262 - 202097 & & 16.21@xmath70.03 & 14.60@xmath70.01 & + 262 - 202098 & 15.69@xmath70.05 & 15.35@xmath70.02 & 14.53@xmath70.01 & + 262 - 202100 & & 17.24@xmath70.05 & 15.92@xmath70.02 & + 262 - 202107 & 16.02@xmath70.05 & 15.67@xmath70.02 & 14.95@xmath70.02 & + 262 - 202109 & & 16.33@xmath70.03 & 15.31@xmath70.02 & + 262 - 202117 & 13.98@xmath70.04 & 13.95@xmath70.01 & 13.00@xmath70.02 & + 262 - 202118 & & 16.08@xmath70.05 & 15.53@xmath70.01 & + 262 - 202120 & & 15.26@xmath70.02 & 14.11@xmath70.02 & + 262 - 202128 & 13.56@xmath70.04 & 13.56@xmath70.01 & 12.62@xmath70.01 & + 262 - 202135 & 16.41@xmath70.05 & 16.15@xmath70.02 & 15.12@xmath70.01 & + 262 - 202141 & 16.66@xmath70.09 & 15.72@xmath70.03 & 14.48@xmath70.01 & + 262 - 202146 & & 16.35@xmath70.03 & 15.13@xmath70.04 & + 262 - 202155 & 12.62@xmath70.04 & 12.67@xmath70.01 & 12.39@xmath70.03 & yes + 262 - 202156 & & 16.79@xmath70.03 & 15.72@xmath70.02 & + 262 - 202159 & 16.45@xmath70.05 & 16.09@xmath70.02 & 15.08@xmath70.01 & + 262 - 202161 & & 17.00@xmath70.03 & 16.07@xmath70.02 & + 262 - 202166 & 14.60@xmath70.04 & 14.44@xmath70.02 & 13.88@xmath70.02 & + 262 - 202170 & 11.59@xmath70.04 & 11.63@xmath70.01 & 10.68@xmath70.03 & + 262 - 202175 & & 17.59@xmath70.08 & 16.33@xmath70.05 & + 262 - 202176 & 12.57@xmath70.04 & 12.61@xmath70.01 & 11.64@xmath70.01 & + 262 - 202177 & 15.22@xmath70.04 & 14.70@xmath70.02 & 13.65@xmath70.04 & + 262 - 202178 & & 16.73@xmath70.04 & 15.81@xmath70.01 & + 262 - 202185 & & 17.45@xmath70.06 & 16.09@xmath70.01 & + 262 - 202187 & & 17.41@xmath70.06 & 16.30@xmath70.01 & + 262 - 202190 & 14.87@xmath70.04 & 14.81@xmath70.01 & 13.80@xmath70.01 & + 262 - 202194 & 15.97@xmath70.05 & 15.68@xmath70.02 & 14.71@xmath70.03 & + 262 - 202197 & & 16.89@xmath70.04 & 15.79@xmath70.01 & + 262 - 202199 & 14.34@xmath70.04 & 14.18@xmath70.01 & 13.13@xmath70.01 & + 262 - 202207 & 14.56@xmath70.04 & 14.39@xmath70.01 & 13.68@xmath70.01 & + 262 - 202212 & 15.56@xmath70.04 & 15.41@xmath70.02 & 14.32@xmath70.02 & + 262 - 202218 & 13.14@xmath70.04 & 13.21@xmath70.01 & 12.33@xmath70.01 & + 262 - 202225 & & 16.96@xmath70.04 & 15.83@xmath70.02 & + 262 - 202224 & & 16.75@xmath70.03 & 15.66@xmath70.01 & + 262 - 202227 & 16.31@xmath70.05 & 15.84@xmath70.02 & 14.89@xmath70.06 & + 262 - 202236 & 16.47@xmath70.05 & 16.18@xmath70.03 & 15.41@xmath70.01 & + 262 - 202239 & & 17.00@xmath70.04 & 15.81@xmath70.04 & + 262 - 202243 & 16.16@xmath70.05 & 15.65@xmath70.02 & 14.80@xmath70.01 & + 262 - 202247 & 13.70@xmath70.04 & 13.72@xmath70.01 & 12.85@xmath70.03 & + 262 - 202248 & 13.57@xmath70.04 & 13.34@xmath70.02 & 12.62@xmath70.07 & + 262 - 202255 & & 16.40@xmath70.03 & 15.56@xmath70.01 & + 262 - 202256 & & 17.86@xmath70.10 & 16.15@xmath70.03 & + 262 - 202260 & & 16.10@xmath70.03 & 14.84@xmath70.01 & + 262 - 202265 & 14.56@xmath70.04 & 14.08@xmath70.02 & 13.27@xmath70.01 & + 262 - 202269 & 12.59@xmath70.04 & 12.67@xmath70.02 & 12.38@xmath70.02 & yes +